265. The Process.—In the treatment of sewage by the activated sludge process the sewage enters an aËration tank after it has been screened and grit has been removed. As it enters the aËration tank it is mixed with about 30 per cent of its volume of activated sludge. The sewage passes through the aËration tank in about two to four hours during which time air is blown through it in finely divided bubbles. The effluent from the aËration tank passes to a sedimentation tank where it remains for one-half an hour to an hour to allow the sedimentation of the activated sludge. The supernatant liquid from the sedimentation tank is passed to the point of final disposal. A portion of the sludge removed from the tank is returned to the influent of the aËration tank. The remainder may be sent to any or all of the following: the sludge drying process, the reaËration tanks, or to some point for final disposal. Sections of the activated sludge plant at Houston, Texas, are shown in Fig. 177.

The biological changes in the process occur in the aËration tank. These changes are dependent on the aËrobic organisms which are intensively cultivated in the activated sludge. When placed in intimate contact with fresh sewage, brought about by the agitation caused by the rising air, and in the presence of an abundance of oxygen, the organic matter is partially oxidized. The putrefactive stage of the organic cycle is avoided. Colloids and bacteria are partially removed probably by the agitation effected in the presence of activated sludge but the exact action which takes place is not well understood.

266. Composition.—Activated sludge is the material obtained by agitating ordinary sewage with air until the sludge has assumed a flocculent appearance, will settle quickly, and contain aËrobic and facultative bacteria in such numbers that similar characteristics can be readily imparted to ordinary sewage sludge when agitated with air in the presence of activated sludge. Copeland described activated sludge as follows:^[172]

The sludge embodied in sewage and consisting of suspended organic solids, including those of a colloidal nature, when agitated with air for a sufficient period assumes a flocculent appearance very similar to small pieces of sponge. AËrobic and facultative bacteria gather in these flocculi in immense numbers—from 12 to 14 million per c.c.—some having been strained from the sewage and others developed by natural growth. Among the latter are species that have the power to decompose organic matter, especially of an albuminoid or nitrogenous nature, setting the nitrogen free; and others absorbing the nitrogen convert it into nitrites and nitrates. These biological processes require time, air, and favorable environment such as suitable temperature, food supply and sufficient agitation to distribute them throughout all parts of the sewage.

Ardern states that the sludge differs entirely from the usual tank sludge. It is inoffensive and flocculent in character. The percentage of moisture is from 95 to 99 per cent. American experience has generally been that the sludge does not readily separate from its moisture by treatment on fine-grain filters, but the results in England and at Milwaukee, Wisconsin, are in conflict with this general experience. Upon standing 24 hours or more partially dried activated sludge may start to decompose accompanied by the production of offensive odors.

Duckworth states:

The activated sludge at Salford contained three times as much nitrogen, twice as much phosphoric acid and one-half as much fatty matter as ordinary sludge.

These results have been roughly checked by American experimenters as shown in Table 91.^[173] In the recovery of nitrogen from sewage the activated sludge process is the most promising for satisfactory results. In all other processes of sewage treatment the sludge is digested to some extent and nitrogen lost in the gases or in the soluble matter which passes off with the effluent. In the activated sludge process a negligible amount of gasification and liquefaction take place and only a small amount of nitrogen passes off with the effluent as compared with the loss from the Imhoff process as shown in Table 91. The percentage of nitrogen in dried activated sludge is shown in Table 92.

Nitrifying bacteria and other species which have the power of destroying organic matter have been isolated from the sludge. An analysis of the dried sludge at Urbana^[174] showed the following results after the weight had been reduced 95.5 per cent by drying: 6.3 per cent nitrogen, 4.00 per cent fat, 1.44 per cent phosphorus, and 75 per cent volatile matter or loss on ignition. Analyses of other domestic sewages have not shown such high contents of these desirable constituents.

The dewatering of activated sludge is a problem which offers serious obstacles to the successful operation of the process. It is its greatest disadvantage. Five to ten times the volume of sludge may be produced by the activated sludge process as by an Imhoff tank, and the activated sludge contains a greater percentage of water. According to Copeland:

The best information now available points to a combination of settling and decantation as a preliminary dewatering process. By this means the water will be cut down from about 99 per cent to 96 per cent. On passing the concentrated residue through a pressure filter the moisture can be cut down to 75 per cent. The press cake can be dewatered in a heat drier to 10 per cent moisture or less.^[175]

The quantity of sludge produced at Milwaukee^[176] is about 15 cubic yards per million gallons of sewage, the sludge having about 98 per cent moisture. On the basis of 10 per cent moisture it produces ½ ton of dry sludge per million gallons of sewage treated. At Cleveland,^[177] 20 cubic yards per million gallons at 97.5 per cent moisture are produced. Methods of drying sludge are discussed in Chapter XX.

