<mff&MHEffiEir 
 
 HX64077268 
 
 RA585H17 Brooklyn, N.Y., sewa 
 
Columbia (Bnftetsftp 
 
 College of iPfrpStcians anb Hmrgeons 
 Uabrarp 
 
Digitized by the Internet Archive 
 
 in 2010 with funding from 
 
 Open Knowledge Commons 
 
 http://www.archive.org/details/brooklynnysewageOOhamm 
 
BROOKLYN, N. Y., SEWAGE 
 TREATMENT EXPERIMENTS 
 
 irtrtrniFnTTTTniniiitniiiiiiitfifitiiHiMMrirniUMniuiiiiinrfiiuiiiiirniiinMiiiuniiiiiuiHii iriiiiiiiiiiniMiriiHiuiiiiiiDiniininHiiiiiiniiniuiiniiiiiiitiirifnuitiuiiiiiiiiiirtTmn 
 
 By 
 
 GEORGE T. HAMMOND, M. Am. Soc. C. E. 
 
 156 BERKELEY PLACE, BROOKLYN, N. Y. 
 
 iiiiuiiiiiuiiiiiiiMiiiinniiiiiiuiiiiiiiiiiiiiiiiiiti iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiimiHniiuiuiiiiiiimiiiiiuiiiiiiiuiiiiiimi mi ihiuiiiihiuiiihiiui 
 
 REPRINTED FROM THE 
 
 PROCEEDINGS OF THE AMERICAN SOCIETY FOR MUNICIPAL IMPROVEMENTS 
 
 NINETEEN HUNDRED AND NINETEEN 
 
BROOKLYN, N. Y., SEWAGE 
 TREATMENT EXPERIMENTS 
 
 nuifniiiiiiiiiiiiiiiiiifiiiiiiMiiiiiHiiiiuiiiiiiiiiiMiiiiiriiiiiiiiiiiiiiiiiniiiiiiiiiiiiMiiiiiiiiiiiiiiiiiiiiiMiiiiiHiMinHMMiiiiiiiiiiiiMiiifiiiiiiiiiiiiiMiiiiiiiiiiiiiifiiiiiiiiiiiiiiiiiiininin 
 
 A Brief Review of Five Year s Work 
 
 By 
 GEORGE T. HAMMOND, M. Am. Soc. C. E. 
 
 156 BERKELEY PLACE, BROOKLYN, N. Y. 
 
 Engineer in Charge of Experiments 
 and of Sewer Design, Brooklyn, N. Y. 
 
 Surgeon (Reserve) U.S. Public 
 Health Service 
 
 REPRINTED FROM THE 
 
 PROCEEDINGS OF THE AMERICAN SOCIETY FOR MUNICIPAL IMPROVEMENTS 
 
 NINETEEN HUNDRED AND NINETEEN 
 

BROOKLYN, N. Y. SEWAGE TREATMENT 
 EXPERIMENTS 
 
 A Brief Review of Five Years' Work. 
 
 By George T. Hammond, Surgeon (Reserve), U. S. Public Health 
 Service, Consulting Engineer, 156 Berkeley Place, Brooklyn, 
 
 N. Y. 
 
 Introduction. 
 
 Bearing in mind the brevity required in a paper intended for 
 presentation at a Convention of this Society, it seems proper to 
 begin it with a statement concerning the subject matter and the 
 limits within which it is confined. 
 
 The Boro of Brooklyn has been making sewage treatment 
 experiments for a number of years past. These experiments 
 were interrupted and brought to a close by the war, at the begin- 
 ning of 1918, when the writer left the city's service to enter the 
 war service of the U. S. Government, as sanitary engineer to the 
 U. S. Shipping Board, Emergency Fleet Corporation, and later 
 to the U. S. Public Health Service. 
 
 Before leaving his duties with the city, the writer made a com- 
 pilation of the data obtained from the work of the sewage treat- 
 ment experiment station, at the request of Mr. E. J. Fort, chief 
 engineer of sewers, who, as one of his last official duties before 
 leaving the city's service to become city manager of Niagara 
 Falls, prepared during the summer of 1918, a final report on the 
 26th Ward Sewage disposal, for presentation to the boro officials. 
 The writer is informed that this report is being prepared for 
 the printer and will soon be available to those who are interested 
 in the details of the work, the conclusions reached, and the recom- 
 mendations made for the solution of the problems for which 
 the work was undertaken. 
 
 Soon after the work of the Experiment Station was initiated, 
 Mr. Fort informed the writer that he had promised to have pre- 
 sented to this society such information in the form of papers 
 as might be available at the time of the annual conventions. It 
 was in accordance with his instructions that the following papers 
 were prepared and presented by the writer : 
 
 Sezvage Treatment Experimental Plant in Brooklyn, N. Y. 
 Presented in 1914 at the Boston Convention. 
 
 Sewage Treatment bx Aeration and Activation. Presented in 
 1916 at the Newark Convention. 
 
 To complete the series of papers on this work, the present 
 paper was planned in 1917, and was to have been presented in 
 
 3 
 
4 Brooklyn, N. Y. 
 
 1918; but on account of the war this intention could not be real- 
 ized. It has recently been revised and made considerably more 
 brief than it was as completed in the spring of 1918; and yet 
 apology seems due for its considerable length. 
 
 The purpose of the writer has been to treat the subject wholly 
 as a study in sewage disposal, far removed from any official 
 character. If it were possible to eliminate from it all reference 
 to the special problems for which the plant was provided, this 
 would be done, as the important matter for us in such a study 
 is to look for the underlying principles of general application, if 
 any such there may be. If such principles are not present, the 
 study can be of but little if any interest. But concrete facts' are 
 necessary for illustrative data. We should, however, beware of 
 data so local and dependent upon the concrete conditions afforded 
 by local problems that they would not apply elsewhere. The 
 final report on this work, when published, will go over the entire 
 ground from the standpoint of the local problems and with all 
 the difficulties involved in meeting local conditions. It will con- 
 tain much of no interest to the engineering profession, along 
 with many facts and data out of which interesting general 
 principles can be laboriously extracted. With its local conclu- 
 sions and recommendations, this paper will have nothing to do. 
 But the attempt will be made herein to present some of the facts, 
 selected carefully from that voluminous mass of material, which 
 the writer hopes and believes will prove of interest as a basis 
 for general discussion, thereby carrying out the original idea 
 of Mr. Fort, as explained, when he asked the writer to prepare 
 the series of papers of which this is the final one. 
 
 When the experiment station was authorized and constructed, 
 and also thru its period of operation, Hon. Lewis H. Pounds 
 was President of the Boro of Brooklyn, going out of office 
 January 1, 1918, and Mr. E. J. Fort was Chief Engineer, from 
 which office he resigned to become City Manager of Niagara 
 Falls, N. Y., late in 1918. The writer was engineer of design 
 of the Bureau of Sewers and in charge of the experiment station 
 under Mr. Fort. Mr. W. T. Carpenter was chief chemist of the 
 plant 1913-1917, and Mr. Murray P. Horowitz, assistant chemist. 
 Mr. George H. Knight, assistant engineer, assisted the writer 
 in all engineering matters, and for a considerable time repre- 
 sented him at the plant. 
 
 Dr. Rudolph Hering gave expert advice and revised the 
 plans. Dr. Karl Imhofr revised the plans of the Imhoff tanks 
 and gave much other advice. Professor Earle B. Phelps gave 
 expert advice on sewage aeration. Much assistance was given 
 during the work by many visiting experts. 
 
 The points that will be briefly touched on are : 
 (1) Some general remarks about the plant and the local con- 
 ditions. 
 
Sewage Treatment Experiments 5 
 
 (2) A statement concerning the laboratory technic and sampling. 
 
 (3) The methods of measurement employed. 
 
 (4) The sewage, how obtained for the experiments, its quan- 
 
 tity per capita, and most important characteristics. 
 
 (5) Some remarks on the operation of Imhoff tanks, trickling 
 
 filters, and settling tanks, sludge digestion and sludge 
 drying, giving results obtained in very short retention 
 periods, and by very rapid operation of the filter beds. 
 
 (6) Some studies on the operation of Riensch-Wurl screens 
 
 and the application of screened sewage to trickling filters. 
 
 A Brief Review of the Work 
 
 Sewage investigations were commenced at this station in 1910 
 by Professor Earle B. Phelps, and General (then Colonel) Wil- 
 liam M. Black, Corps of Engineers U. S. A. Their report on 
 the aeration treatment of sewage, now a valued classic in sanita- 
 tion literature, was presented in 1911. The studies were con- 
 tinued by the Bureau of Sewers under Mr. E. J. Fort, assisted 
 by the writer, and in 1912 an extensive experimental plant was 
 authorized. At the same time a new sewage disposal plant was 
 provided for by the Board of Estimate — the design to await the 
 results of experimental study. Local conditions were given care- 
 ful study, and all American sewage disposal plants of note were 
 visited by Mr. Fort and the writer. 
 
 During the spring and summer of 1912, Mr. Fort and the 
 writer, in company with Dr. Rudolph Hering, visited the more 
 important sewage disposal plants in Great Britain and on the 
 European continent, — taking in all of the principal cities in Eng- 
 land, Germany, France, Switzerland, etc. — gathering data for 
 the design of the experimental units and also for the permanent 
 plant. 
 
 In 1913 the experimental units and laboratory were completed, 
 except the screens, which were not placed until 1916. Operation 
 of the three Imhoff tanks, plans of which had been revised and 
 approved by Dr. Karl Imhoff himself, was started in 1913, and 
 at the same time all of the trickling filters and aeration experi- 
 ments. This work was continued through 1914 and 1915, and 
 indeed until the close down of the plant in January, 1918. 
 
 Sewage aeration experiments were undertaken concerning ac- 
 tivated sludge from 1915 to 1918. 
 
 The Riensch-Wurl screens were installed in 1916, and experi- 
 ments with them continued to 1918. 
 
 There was much other experimental work done on various 
 other methods of sewage treatment, and on variations proposed in 
 the standard methods, but this work may be considered as rela- 
 tively unimportant, and led to no conclusions worth stating here. 
 
Brooklyn, N. Y. 
 
Sewage Treatment Experiments 7 
 
 It may be observed that every kind of an experimental unit 
 designed was itself an experiment. It was only by installing such 
 a plant that the actual conditions under which it would be oper- 
 ated could be fully discovered. It was anticipated by the writer 
 that as much could be learned from the troubles and limitations 
 discovered in operation, as from the successful performance of 
 the different units. Long and careful study, and many changes 
 in experimental units, are required in order to obtain dependable 
 results in these studies, and a careful recognition of existing con- 
 ditions and circumstances under which a plant shall operate, and 
 the adaptation of the plant to these conditions, is essential to the 
 success of these investigations. 
 
 The experiment station consisted of a laboratory for the chemi- 
 cal and bacteriological work connected with the study of sewage, 
 and a sewage treatment experimental plant. The size of each 
 of the various units of the plant was made large enough to permit 
 operation on a scale sufficiently great to take it out of the class of 
 a mere laboratory test. All of the work was carried out on such 
 a scale that each experiment was made under conditions designed 
 to approach actual sewage-disposal-plant conditions. 
 
 The plant included three Imhoff tanks of different depths, but 
 with other corresponding dimensions equal ; six trickling filter 
 beds, two of which were designed for discharging compressed air 
 within the mass of the stone medium ; various tanks and appara- 
 tus for the investigation of sewage treatment by forced aeration ; 
 secondary settling tanks for the trickling filters, and for the 
 sewage treated by aeration ; a plain settling tank for crude sewage 
 in connection with an airtight sludge digestion tank, which re- 
 ceived the settled matters and sludge from the settling tank, being 
 
 Fig. 2. View of Plant When Put in Service, Oct. 1, 1913. 
 
8 Brooklyn, N. Y. 
 
 in effect the two essential portions of an Imhoff tank separated ; 
 ten sludge-drying beds, and various units for experimenting with 
 fine-screened sewage, and for drying sludge. 
 
 The mechanical plant consisted of steam-actuated sewage 
 pumps and an air compressor ; with attachments and appliances 
 for measuring volumes of air delivered. 
 
 The pumps that furnished the sewage for the experiments were 
 located within the existing sewage disposal works buildings, and 
 steam was obtained from the boilers of that plant. There were 
 two sewage pumps, both of direct-acting piston type, so installed 
 that either one could be cut out and cleaned or repaired without 
 stopping the other. The larger pump had a capacity of 1,200,000 
 gallons per day, and the smaller 650,000 gallons. 
 
 The compressor installed was a duplex crank and fly wheel 
 machine ; automatic in starting, stopping and speed ; with a dis- 
 placement capacity of 228 cubic feet of air per minute at not ex- 
 ceeding 210 r.p.m. for 30 lb. air pressure and with 100 lbs. of 
 steam at the throttle. It was equipped with a combination speed 
 and pressure governor, arranged to bring the machine to a dead 
 stop if no air was demanded, and to start up automatically upon 
 drop of the pressure. 
 
 The first thing determined, in making the design, was the re- 
 quired elevation of the surfaces of the sewage in each unit of 
 the plant. This determined, the unit was designed to comply 
 with it. 
 
 The low elevation of the ground at the site, but a few inches 
 above ordinary high tide— a tidal marsh, in fact — necessitated 
 the placing of the experimental units above the reach of the 
 highest tide, and pumping the sewage up to such a level that a 
 gravity flow could be obtained for every unit of the plant. 
 
 The datum line was at mean high water and the surface of 
 the sludge beds was placed at an elevation of 2.67 feet above this 
 datum. All of the other units of the plant were given such eleva- 
 tions above this as their operation required. Every unit was pro- 
 vided with a measuring device, for the determination of the flow, 
 usually consisting of an adjustable, calibrated orifice above which 
 a constant head was maintained by a system of overflow weirs. 
 Surplus flow wasted back to the main sewer. There was always 
 a considerable surplus flow, so as to keep a good velocity in pas- 
 sages. 
 
 For measuring compressed air, Venturi meters were provided. 
 The accuracy of all the measuring devices was carefully tested in 
 place. 
 
 The construction of the plant, for the most part above the 
 surface of the ground, afforded opportunity for studying the 
 flow of sludges of different kinds, Imhoff especially, which could 
 not as easily have been studied otherwise ; also, the effect of cold 
 
Sewage Treatment Experiments 9 
 
 weather on exposed sprinkling filter beds ; the danger of freezing 
 of the various channels carrying sewage, and the proper method 
 of protecting and operating the same. These studies were of 
 much interest from the engineer's standpoint, in view of the pro- 
 jected construction of a trickling filter plant on concrete pile 
 foundations over an extensive marshland. 
 
 ADJUSTABLE ORIFICE 
 
 Fig. 3. 
 
 Laboratory Control and Sampling 
 
 The laboratory was erected as part of the experiment station, 
 and supplied with heat and power from the plant of the existing 
 sewage treatment works. It was provided with the usual ap- 
 paratus for sewage investigations and for the filing and classify- 
 ing of results. The laboratory work conformed to standard 
 practice. The standard methods of the American Public Health 
 Association were followed. The analytical work and record keep- 
 ing were organized by Mr. W. T. Carpenter, Chief Chemist, and 
 the methods introduced by him were followed thruout the entire 
 period. 
 
10 Brooklyn, X. Y. 
 
 Samples having to do with the "oxygen balance" of the sewage, 
 and various effluents, are from their nature, incapable of being 
 composited. Effort was therefore made to take them at such a 
 time of day that the sewage would be most likely to be of average 
 quality. The monthly averages from which conclusions were 
 drawn were averages of ten or more samples. 
 
 Samples for the determination of such constituents as are prop- 
 erly made on composites were collected as follows : 
 
 A full set of half-pint samples was taken every four hours, 
 midnight to midnight, and put into bottles with chloroform. 
 These bottles, being the primary composites, were renewed daily, 
 and a pint portion taken to form the secondary composites for 
 analysis. The analyses of the composites were never made less 
 frequently than five times monthly, and ten times for all the 
 more variable conditions of sewage flow. Monthly averages 
 were determined by the composition of about 250 individual por- 
 tions. 
 