Chemical analyses and biological tests indicate that the fertilizing value of the sludge is appreciable. Professor C. B. Lipman states, as the result of a series of tests in which a sludge and a soil were incubated for one month, as follows:^[178]

The amounts of nitrates produced in one month’s incubation from the soil’s own nitrogen and from the nitrogen from the sludge mixed with the soil in the ratio of one part of sludge to 100 of soil is, in milligrams of nitrate, as follows: Anaheim soil without sludge 6.0, with sludge 10.0; Davis soil without sludge 4.2, with sludge 14.0; Oakley soil without sludge 2.2, with sludge 4.0.

The effect of the sludge on plant growth is shown in Table 93.^[179] The results represent the growth obtained after fifteen weeks from the planting of 30 wheat seeds in each pot.

267. Advantages and Disadvantages.—Some of the advantages of the process are: a clear, sparkling, and non-putrescible effluent is obtained; the degree of nitrification is controllable within certain limits; the character of the effluent can be varied to accord with the quantity and character of the diluting water available; more than 90 per cent of the bacteria can be removed; the cost of installation is relatively low; and the sludge has some commercial value.

Among the disadvantages of the process can be included, uncertainty due to the lack of information concerning the results to be expected under all conditions, high cost of operation under certain conditions, the necessity for constant and skilled attendance, and the difficulty of dewatering the sludge.

268. Historical.—The most notable work in the aËration of sewage within recent years was that performed by Black and Phelps for the Metropolitan Sewerage Commission of New York, in 1910,^[180] and by Clark and Gage at the Lawrence, Massachusetts, Sewage Experiment Station in 1912 and 1913.^[181] The results of these investigations showed that the treatment of sewage by forced aËration might give a satisfactory effluent, but that the time and expense in connection thereto rendered the method impractical.

It remained for Messrs. Ardern and Lockett of Manchester, England, to introduce the process of the aËration of sewage in the presence of activated sludge, as a result of their connection with Dr. Fowler, who attributes his inspiration to his visit to the Lawrence Experiment Station and observing the work of Clark and Gage. Ardern and Lockett commenced their experiments in 1913. Their results were published in the Journal of the Society of Chemical Industry, May 30, 1914, Vol. 33, p. 523. Shortly thereafter experiments were started at the University of Illinois by Dr. Edw. Bartow and Mr. F. W. Mohlmann of the Illinois State Water Survey. At about the same time an experimental plant was started at Milwaukee, by T. C. Hatton, Chief Engineer of the Milwaukee Sewerage Commission. The United States Public Health Service became actively interested in December, 1914, and on February 20, 1915, announced its intention to co-operate with the Baltimore Sewerage Commission in the conduct of experiments. In May, 1915, patent number 1,139,024 was granted to Leslie C. Frank, Sanitary Engineer of the U. S. Public Health Service, covering certain features of the process. Mr. Frank generously donated this patent to the public for the use of municipalities.

The first full sized plant for the treatment of sewage by this method was erected in Milwaukee in December, 1915. This plant had a capacity of 1,600,000 gallons per day. It was used for experimental purposes and is not now in use. The Champaign, Illinois, septic tank, among the first of its kind in the country, was converted into an activated sludge tank on April 13, 1916. The changes, developments, and the results obtained from these and other plants have been reported in the technical press from time to time.

269. AËration Tank.—The sewage on leaving the screen and grit chamber enters the aËration tank, which is usually operated on the continuous-flow principle, although in the early days of experimentation the fill and draw method was practiced. This tank should be rectangular with a depth of about 15 feet and a width of channel not to exceed 6 to 8 feet. Such proportions allow better air and current distribution than larger tanks. The bottom should be level to insure an even distribution of air. The velocity of flow of sewage through the tank is usually in the neighborhood of 5 feet per minute, dependent on the length of the tank and the period of retention. The period of retention is in turn dependent on the desired quality of the effluent. The process is flexible and the quality of the effluent can be changed by changing the period of retention or by changing the rate of application of the air, or both. The period of retention in the aËration tank is usually about 4 hours.

The bottom of the aËration tank is usually made of concrete arranged in ridges and valleys, or small shallow hoppers, at the bottom of which the air-diffusing devices are located, as shown in Fig. 177. The inlet and outlet devices are similar to those in a plain sedimentation tank.

270. Sedimentation Tank.—It is evident that as no sedimentation is permitted in the aËration tank, the settleable particles will be discharged in the effluent unless some provision is made for their detention. The effluent from the aËration tank is therefore run through a plain sedimentation tank, usually with a hopper bottom, which has been arranged to permit frequent and easy cleaning. An air lift or a centrifugal sludge pump is satisfactory for this purpose. Another type of sedimentation tank which has been used has a smooth bottom with a slight slope towards the center. A revolving scraper collects the sludge continuously, scraping it towards the center of the tank. Although this arrangement gives better results than the hopper-bottom tank, its expense has usually prevented its installation.^[182]

The period of sedimentation in different plants varies from 30 minutes to one hour, although the longer periods usually give the better results. Approximately 65 per cent of the sludge will settle in the first 10 minutes, 80 per cent in the first 30 minutes, and about 5 per cent more in the next half hour.

The effluent from the sedimentation tank is ready for final disposal or if desired, for further treatment by some other method. The sludge, or a portion of it, is pumped back into the influent of the aËration tank, provided the sludge is in a satisfactory state of nitrification. Otherwise it should be pumped to the reaËration tanks. The remainder of the sludge which is not to be used in the process is ready for drying and final disposal.

271. ReaËration Tank.—The purpose of the reaËration or sludge aËration tank is to reactivate the sludge which has gone through the aËration tank. During the process of the aËration of the sewage in the aËration tank the activated sludge may lose some of its qualities because of the deficiency of oxygen to maintain aËrobic conditions. By blowing air through the sludge in the reaËration tank these properties are returned and the sludge made available to be pumped back into the aËration tank. The reactivation of the sludge obviates the necessity for supplying sufficient air to the entire mass of the sewage to maintain aËrobic conditions, and results in an economy in the use of air. The use of mechanical agitators has also been attempted both in the reaËration and the aËration tanks with the expectation of saving in the use of air, but with indifferent success.

It is difficult to say, without experimentation, what the size of the reaËration tank should be, as the necessary amount or reactivation is uncertain. In the experimental plant at Milwaukee, there were eight units of aËration tanks, one sedimentation tank, and two reaËration tanks, all of the same capacity and general design. This represents a ration of about one reaËration tank to four aËration tanks.

272. Air Distribution.—Air is applied to the sewage at the bottom of the aËration tank at a pressure in the neighborhood of 5.5 to 6.0 pounds per square inch, dependent on the depth of the sewage, the loss of head through the distributing pipes, and the rate of application. In different experimental plants the pressure has varied from 3 to 30 pounds per square inch. Such pressures are on the line which divides the use of direct blowers for low pressures from turbo and reciprocating pressure machines for pressures above 10 pounds per square inch. Positive-pressure blowers or direct blowers operate on the principle of a centrifugal pump and because of the lighter specific gravity of air they rotate at a very high speed. The Nash Hytor Turbo Blower consists of a rotor with a large number of long teeth slightly bent in the direction of rotation. The rotor, which has a circular circumference, revolves in an elliptical casing. At the commencement of operation the rotor and casing are partially filled with water. The revolution of the rotor throws the water to the outside of the elliptical casing thus forming a partial vacuum between any two teeth as the water is thrown from near the center of the short diameter of the casing to the extremity of the long diameter of the casing. Air is allowed to enter through the inlet port to relieve the vacuum. As the teeth pass from the long diameter to the short diameter of the ellipse, the water again approaches the center of the rotor compressing the air trapped between the teeth and forcing it out under pressure into the exhaust pipe. Among the advantages of this compressor are the washing of the air, cooling, and ease in operation. Reciprocating air compressors operate similarly to direct-acting steam pumps or crank-and-fly-wheel pumps but at much higher speeds, and they require more floor space than either of the other types. Fig. 178 shows the field of serviceability of various types of air compression machinery.