 In order that there may be no misunderstanding as to the 
 meaning of certain terms used in this paper, which are not always 
 given the same definition, the following explanation, prepared by 
 Mr. Carpenter, is appended : 
 
 Settling Matter is the volume of matter, solid and water con- 
 tent, which will settle to the bottom of an Imhoff settling cone, 
 and is expressed in cubic centimeters per liter. 
 
 The time period used in all of these determinations was two 
 hours. After an hour's settling the cones were rolled between 
 the hands a few times, to send to the bottom such of the solids as 
 adhered to the sloping sides. 
 
 By Total Suspended Solids is meant the gravimetric amount 
 of dry material which will be retained upon the asbestos mat in a 
 Gooch crucible, after filtration with suction. Its weight is ex- 
 pressed in parts per million parts of the sewage from which it is 
 taken. 
 
 By Volatile Suspended Solids is meant that portion of the dry 
 material in the Total Suspended Solids which is lost by incin- 
 eration at cherry red heat, in a muffle furnace. It is expressed 
 in weight in parts per million, and by per cent, of the Total Sus- 
 pended Solids. 
 
 It is doubtless more than the organic matter, but the difference 
 is not readily determinable, and probably constant within very 
 narrow limits, and unimportant, as the figures are only of inter- 
 est in making comparisons. During the first three years the char- 
 acter of the suspended solids was analyzed closely to determine 
 what portion of the solids was settleable, and what portion non- 
 settleable, or colloidal. The Gooch-crucible method was used. 
 The total suspended solids were determined on an unsettled sam- 
 ple, and the colloidal solids on a sample siphoned from the upper 
 
Sewage Treatment Experiments 11 
 
 portion of the contents of an Imhoff cone after two hours' 
 sedimentation. The latter are called Colloidal Suspended Solids, 
 though they undoubtedly contain a minute and scarcely apprecia- 
 ble amount of material capable of settling in longer periods of 
 time. The Settling Suspended Solids are the difference between 
 the two. 
 
 Oxygen Consumed was determined in accordance with the 
 1913 Standard Methods of the American Public Health Associa- 
 tion, using a 30-minute digestion period, the heating being done 
 in an Arnold steam sterilizer. The determinations were made 
 on raw and on paper-filtered samples of sewage. 
 
 Nitrogen as Free Ammonia was obtained by direct nessleriza- 
 tion. 
 
 Organic Nitrogen was determined colormetrically by the 
 Kjeldahl process after direct nesslerization. 
 
 Nitrogen as Nitrites was determined according to "Standard 
 Methods". 
 
 Nitrogen as Nitrates was determined by the phenol-sulphonic 
 method. 
 
 This procedure was adopted after experimenting upon the 
 brucine and the narcotine methods. The reduction method was 
 considered too time-consuming to be justified with the force avail- 
 able. Care was used at all times to insure the evaporation of the 
 last drops of the sample in the air, instead of on the steam 
 bath, in order to avoid loss of nitrogen in consequence of a high 
 chlorine content. This determination is most significant where 
 oxidized effluents are concerned, and the nitrate figures are to 
 be interpreted as being at least the recorded values. The deter- 
 mination of both nitrite and nitrate nitrogen was preceded by 
 clarification of the samples by alum. 
 
 Temperature was recorded in degrees centigrade. 
 
 Dissolved Oxygen was determined by the use of the Hale and 
 Melia modification of the Winkler process. 
 
 Relative Stability was determined in bottles containing 250 c.c. 
 of the sample, when full, which were stoppered with corks, after 
 adding y 2 c.c. of a 0.05 per cent, aqueous solution of methylene 
 blue. The bottles were incubated at the laboratory room tem- 
 perature, and the time of decolorization noted. The relative 
 stability number, or value in terms of the per cent, stable, was 
 obtained from the table in Standard Methods, A. P. H. A. 
 
 Oxygen Demand (Biochemical), or more simply Demand, is 
 the amount of dissolved oxygen in parts per million parts of the 
 sewage, which sample of sewage, or treated effluent, will take up 
 from thoroly aerated clean water, when incubated in a glass- 
 stoppered bottle, at room or standard temperature for five days. 
 It is assumed that it measures the amount of dilution necessary 
 to prevent a nuisance when sewage is discharged into a body of 
 
12 Brooklyn, N. Y. 
 
 water, and also to approximate the condition of a stream, if 
 re-aeration did not take place. The determination was made ac- 
 cording to Standard Methods, A. P. H. A. 
 
 Local Conditions 
 
 The Twenty-sixth Ward of the boro of Brooklyn constitutes 
 the extreme easterly portion of the boro. The population (1918) 
 was estimated from police and school records to be 220,000. The 
 surface slopes from high lands along the northerly boundary, at 
 first with short but sharp grades, into a gently sloping plain, which 
 constitutes about 85 per cent, of the area. This has resulted in 
 the design of a sewerage system with very flat grades. The 
 drainage and sewage are discharged into the waters of Jamaica 
 Bay. 
 
 The hourly rate of flow in dry weather, based upon data de- 
 termined by weir measurements in the outfall, varies at present 
 (1918) from about 18,000,000 to 27,000,000 gallons per day. 
 The storm flow from the area at present provided with sewers 
 ranges up to 1,000 cubic feet per second, altho the ordinary 
 storm does not give more than 300 to 500 cubic feet per second. 
 The dry weather flow is mainly of a domestic character. The 
 suspended matters change widely in quantity at different seasons, 
 and at different hours of the day. The increase of population 
 occupying the drainage area is notable and of interest in show- 
 ing how rapidly New York suburban districts become urban, as 
 may be seen from Table I. 
 
 The sewerage system was mainly installed between the years 
 1890 and 1896 and includes a chemical-precipitation sewage-treat- 
 ment plant, designed in 1888 and intended to take care of a 
 maximum population of 35,000, which, at the time, was consid- 
 ered ample provision for the future. This plant was completed 
 in 1896, when the population had already reached about 60,000. 
 It, therefore, was inadequate from the beginning of its operation, 
 altho for several years it rendered fairly good service. 
 
 The sewage passing thru the plant flows thru fixed screens 
 and narrow settling tanks, with considerably velocity, to a cen- 
 tral well, from which it is pumped into the outfall sewer. The 
 pumps have a nominal capacity of about 20,000,000 gallons per 
 day, but on account of depreciation are incapable of pumping 
 more than half this quantity at present. 
 
 The twin outlet sewers, which are combined sewers with very 
 fiat grades, terminate at the location of the plant in a silt basin, 
 which, in dry weather, serves as a grit chamber for the sanitary 
 sewage. This basin, which is covered, is 80 ft. by 60 ft. in plan, 
 and is 9 ft. deep, acts as a trap to divert the sewage in dry 
 weather into the treatment plant. The basin is provided with a 
 hisfh level overflow for storm water, and the line of the sewer is 
 
Sewage Treatment Experiments 13 
 
 continued by means of an outfall flume, beginning at the high- 
 level overflow mentioned and extending from the basin to Jamaica 
 Bay. The outfall flume is a single channel section, 26 feet wide ; 
 its invert grade begins at the basin 3.40 feet above the invert 
 grades of the twin sewers and, as it is inadequate in storms of 
 much severity, it backs up the flow in the twin sewers and causes 
 flooding. In dry weather the sewage which cannot be taken care 
 of by the existing treatment plant overflows thru the storm outlet 
 and discharges into the bay without treatment. 
 
 A passageway.. 48. inches in diameter, is provided from the 
 grit chamber to the existing treatment plant, with the invert at 
 the floor elevation of the grit chamber, for carrying the dry- 
 weather flow into the plant, where, after receiving its charge of 
 lime, and having passed thru the screens and tanks, it flows into 
 the central pump well. From the well the treated sewage is 
 pumped against an average head of 20 feet into the outfall flume 
 at a point about 100 feet downstream from the silt basin. During 
 storms the by-pass valves are closed and the entire flow dis- 
 charges directly from the grit chamber thru the outfall flume 
 to the bay. 
 
 During a moderate rainfall the volume of flow carried off by 
 the sewers is about ten times the ordinary volume in dry weather. 
 To make matters worse, the existing sewers are not of modern 
 design and do not adapt themselves well to this double service and 
 are usually too small for present conditions of storm service. 
 They are, however, much too large for efficient operation in dry 
 weather, and this is especially the case in the outfall sewers, 
 which on account of low street grades and lack of sufficient 
 cover, are flat and broad, and blanketed by the tide at their outlet. 
 