For pressures up to about 10 pounds per square inch the positive blower seems most desirable. It has a low first cost and a relatively high efficiency of about 75 to 80 per cent of the power input. No oil or dirt is added to the air to clog the distributing plates, as in the reciprocating machine. A disadvantage is the difficulty of varying the pressure or quantity of the output of the machine. As the required pressure and volume of air increases the turbo blower becomes more and more desirable within the limits of pressure which are ordinarily used in this process. For small installations the best form of power is probably the electric drive, but when the capacity becomes such as to make turbo blowers advisable they should be driven by directly connected steam turbines.

The quantity of air required varies between 0.5 to 6.0 cubic feet per gallon of sewage, with from 3 to 6 hours of aËration. The quantity of air depends on the degree of treatment required, the strength of the sewage, the depth of the tank, and the period of aËration. The deeper the tank the less the amount of air needed because of the greater travel of the bubble in passing through the sewage, but the higher the pressure at which the air must be delivered. Shallow tanks usually require a longer period of retention. The depth of the tank then has very little to do with economy in the use of air. Hatton states:^[183]

The purification of sewage obtained varies decidedly with the volume of air applied. Small volumes applied for 5 or 6 hours do as well as larger volumes applied for 3 or 4 hours, but the time of aËration required to obtain a like effluent does not vary directly with the volume of air applied per unit of time. For instance air applied at a rate of 2 cubic feet per minute purifies the sewage in less time than one cubic foot of air per minute, but will not accomplish an equal degree of purification in half the time.

It has been found that although a low temperature has a deleterious effect on the process, by the use of an additional quantity of air good results can be maintained. The effect of changing the quantity of air and the period of aËration are shown in Table 94 taken from Hatton.

The velocity of the air in the pipes should be about 1,000 feet per minute. There should be relatively few sharp turns in the line, and the distributing mains should be arranged without dead ends. It is desirable to use as little piping as possible and at the same time to make the travel of the sewage long in order to maintain a non-settling velocity and intimate contact with the air. The piping should be accessible and well provided with valves. It should be non-corrodible, particularly on the inside, as flakes of rust will quickly clog the air diffusers. It should drain to one point in order that it can be emptied when flooded, as occasionally happens.

It is desirable to diffuse the air in small bubbles as by this means the greatest efficiency seems to be obtained from the amount of air added. A diameter 1
16 to ? of an inch is approximately the maximum limit for the size of an effective bubble. Monel metal cloth, porous wood blocks, open jets, paddles, and other forms of diffusers have been tried, but none have given the satisfaction of the filtros plate. The relative value of different types of diffusers is shown in Table 95 taken from Hatton.^[184] The Filtros plates are a proprietary article manufactured by the General Filtration Company of Rochester, N. Y. They are made of a quartz sand firmly cemented together and can be obtained with practically any degree of porosity, size of pore opening or dimension of plate, but they are made in a standard size 12 inches square by 1½ inches thick. The frictional loss through the plate is not very great for the amount of air ordinarily used. The plates are classified in accordance with the volume of air which will pass through them, when dry, per minute when under a pressure of 2 inches of water. These classes run from ½ to 12 cubic feet of air per minute. The type usually specified passes about 2 cubic feet of air per minute. The loss of head through these plates as tested at Milwaukee showed an initial loss of ¾ of a pound and an additional loss of about ¼ of a pound for every cubic foot of air per minute per square foot of surface. It is necessary to screen and wash the air before blowing it through the filtros plate as ordinary air is so filled with dirt as to clog the pores of the diffuser quite rapidly.

The area of filtros plates required in the bottom of the tank is usually expressed in terms of the free surface of the tank or as a ratio thereto. In the Urbana tests the best ratio was found to be less than 1 : 3 and more than 1 : 9. In Milwaukee^[185] the ratio adopted is in the neighborhood of 1 : 4 or 1 : 5. At Fort Worth the ratio will be about 1 : 7 and at Chicago it will be 1 : 8. The exact ratio should be determined by experiment and will depend on the construction of the tank and the character of the raw sewage and the desired effluent. It is essential that the filtros plates be placed level and at the same elevation as otherwise the distribution of air will be uneven.

273. Obtaining Activated Sludge.—After a plant is once started activated sludge is generated during the process of treatment and with careful management a stock of activated sludge can be kept on hand. When a plant is new, or if shut down for such a length of time that the sludge loses its activation, it is necessary to activate some new sludge. This is done by blowing air continuously through sewage either on the fill and draw method with periodic decantations of the supernatant liquid, or by the continuous-flow process, but more preferably by the latter. Where activated sludge is to be obtained from fresh sewage alone the time required is in the neighborhood of 10 to 14 days, and purification begins at the start. An estimate of the quantity which will be obtained can not be made with accuracy. After the initial quantity of sludge has been obtained activated sludge can be maintained during the process of aËration of the raw sewage, or by means of the reaËration tanks previously described.

The volume of activated sludge present in the aËration tank should be about 25 per cent of the volume of the tank. The volume of the sludge is measured in a somewhat arbitrary manner as the amount by volume which will settle in 30 minutes in an ordinary test tube. It is found that this is almost 90 per cent of the solids settling in 4 to 6 hours.

274. Cost.—The available information on the cost of the activated sludge process is meager and unreliable. The factors entering into the cost are: the price of fuel, the size of the plant, the period of sedimentation, the amount of air per gallon of sewage, the air pressure, and the percentage of sludge to be aËrated in the mixture. In Milwaukee^[186] the cost of construction is estimated at $44,000 per million gallons, and $4.75 per million gallons for operation. At Houston, Texas, the cost is estimated at $24,000 per million gallons, exclusive of the sludge drying plant, which may cost $40,000 per million gallons. At Milwaukee, the cost of pressing the sludge is $4.82 per dry ton and of drying is $3.93 per dry ton. The sludge may be sold at the normal rate of $2.50 per unit of nitrogen. Based on the normal value the evident profit will be $3.75 per ton. The net cost of disposing of Milwaukee sewage is estimated at $9.64 per million gallons of which $4.89 is chargeable to overhead and $4.75 to repairs, operation and renewal. In a comparison of the costs of activated sludge and Imhoff tanks with sprinkling filters,^[187] the information given by Eddy has been summarized in Table 96. In comparing the relative areas required for different methods of sewage treatment, activated sludge should be allowed about 15 million gallons per acre per day on the basis of aËration tanks 15 feet deep. This figure represents approximately the gross area of the plants at Milwaukee and at Cleveland.

The following abbreviations will be used: A.S. for Activated Sludge, E.C. for Engineering and Contracting, E.N. for Engineering News, E.R. for Engineering Record, E.N.R. for Engineering News-Record, p. for page, and V. for volume.

No.

1.

Cooperation Sought in Conducting A.S. Experiments at Baltimore, by Franks and Hendrick. E.R. V. 71, 1915, pp. 521, 724, and 784. V. 72, 1915, pp. 23, and 640.

2.

Sewage Treatment Experiments with AËration and A.S., by Bartow and Mohlman. E.N. V. 73, 1915, p. 647, and E.R. V. 71, 1915, p. 421.

3.

A.S. Experiments at Milwaukee, Wisconsin, by Hatton. E.N. V. 74, 1915, p. 134.

4.

A.S. in America, An Editorial Survey, by Baker. E.N. V. 74, 1915, p. 164.

5.

Choosing Air Compressors for A.S., by Nordell, E.N. V. 74, 1915, p. 904.

6.

A Year of A.S. at Milwaukee, by Fuller. E.N. V. 74, 1915, p. 1146.

7.

A.S. Experiments at Urbana. E.N. V. 74, 1915, p. 1097.

8.

Experiments on the A.S. Process, by Bartow and Mohlman. E.C. V. 44, 1915, p. 433.

9.

Milwaukee’s A.S. Plant, the Pioneer Large Scale Installation, by Hatton. E.R. V. 72, 1915, p. 481 and E.C. V. 44, 1915, p. 322.

10.

A.S. Experiments at Milwaukee, by Hatton. Journal American Waterworks Association and Proceedings Illinois Society of Engineers, 1916. Also E.R. V. 73, 1916, p. 255. E.C. V. 45, 1916, p. 104, and E.N. V. 75, 1916, pp. 262 and 306.

11.

A.S. Defined. E.N. V. 75, 1916, p. 503, and E.N.R. V. 80, 1918, p. 205.

12.

Status of A.S. Sewage Treatment, by Hammond. E.N. V. 75, 1916, p. 798.

13.

Trial A.S. Unit at Cleveland, by Pratt. E.N. V. 75, 1916, p. 671.

14.

Air Diffuser Experience with A.S. E.N. V. 76, 1916, p. 106.

15.