 This condition results in making the sewers elongated settling 
 tanks as they approach the outlet of the system, and the dry 
 weather flow passes thru them with progressively falling velocity 
 from the lateral sewers to the outlet ; they are all, therefore, 
 sewers of deposit in dry weather and would cause great trouble 
 from smells as well as from clogging with banks of sedimented 
 solids, if it were not that the periodical storms flush them out, 
 carrying the matters that have collected in dry weather with the 
 rush of the flood wave into Jamaica Bay. The first part of the 
 storm flow is, therefore, exceedingly foul, much more so than the 
 ordinary dry-weather flow. 
 
 Measurement of Seivage. 
 
 The measurement of the total flow of sewage discharged by 
 the main outfall sewer in dry weather, and its hourly and daily 
 variations was obtained by means of a knife-edge weir installed 
 in the outfall flume. It was sharp crested with end contractions 
 suppressed. The crest length was 25.84 ft. and the height 2.17 
 
14 Brooklyn, N. Y. 
 
 ft. The horizontal knife-edge was installed by means of a wye 
 level ; the adjustments were made from slotted bolt holes, and 
 finally by filing off very small irregularities. 
 
 The device for recording the heads of flow on the weir con- 
 sisted of an automatic water stage register, with a specially de- 
 signed apparatus for magnifying the gage heights placed in a 
 Bazin pit installed just outside of the channel, and 16 feet up 
 stream from the crest of the weir. The pit was connected with 
 the flume by means of an iron pipe 3 feet long, laid on the floor, 
 and perpendicular to the side of the approach channel to the weir. 
 
 The zero of the register was determined by means of a hook 
 gage and wye level. The correction for slack motion of the 
 register was determined by hook gage to be ^ of 1 per cent. 
 
 A similar knife-edge weir, with end contractions suppressed, 
 with a crest 8 feet in horizontal length, was used to measure efflu- 
 ent from the Riensch-Wurl Screens. 
 
 This weir was placed in a specially prepared flume located 
 parallel to and alongside of the main outfall flume, the flow re- 
 entering the main flume several hundred feet below the experi- 
 ment station. 
 
 The details of this weir are similar to those of the larger weir. 
 
 In computing the discharge over these weirs the formulae of 
 Bazin, Fteley and Stearns, Francis, and Hamilton Smith, were 
 employed. The formula of Bazin was selected as probably the 
 most applicable to the kind of weir used under the observed con- 
 ditions. Taking Bazin at 100 per cent we have the following 
 comparative values : 
 
 Bazin .100 % 
 
 Fteley and Stearns.. 96.2% 
 
 Francis 95.6% 
 
 Hamilton Smith 95. % 
 
 Calibrated Orifices. — For the measurement of sewage delivered 
 to the various units of the experimental plant, calibrated orifices, 
 discharging under a constant head, were selected as the most sat- 
 isfactory, as errors would be less than would be the case with 
 weirs. In the use of such orifices, the calibration could be 
 checked experimentally with the apparatus in place by discharg- 
 ing into a measuring tank, or into the unit which the given orifice 
 was designed to supply with sewage. The latter was found to 
 be a satisfactory method. The calibration was performed against 
 the tank which was to be controlled by the orifice, and the head 
 was held by allowing a slight excess to waste over a weir above 
 the orifice. 
 
 The orifices originally designed were made of bronze ; they 
 were adjustable, and provided with scales for setting the size of 
 the opening for various rates of discharge. To measure small 
 
Sewage Treatment Experiments 
 
 15 
 
 rates of flow, orifices were made in copper plate in the machine 
 shop of the treatment plant, which gave entire satisfaction. 
 
 The accurate measurement of sewage is at best a difficult mat- 
 ter. The suspended material will affect almost any form of ap- 
 paratus used for the purpose ; orifices not excepted. However, 
 with regular and frequent cleaning, the types of orifices adopted 
 were satisfactory and gave results that were well within the ac- 
 curacy of the chemical tests and their interpretation. 
 
 Measurement of Compressed Air. — On account of the volume 
 of air to be metered, and the pressure, no form of available gas 
 meter was found fully to answer the requirements ; therefore, 
 several forms of Venturi meters were studied and a meter was 
 developed, which worked very well. The contract to make and 
 install these meters, and to test them in place, was awarded to 
 Messrs. Wallace & Tiernan, of New York. 
 
 The quantity of compressed air that would be required in 
 sewage aeration was unknown, at the beginning of the experi- 
 ments, and it was therefore necessary to provide for a wide range 
 in measuring capacity. This was accomplished by having several 
 interchangeable throats of different sizes for each Venturi meter, 
 and a specific manometer scale for each throat. 
 
 Three sets of meters were obtained, each of which was stand- 
 ardized as to measuring capacity, against an operating head of 
 40 lbs. air pressure, and calibration scales were worked out for 
 pressures from a minimum of 5 lb. to a maximum of 40 lb. per 
 
 AIR VENTURI METER 
 
 Fig. 4. 
 
16 
 
 Brooklyn, N. Y. 
 
 square inch, in order that by exchanging scales any pressure 
 within these limits might be used. The measuring capacity of 
 each set of meters under the 40 lb. pressure standard was as 
 follows : 
 
 Set No. 1 capacity 20 to 150 cu. ft. per minute. 
 
 Set No. 2 capacity 5 to 20 cu. ft. per minute. 
 
 Set No. 3 capacity 1 to 30 cu. ft. per minute. 
 
 The meters were each assembled in cylinders of the same size, 
 so that any one of them could be installed in the body of an or- 
 dinary 3-inch check valve, put in the air line. There were three 
 Venturi tubes complete, supplied for each meter ; two of these 
 were of small capacity, of which each had two interchangeable 
 throats, while the third was larger and had but one throat. 
 
 Referring to the drawing, Fig. 4, air passes along the line indi- 
 cated by the arrows, thru the throat and out thru the air line. 
 
 The manometer used was of special design. The cross section 
 of the oil reservoir was made many times greater than that of the 
 manometer tube, so that the zero point was practically unaffected 
 thru the range of the scale. The upper end of the manometer 
 tube was connected with the Venturi throat chamber, and had an 
 upper reservoir interposed, of slightly greater capacity than that 
 of the lower reservoir. This was provided to prevent the blowing 
 of any oil into the pipe line by sudden rushes of air. 
 
 The calibration of these meters was an interesting problem. 
 
 CALIBRATION BOX 
 
 s= 
 
 
 
 / 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 Fig. 5. 
 
 On account of the quantity of air it was impracticable to use a 
 gasometer. Therefore a calibrating box, consisting of a special 
 multiple-orifice box, was designed to receive and measure the air 
 
Sewage Treatment Experiments 
 
 17 
 
 passing thru the meter. This box is shown in Fig. 5. The top 
 of the box was a metal plate having in all 150 orifices, each ex- 
 actly equal in size and all made with the same die, and equally 
 spaced. A cover plate was provided, so designed that any num- 
 ber of orifices from 1 to 150 might be left open. The orifices 
 were each computed and made of proper size to discharge 1 cu. 
 ft. of air per minute under a head of 10 inches of water pressure. 
 A perforated metal screen was placed diagonally across the box 
 to prevent eddy currents. Preliminary readings were made for 
 rating this work, with a small gasometer, to determine the air 
 flows from different orifices and from different groups of orifices 
 to see if there was any variation because of location or number 
 of those discharging. The flows were found to agree within 1/5 
 of 1 per cent. 
 
 In calibrating a meter, the box was connected with the air 
 line thru the meter, and a certain number of orifices uncovered 
 on the top of the box, a stop-valve being opened until the desired 
 head was shown on the box manometer. 
 
 Knowing the discharge for this number of orifices, and the 
 corresponding pressure, it was a simple matter to mark on the 
 Venturi manometer scale the flow corresponding to the height of 
 the liquid. 
 
 After having determined several of these points, the scale 
 was graduated in cubic feet. 
 
 Distribution Control. — In order that the sewage coming from 
 the pumps might pass by gravity to every unit, and that the 
 impulse imparted by the pump might be removed, a quieting and 
 distributing tank was provided, thru which the entire plant could 
 be supplied except the R. W. fine screens. 
 