Nitrogen from Sewage Sludge, Plain and Activated, by Copeland, Journal American Chemical Society, Sept. 28, 1916. E.N. V. 76, 1916, p. 665. E.R. V. 74, 1916, p. 444.

16.

Tests Show A.S. Process Adapted to Treatment of Stock Yards Wastes. E.R. V. 74, 1916, p. 137.

17.

AËration Suggestions for Disposal of Sludge, by Hammond. Journal American Chemical Society, Sept. 25, 1916. E.R. V. 74, 1916, p. 448.

18.

Cost Comparison of Sewage Treatment. Imhoff Tank and Sprinkling Filters vs. A.S., by Eddy. E.R. V. 74, 1916, p. 557.

19.

Large A.S. Plant at Milwaukee. E.N. V. 76, 1916, p. 686.

20.

A.S. Novelties at Hermosa Beach, Cal. E.N. V. 76, 1916, p. 890.

21.

A.S. Experiments at University of Illinois, by Bartow, Mohlman, and Schnellbach. E.N. V. 76, 1916, p. 972.

22.

A.S. Results at Cleveland Reviewed, by Pratt and Gascoigne. E.N. V. 76, 1916, pp. 1061 and 1124.

23.

Sewage Treatment by AËration and Activation, by Hammond. Proceedings American Society Municipal Improvements, 1916.

24.

A.S., by Bartow and Mohlman, Proceedings Illinois Society of Engineers, 1916.

25.

The Latest Method of Sewage Treatment, by Bartow. Journal American Waterworks Association, V. 3, March, 1916, p. 327.

26.

Winter Experiences with A.S., by Copeland. Journal American Society of Chemical Engineers, April 21, 1916. E.C. V. 45, 1916, p. 386.

27.

A.S. Process Firmly Established, by Hatton. E.R. V. 75, 1917, p. 16.

28.

Operate Continuous Flow A.S. Plant, by Bartow, Mohlman, and Schnellbach. E.R. V. 75, 1917, p. 380.

29.

Chicago Stock Yards Sewage and A.S., by Lederer. Journal American Society of Chemical Engineers, April 21, 1916. E.C. V. 45, 1916, p. 388.

30.

The Patent Situation Concerning A.S. E.C. V. 45, 1916, p. 208.

31.

“Sewage Disposal” by Kinnicutt, Winslow, and Pratt, published by John Wiley & Sons. 2d Edition, Chapter 12.

32.

A.S. Tests Made by California Cities. E.N.R. V. 79, 1917, p. 1009.

33.

Conclusions on the A.S. Process at Milwaukee. Journal American Public Health Association, 1917. E.N.R. V. 79, 1917, p. 840.

34.

Dewatering A.S. at Urbana, by Bartow. Journal American Institute of Chemical Engineers, 1917. E.N.R. V. 79, 1917, p. 269.

35.

Milwaukee Air Diffusion Studies in A.S. E.N.R. V. 78, 1917, p. 628.

36.

A.S. Bibliography (up to May 1, 1917) by J. E. Porter.

37.

Air Diffusion in A.S. E.N.R. V. 78, 1917, p. 255.

38.

A.S. Plant at Houston, Texas. E.N. V. 77, 1917, p. 236, E.N.R. 83, 1919, p. 1003, and V. 84, 1920, p. 75.

39.

A.S. Power Costs, by Requardt. E.N. V. 77, 1917, p. 18.

40.

A.S. at San Marcos, Texas, by Elrod. E.N. V. 77, 1917, p. 249.

41.

Filtros Plates Made the Best Showing in Air Diffuser Tests. E.N.R. V. 79, 1917, p. 269.

42.

Results of Experiments on A.S., by Ardern and Lockett. Journal Society for Chemical Research, V. 33, May 30, 1914, p. 523.

43.

Final Plans at Milwaukee. E.N.R. V. 84, 1920, p. 990.

44.

A.S. Bibliography, published by General Filtration Co., Rochester, N. Y., 1921.

45.

A.S. at Manchester, Eng. by Ardern. Journal Society Chemical Industry, 1921. E.C. V. 55, 1921, p. 310.

46.

The Des Plaines River A.S. Plant, by Pearse. E.N.R. V. 88, 1920, p. 1134.

47.

Sewage Treatment by the Dorr System, by Eagles. Proceedings, Boston Society of Engineers, 1920. Public Works V. 50, 1920, p. 53.