 PLAN 
 
 •*C>JUST*&U WFICt . 
 
 QUIETING TANK 
 SECTION A-A 
 
 SECTION B-B 
 
 CKSM una&t aimj I 
 
 
 Fig. 6. 
 
18 Brooklyn, N. Y. 
 
 Quieting Tank. — The quieting tank (Fig. 6) was designed to 
 maintain a constant head in the supply, at elevation 33.42 feet. 
 A platform around the tank was provided for convenient inspec- 
 tion and operation, and connected by means of a bridge with the 
 tops, or "decks", of the three Imhoff tanks. 
 
 Obtaining Seivage for the Experiments 
 
 When the work was commenced on the design of the experi- 
 mental plant in 1912, one of the first problems presented was 
 how to obtain sewage from the large sewers of fair average 
 strength and condition. 
 
 The following local conditions prevailed : The passages of 
 the large twin sewer, each 12 ft. 6 in. wide by 9 ft. high, entered 
 the grit chamber with their inverts at — ■ 2.40, datum being mean 
 high water. The surface of the flow of sewage in the sewers and 
 grit chamber at this point averaged about elevation -\- 2.37 for 
 dry-weather flow sewage, but was subject to changes of level 
 during the day between elevation -\- 1.75 and -|- 3.00. 
 
 The surface velocity of the sewage as it entered the grit 
 chamber ranged from about half a foot to more than a foot per 
 second, under different conditions of flow. 
 
 The sewers and grit chamber acted to some extent as a sedi- 
 mentation tank, sewage solids, as well as grit, being deposited 
 a^ all times during dry weather in greater or less quantities. The 
 inverts of the larger sewers of the Twenty-sixth Ward are shal- 
 low and wide, and the grades too flat, as a rule, to afford suffi- 
 cient velocity of flow to prevent considerable of the suspended 
 solids in the sewage from settling and forming deposits, which 
 during storms are picked up by the high velocity of flow and 
 flushed out into Jamaica Bay. During the first rush of the storm 
 the discharge is even more foul than the ordinary discharge of 
 sewage in dry weather. Following this, however, after the flush- 
 ing out of these deposits has been completed, the storm flow 
 rapidly improves in quality and soon becomes mere dirty water, 
 which is incapable of causing a nuisance, and does not require 
 treatment. 
 
 The by-pass, provided to carry the dry- weather flow from the 
 grit chamber of the treatment plant, is given off from the lower 
 portion of the chamber, the bottom of which agrees with the 
 invert of the by-pass, and is elevation — 3.00. Since the by-pass 
 is 48 in. in diameter, the crown of its arch is elevation -f- 1.00. It 
 enters the treatment plant with its gates at invert elevation — 3.50, 
 which is above the water line in the receiving tanks of the plant, 
 so that sewage enters with a free flow. 
 
 The name "grit or silt chamber" is really a misnomer, for 
 most of the grit and silt settles from the sewage in the large 
 sewers before the chamber is reached. 
 
Sewage Treatment Experiments 19 
 
 All of the storm flow and about two-thirds of the dry weather 
 flow, which cannot be admitted to the treatment plant, discharges 
 from the grit chamber at elevation -[-1.00 into the outfall flume 
 which runs 3,230 feet to the bay. 
 
 It was observed that the sewage passing into the treatment 
 plant, thru the by-pass, gave a more consistent average proportion 
 of the total and suspended solids, than that which passed into the 
 bay thru the outfall sewer, which, on the average, was a weaker 
 sewage, altho at times it was much stronger, and carried more 
 floating matter, as well as settleable solids, but was subject to 
 greater variation in suspensa, as well as volume of flow. These 
 conditions indicated that the mo-st appropriate sewage for study 
 in the experimental work was that entering the treatment plant, 
 as near the outlet end of the by-pass as practicable. For, altho 
 this sewage was slightly stronger than the average of the entire 
 flow, this would give results on the safe side ; while the other flow 
 was weaker than the average, and would, therefore, give more 
 fallacious results. 
 
 TABLE I 
 
 Population of the 26th Ward of Brooklyn, N. Y. 
 
 Increase of Population, 
 26th Ward, Brooklyn. 
 Year Population to nearest 1,000 
 
 1880 13,000 
 
 1890 30,000 
 
 1900 66,000 
 
 1905 94,000 
 
 1910 178,000 
 
 1914 200,000 
 
 1918 : 220,000 
 
 TABLE II 
 
 Daily and Hourly Flow of Sewage 
 
 Dry-weather flow of sewage contributed by 220,000 peo- 
 ple REACHING THE OUTLET IN HENDRIX STREET, BASED UPON 
 DATA OBTAINED BY WEIR MEASUREMENTS, REDUCED TO HOURLY PER- 
 CENTAGE OF FLOW PER CAPITA : MAXIMUM DAILY RATE PER CAP- 
 ITA 121 gallons. Minimum rate 81 gallons. Prepared for 
 January 1, 1918. 
 
 Hour Pet. of max. flow Cu. ft. per sec. Rate in M.g.d. 
 
 12 a. m 78.5 32.49 21 
 
 1 a. m 74.4 30.92 20 
 
 2 a. m 71. 29.40 19 
 
 3 a. m 67.8 28.62 18.5 
 
 4 a. m 67. 27.85 18 
 
 5 a. m. min 66.1 27.07 17.6 
 
 6 a. m 67.8 27.85 18 
 
 7 a. m 72.8 30.17 19.5 
 
 8 a. m 81. 33.65 21.75 
 
20 Brooklyn, N. Y. 
 
 Hour Pet. of max. flow Cu. ft. per sec. Rate in M. g. d. 
 
 9 a. m 87.6 36.36 23.5 
 
 10 a. m 91.7 37.90 24.5 
 
 11 a. m 93.4 38.68 25 
 
 12 p. m 96.7 40.23 . 26 
 
 1 p. m 99.2 41.00 26.5 
 
 2 p. m 99.2 41.00 26.5 
 
 3 p. m. max 100. 41.29 26.62 
 
 4 p. m 97.5 40.61 26.25 
 
 5 p. m 96.9 39.84 25.75 
 
 6 p. m 94.2 39.06 25.25 
 
 7 p. m 89.2 37.13 24 
 
 8 p. m 85.9 35.59 23 
 
 9 p. m 84.3 34.81 22.5 
 
 10 p. m 84.3 34.81 22.5 
 
 11 p. m 82.6 34.42 22.25 
 
 12 p. m 78.5 32.49 21 
 
 The weekly cycle of daily and hourly per capita flow of sewage 
 is shown by Table III, as well as the quantity of flow for each 
 day of the week. The figures are averages. 
 
 TABLE III 
 
 Per Capita Flozv of Sewage in Gallons for Every Hour of the 
 
 Day and Week 
 
 1914-1915 
 
 Sun. Mon. Tues. Wed. Thurs. Fri. Sat. 
 
 Hour gal. gal. gal. gal. gal. gal. gal. 
 
 5 a .m 3.3 3.4 3.6 3.2 3.4 3.6 3.0 
 
 6 a. m 3.3 3.6 3.8 3.3 3.5 3.6 3.2 
 
 7 a. m 3.4 3.8 4.1 3.6 3.6 3.6 3.5 
 
 8 a. m 3.9 4.2 4.4 4.3 4.2 4.1 3.9 
 
 9 a. m 4.5 4.8 4.8 4.5 4.8 4.5 4.3 
 
 10 a. m 4.8 5.8 5.0 4.6 5.0 4.7 4.4 
 
 11 a. m 5.0 5.1 4.8 4.7 5.0 5.0 4.4 
 
 12 a. m 5.1 5.4 5.0 4.7 5.0 5.0 4.7 
 
 1 p. m 5.0 5.5 5.0 4.7 5.1 5.0 4.8 
 
 2 p. m 4.8 5.2 4.8 4.8 5.3 5.1 5.0 
 
 3 p. m 4.8 5.1 5.1 4.8 5.3 5.2 4.0 
 
 4 p. m 4.7 4.8 5.1 4.8 5.0 4.4 4.8 
 
 5 p. m 4.5 4.7 5.0 4.7 5.0 4.3 4.7 
 
 6 p. m 4.4 4.6 4.7 4.6 4.8 4.3 4.5 
 
 7 p. m 4.2 4.5 4.5 4.6 4.7 4.2 4.4 
 
 8 p. m 4.2 4.5 4.5 4.4 4.5 4.2 4.4 
 
 9 p. m 4.1 4.5 4.4 4.4 4.1 4.1 4.2 
 
 10 p. m 4.1 4.4 4.3 4.4 4.4 4.1 4.1 
 
 11 p. m 4.1 4.3 4.2 4.3 4.3 4.1 3.8 
 
 12 p. m 3.9 4.2 3.9 3.9 4.2 3.9 3.7 
 
 1 a. m 3.7 3.9 3.9 3.7 3.7 3.9 3.5 
 
 2 a. m 3.5 3.6 3.8 3.5 3.5 3.8 3 3 
 
 3 a. m 3.4 3.5 3.7 3.3 3.3 3.6 3.3 
 
 4 a. m 3.3 3.4 3.6 3.2 3.3 3.6 3.1 
 
 Totals 100.0 106.0 106.0 101.0 105.0 101.9 97.8 
 
 Mean flow for week, 102. S gallons per capita. 
 
Sewage Treatment Experiments 21 
 
 The Storm Water Sewage 
 
 The effect of storms on the quality of the sewage was always 
 greatly to increase the matters in suspension during the first 
 period of the storm. Where the storm was of short duration, 
 this increase continued thruout the storm, but if the duration 
 was continued over several hours, a marked improvement took 
 place in the flow. 
 
 This condition, as already mentioned, was probably due to the 
 large size of the main sewers, which are on the combined plan, 
 and have very flat grades thru much of their extent, and shallow 
 inverts, into which considerable settling matter falls in dry 
 weather, being flushed out by the flood wave of the storm. 
 
 Two storms, of ordinary severity such as frequently occur 
 at this place, and a rather high rate of precipitation, may be given 
 here in illustration of the phenomena attending upon the flow of 
 storm sewage, during the usual summer shower. 
 
 The first storm referred to ("A" in Table IV) took place on 
 May 27, 1914. The total rainfall recorded was .15 inches, of 
 which .10 fell in the first half hour. The second, ("B" in the 
 table) occurred on August 21 of the same year. The rainfall, 
 according to the gage, was 1.0 inch in all, 0.9 inch falling in 
 the first half-hour. 
 
 The following phenomena were common to both storms. After 
 a lapse of from twenty-five minutes to an hour from the begin- 
 ning of rainfall, the sewage became very foul, as shown by the 
 remarkable leap in suspended solids. The persistence of this 
 abnormal content was only about an hour in the heavier storm, 
 and about two hours in the lighter one. The presence of consid- 
 erable quantity of gritty street washings is indicated by a marked 
 drop in the percentage of volatile matter in the suspensa. In 
 the storm of May 27, this drop took place some time later than 
 the time of maximum suspended matter, but in the storm of 
 August 21, the street wash appeared to come coincidently with 
 the outflush of sludge from the inverts of the trunk sewers. The 
 accompanying Table IV gives the figures : 
 
 TABLE IV 
 
 Hrs. after Settling 
 
 beginning Matter Suspended Solids Dissolved 
 
 of rain c.c.l. Total p. p.m. Volatile % Oxygen p.p.m. 
 
 A B A B A B A B 
 
 Date 5-27 8-21 5-27 8-21 5-27 8-21 5-27 8-21 
 
 2.7 3.2 226 252 74 76 .6 .7 
 
 Va 1.5 3.0 208 254 78 75 .5 .6 
 
 Vi 2.1 6.2 248 412 73 77 .4 
 
 54 2.5 . 16.5 250 2250 72 40 .4 
 
 1 29.0 12.8 1880 1976 72 41 
 
 W2 2.7 4.0 588 820 35 27 
 
 2 5.4 2.5 1126 480 28 30 .1 1.9 
 
 2J4 5.5 2.6 1096 480 29 38 .6 1.1 
 
 3 4.2 494 39 .... .7 
 
 4 2.2 204 47 .... 2.5 
 
 5 9 118 63 .... 1.9 
 
 6 5 84 52 .... 2.4 
 
22 
 
 Brooklyn, N. Y. 
 
 The most important characteristics of the dry-weather flow 
 sewage are exhibited by Table V, which for the data given covers 
 the period of the experimental work of the station, upon which 
 the report is based. It should be noted that the figures given 
 are averages, and as such do not show either extreme of the con- 
 ditions. 
 
 TABLE V 
 General Characteristics of 26th Ward sewage 
 
 From Monthly Averages for the Period of Experiments 
 
 Oxygen 
 
 Suspended Solids — Demand 
 
 Month Temp. Diss. Oxygen Total Volatile Non-Vol. Cone Biochem 
 
 °C p.p.m. % sat. p. p.m. p. p.m. p. p.m. c.c.l. p.p.m. 
 
 Jan 11.0 4.5 41 175 140 35 2.1 209 
 
 Feb 10.0 4.5 40 172 134 38 2.0 262 
 
 March 11.6 3.6 33 192 142 50 1.8 220 
 
 April 14.7 2.8 27 162 124 38 1.7 195 
 
 May 18.2 1.9 20 178 136 42 1.9 213 
 
 June 20.9 1.2 13 153 118 35 1.6 250 
 
 July 26.0 0.8 10 163 118 45 1.7 202 
 
 Aug 23.8 0.6 7 146 112 34 1.7 203 
 
 Sept 22.1 0.5 6 168 136 32 2.1 237 
 
 Oct 17.5 1.9 20 146 108 38 1.8 205 
 
 Nov 13.9 3.8 37 160 132 28 1.7 254 
 
 Dec 10.8 3.8 34 203 155 48 2.6 223 
 
 Average 16.7 168 129 39 1.9 223 
 
 Dec-Mar 10.8 4.1 37 186 143 43 2.1 228 
 
 Apr.-June 18.0 2.0 21 164 126 38 1.7 219 
 
 July-Sept 24.0 0.6 7 159 122 37 1.8 214 
 
 Oct. -Nov 16.C 2.9 29 153 120 33 1.8 229 
 
 Seasonal variations by averages 
 
 Month 
 & Year 
 
 TABLE VI 
 Showing Oxygen Relations of Sewage for One Year 
 
 Oxygen Consumed in 30 Minutes Digestion 
 
 Susp'd Solids Nitrogen as 
 
 Oxygen 
 Consumed 
 
 Diss. 
 Oxygen" 
 
 Temp. C 
 
 1914 
 
 Is 
 
 H & 
 
 ■a s 
 
 .ti d 
 
 Zd 
 
 .t; d 
 Zd 
 
 3S 
 
 C D. 
 
 id 
 
 iS d 
 
 April 
 
 May 
 
 June 
 
 July 
 
 Aug. 
 
 Sept. 
 
 Oct. 
 
 Nov. 
 
 Dec. 
 
 1915 
 
 Jan. 
 
 Feb. 
 
 126 
 164 
 125 
 152 
 128 
 164 
 127 
 158 
 201 
 
 175 
 219 
 
 March 203 
 
 129 
 96 
 97 
 101 
 130 
 101 
 126 
 155 
 
 142 
 168 
 155 
 
 .09 
 .37 
 .21 
 .07 
 .13 
 .28 
 .37 
 
 .33 
 .13 
 .31 
 
 .31 
 .09 
 .02 
 .01 
 .05 
 .05 
 .07 
 .14 
 .13 
 
 .13 
 .13 
 .16 
 
 53 
 60 
 48 
 56 
 53 
 58 
 51 
 51 
 71 
 
 74 
 82 
 
 75 
 
 35 
 44 
 38 
 37 
 35 
 39 
 44 
 
 2.2 
 1.5 
 1.8 
 1.2 
 
 1.2 
 2.2 
 1.9 
 
 2.0 
 
 2.6 
 2.5 
 1.4 
 
 22 
 17 
 
 21 
 14 
 14 
 24 
 19 
 18 
 
 13.7 
 18.3 
 21.1 
 22.3 
 23.9 
 21.4 
 19.4 
 15.9 
 12.0 
 
 11.6 
 11.0 
 13.3 
 
 * Comparative dissolved oxygen values during an average day, by two-hour periods. 
 Lowest hourly value assumed to equal 1,000. Twenty-sixth Ward Sewage: 
 
 Midnight 1.100 
 
 2 a. m 1.100 
 
 4 a. m 1.205 
 
 6 a. m 1.300 
 
 8 a. m 1.545 max. 
 
 10 a. m 1.478 
 
 Noon 1.363 
 
 2 p. m 1.182 
 
 4 p. m 1.341 
 
 6 p. m 1.000 min. 
 
 8 p. in 1.100 
 
 10 p. m 1.114 
 
 Midnight 1.100 
 
Sewage Treatment Experiments 23 
 
 TABLE VII 
 Average Nitrogen Contents of Sewage 
 
 Free ammonia 32.00 p.p.m. 
 
 Organic nitrogen: 
 
 Total 27.00 p.p.m. 
 
 Dissolved 18.00 p.p.m. 
 
 Nitrites 0.26 p.p.m. 
 
 Nitrates 0.00 p.p.m. 
 
 Nitrites and nitrates are frequently absent, and the sewage was 
 frequently septic during the summer months, but very seldom 
 during the other seasons. 
 
 TABLE VIII 
 Daily cycle of sewage, showing hourly changes in strength 
 
 Suspended Solids Oxygen Consumed* 
 
 Settling Total Volatile Total Diss. 
 
 **c.c. perl, p.p.m. p.p.m. p.p.m. p.p.m. 
 
 Midnight 1.6 135 111 58 42 
 
 2 a. m 1.4 168 137 58 39 
 
 4 a. m 1.8 128 107 51 36 
 
 6 a. m 1.0 93 74 39 27 
 
 8 a. m 1.4 88 64 36 25 
 
 10 a. m 2.0 146 110 67 46 
 
 Noon 2.0 196 159 81 48 
 
 2 p. m 1.7 177 139 72 47 
 
 4 p. m 2.3 190 147 84 51 
 
 6 p. m 2.1 187 147 96 59 
 
 8 p. m 2.2 183 140 88 56 
 
 10 p. m 1.5 149 115 70 48 
 
 Midnight 1.6 135 111 58 42 
 
 * Oxygen consumed in thirty minutes at 100°C. Diss, is from filtered sample. 
 ** The settling suspended solids in the first column are obtained by means of the 
 Imhoff cone. 
 
 One of the notable features of the local sewage is the large 
 proportion of colloidal suspensa in very finely divided condition, 
 which settles very slowly if at all. The different tanks removed a 
 satisfactory percentage of settleable solids, but to a large extent 
 this non-settleable material remained in the sewage. 
 
 That portion of this fine material which disappeared in passage 
 thru the tanks probably did not settle but was dissolved. 
 
 Experiments in Imhoff settling cones were made to determine 
 the average rate of sedimentation. The volume of material in 
 the apex of the cone was read after the expiration of varying 
 intervals. The following results, obtained from an extensive 
 series of observations, are averages of the quantity of settling 
 matter in cubic centimeters per liter, and, assuming the average 
 amount settled in two hours equal to 100 per cent., the percent- 
 ages, for shorter periods are given. 
 
 Time Settlings Per Cent. 
 
 Minutes c.c. per liter Settled 
 
 5 0.629 33.1 
 
 10 1.109 58.3 
 
 15 1.188 62.5 
 
 30 1.359 71.5 
 
 45 1.480 78.0 
 
 60 1.490 78.4 
 
 120 1.900 100. 
 
24 Brooklyn, N. Y. 
 
 The above matters, settling per liter, were equivalent only to 
 about 69 p. p. m. of the sewage, which contained 162 p. p. m. 
 of suspended solids. In other words, 1.90 c.c. per liter represents 
 only about 43 per cent, of the suspended solids. The remainder 
 would not settle in a two-hour period, and even in six hours but 
 a small portion of it would settle. It is obvious, therefore, that 
 the tanks would not remove in two hours a greater proportion 
 than could be removed in the cones. In the results of removal 
 effected by the tanks given in the tables which follow, this should 
 be understood. The tabulated figures refer to the constituents 
 observed in the sewage and in the effluents'. In Table XIV the 
 percent of removal is also shown. 
 
 The removal of 43 per cent, of total suspended solids might be 
 stated as far as the tank is concerned, as equal to a removal 
 of 100 per cent., because it is quite clear that the tanks cannot 
 remove in two hours any more solids by sedimentation than a 
 cone can in the same interval of time. 
 
 That the Imhoff cone should be a more efficient remover of 
 suspended solids than the tanks is not in the least surprising, 
 when it is recalled that in the cone, sedimentation is quiescent 
 and theoretic retention is 100 per cent, efficient — -while in the 
 tank there is always some movement in the settling sewage and 
 the depth is several times greater than in the cone, and it follows 
 that particles reach the bottom of the cone considerably quicker 
 than they do the bottom of the tank. 
 
 In the studies made on the Brooklyn sewage, experiments were 
 undertaken by Mr. W. T. Carpenter to ascertain some of the 
 phenomena attending the sedimentation of sewage solids. These 
 studies were too elaborate to be given space in this paper, but 
 as one set of experiments illustrates the effect of depth on sedi- 
 mentation, and fineness, or colloidal condition of suspensa, suffi- 
 cient of it will be given to bring out this point. 
 
 The apparatus used consisted of 2-inch diameter galvanized 
 iron pipe terminating in a coupling into which was cemented the 
 lower part of an Imhoff cone, thus giving a chamber in which 
 settling matter was visible and could be measured volumetrically. 
 The apparatus afforded a sedimentation depth of 74 inches. The 
 capacity was that of a measured sample bottle, which was filled 
 with the sewage or tank effluent to be tested. The contents of 
 the bottle were discharged into the apparatus as quickly as pos- 
 sible. The volume of sediment forming at the apex was read at 
 intervals of 15 minutes, 30 minutes, and at 1, 2, 3, 4, 5, and 6 
 hours. 
 
 A large number of such tests were made on crude sewage, and 
 also on Imhoff tank effluent, using that from tank 3. 
 
 The figures given are averages of all the tests. For compari- 
 son, figures showing rate of deposit in the ordinary Imhoff cone 
 on the same sewage and effluents are also given. 
 
Sewage Treatment Experiments 
 
 25 
 
 TABLE IX 
 
 Volume of Deposit and Time of Settling 
 
 Observed in Im/iofi and Special Cones 
 Quantities in c.c. per liter 
 
 Time 
 
 
 Imhoff Cone 
 
 Special 74- 
 
 in. Cone 
 
 
 
 Sewage 
 
 Effluent 
 
 Sewage 
 
 Effluent 
 
 Minutes 
 
 
 
 
 
 
 S 
 
 
 0.63 
 
 0.0 
 
 
 
 10 
 
 
 1.11 
 
 .12 
 
 
 
 IS 
 
 
 1.19 
 
 .20 
 
 0.52 
 
 0.07 
 
 30 
 
 
 1.36 
 
 .29 
 
 .90 
 
 .10 
 
 45 
 
 
 1.48 
 
 .36 
 
 
 .13 
 
 Hours 
 
 
 
 
 
 
 1 
 
 
 1.49 
 
 .40 
 
 1.33 
 
 .15 
 
 2 
 
 
 1.90 
 
 .47 
 
 1.69 
 
 .20 
 
 3 
 
 
 2.30 
 
 .50 
 
 1.89 
 
 .25 
 
 4 
 
 
 
 
 1.97 
 
 .28 
 
 5 
 
 
 
 
 2.02 
 
 .32 
 
 6 
 
 
 
 
 2.03 
 
 .a 
 
 
 
 Sewage and effluent tested* 
 
 
 
 Contents 
 
 of 
 
 
 Sewage 
 
 Effluent 
 
 % Removed 
 
 Total solids 
 
 
 162 p. p.m. 
 
 97 p. p.m. 
 
 40 
 
 Settling 
 
 solids 
 
 
 92 p. p.m. 
 
 23 p. p.m. 
 
 66 
 
 Colloidal 
 
 solids 
 
 
 93 p. p.m. 
 
 74 p. p.m. 
 
 20 
 
 * Effluent is from Imhoff Tank 3. 
 
 % of removal given was effected in the Imhoff tank. 
 
 Experimental Data From the Imhoff Tanks and Settling Tanks 
 (Dortmund Type) With Short Retention Periods. 
 
 The station was provided with three Imhoff tanks, so designed 
 that they differed in depth and cubic capacity only. Each tank 
 was a wooden cylinder, 15 ft. in diameter. They were placed 
 along the northerly side of the plant, extending east from the 
 quieting tank, from which each received sewage thru an inde- 
 pendent flume, the sewage entering the flume thru an adjustable 
 measuring orifice. The plant was also provided with four plain 
 settling tanks (Dortmund type), which received sewage from 
 the quieting tank in the same manner. 
 
 The principal dimensions of these tanks are shown by means 
 of Table X, and the accompanying sketches. 
 
 For the purposes of this paper the three Imhoff tanks and 
 settling tank 3, will be considered in one group, as, during the 
 period of the tests described on the Imhoff tanks, settling tank 3 
 was operated on the same sewage in connection with a separate 
 digestion tank, which daily received the solids settled out. The 
 two tanks acted the part of an Imhoff tank with the two stories 
 separated. 
 
 The direction of flow thru the Imhoff tanks was from north to 
 south. The tanks were so connected with the other units of the 
 experimental plant, that Imhoff effluent could be obtained by grav- 
 ity flow for all. The three tanks were similar in design. Inlet 
 and outlet weirs were full width of the flowing-thru chamber, 
 and were of the same design in each case, as were also the hopper 
 
26 
 
 Brooklyn, N. Y. 
 
 Fig. 7. View of Experimental Tanks from the North. 
 
 1. Quieting Tank. 
 
 2. Imhoff Tank No. 1. 
 
 3. Imhoff Tank No. 2. 
 
 4. Imhoff Tank No. 3. 
 
 5. Sprinkling Filter Beds. 
 
 Capacity of Plant — 1,200,000 gallons per day of sewage was used in the 
 various experimental processes. 
 
 Man is seen taking samples on top deck of Imhoff Tank No. 3, just 
 ui\der Fig. 4. 
 
 The Imhoff Tanks are 15 feet in diameter inside. 
 
 bottoms, the 8-inch-diameter sludge outlet, the settling-chamber 
 floors, slopes and slots. The tanks differed only in the matter of 
 depth. The water line in each was at elevation 31.17. 
 
 The slots for the passage of settled matters from the slopes 
 into the digesting chamber were so • designed that the plane of 
 each slope was carried without obstruction directly thru into 
 the digesting chamber, on each side. The slopes did not pass the 
 one under the other, as is often observed in American practice. 
 The latter design was avoided, as with it the settlings on one 
 slope must turn over upon the other slope in order to pass thru 
 into the digesting chamber, thus making obstruction probable and 
 calling for frequent cleaning of slopes and slots. See Fig. 9, at 
 1 and 3. The design adopted to avoid this (Fig. 9, at 2) was 
 to guard the slots, or openings from below, from rising gases by 
 means of an A-shaped shield or baffle which afforded a slot on 
 each side. The upper surfaces of the A-shaped shield were ex- 
 actly in the same plane as the slopes of the flowing-thru chamber, 
 so that when sediment started to slide, it found an opening di- 
 rectly in its path thru which it could pass without stopping or 
 turning in its course. This is believed to be an important prac- 
 
Sewage Treatment Experiments 
 
 27 
 
 tical point in the design of these tanks. It was so successful that 
 slopes, inclined only about 42 degrees from level, did not at any 
 time become clogged in five years' service, and never needed to 
 be squeegeed down. The width of each slot on the plane of the 
 slope was 6 inches. The inclination of slope above mentioned 
 was put in as the least slope that might be expected to work with 
 success. Arrangements had been made to increase it if neces- 
 sary, but no such change was made. 
 
 SECTIONS OF IMHOFF TANKS 
 IMHOFF TANK N? I IMHOFF TANK N? 2 
 
 fwvf 
 
 SLUW.E PSCH&AbE 
 
 IMHOFF TANK N? 3 
 
 ■=» 
 
 o a ti 
 
 "3 H 5 
 
 Fie. 8. 
 
 The bottom of each digesting chamber was formed inside of 
 the cylindrical tank in the shape of an inverted truncated hexa- 
 gonal pyramid, made in two sections, the upper overlapping the 
 lower. A perforated lead pipe 1% inches in diameter was placed 
 entirely around the overhanging edge of the upper section of 
 pyramid, and connected with the city water supply, for use in 
 starting sludge, etc. The water was controlled by a gate valve. 
 The use of this pipe never was called for in the operation of the 
 tanks. 
 
 Scum boards 18^ inches deep were placed at inlet and outlet 
 weirs, one foot from each weir, and were used thruout the ex- 
 periments. 
 
 Baffles were not provided at first, but the necessity for them 
 was demonstrated as the experiments progressed, and they were 
 then put in, after considerable study as to their position, shape 
 and depth. 
 
Fig. 9. 
 
 1. Type of Slot Frequently Used in American Practice. 
 
 2. Type of Slot Used on all the Brooklyn Imhoff Tanks. 
 
 3. Type Sometimes Used. 
 
Sewage Treatment Experiments 
 
 29 
 
 PLAN OF IMHOFF TANKS 
 IMHOFF TANK N9 I IMHOFF TANK N<?£ IMHOFF TANK N? 3 
 
 FLUMES fTOM QUIETtS6 TftHK 
 
 , EFFLUENT FROM IMHOFF TANKS TO i &al TRICKLW6 
 
 yCX __^_ __ {J = 
 
 nus&um Wfftct tat ■ 
 
 01 e 3 4 s ;. 
 
 FEET 
 
 Co? 
 
 Fig. 10. 
 
 HasnmMRcei 
 
 Fig. 11. 
 
 The gas vents, being segments of the 15 ft. diameter circle, 
 cut off by the walls of the flowing-thru chambers, were 2 ft. wide 
 at their greatest width. They were at all times ample for the 
 purpose. 
 
 The tanks were completed and put in service October 1, 1913, 
 and a few days later, at the suggestion of Dr. Rudolph Hering, 
 ripe sludge was obtained from the tanks of the Pennypack Creek 
 
30 Brooklyn, N. Y. 
 
 sewage disposal plant near Philadelphia, and with this each tank 
 was seeded. The ripening period extended thru the severe win- 
 ter of 1913-1914 and was not completed in either of the tanks 
 until May 21, 1914, when No. 2 and No. 3 gave ripe sludge, but 
 No. 1 did not ripen until June 11. The tanks, however, all con- 
 tinued to improve in operation during the year 1914. Early in 
 the spring there was considerable smell from all of them, but 
 this disappeared during the summer, and did not return until the 
 following spring. Improvement continued thruout the entire 
 period of operation, and the tanks were doing their best when 
 finally shut down in 1918. At no time during the experiments 
 was there any evidence of sludge accumulating on the slopes of 
 the settling chambers. As stated elsewhere, the slopes inclined 
 only 41 degrees 43 minutes from level. It had been intended to 
 try a number of different angles of slope, beginning with the least 
 which it was convenient to place, with the expectation that clean- 
 ing would be necessary with slopes inclined from level less than 
 50 degrees, but there seemed no reason to increase the inclination 
 after installing the slopes. The surfaces of the slopes were of 
 planed boards, with the grain running with the slope. It is quite 
 possible that this surface is less subject to affording points of ad- 
 hesion for settling matters than concrete would be, but as it was 
 considered perfectly feasible to use an inclined floor of planed 
 boards for the slopes in a full size plant, it was decided not to go 
 further in this study, but to adopt an angle of inclination in future 
 tanks as much greater than 42 degrees as the other conditions 
 would permit. The double slot placed at the lowest point, i. e. 
 the intersection of the planes of each side, without permitting the 
 crest of the baffle to emerge into the settling tank and form an 
 obstruction to sliding matters, — was of much importance in keep- 
 ing the slopes clean. The settlings do not appear to go thru the 
 slot at the foot of the slope down which they slide, unless they 
 are heavy. The light-weight materials and much of the heavy, 
 appear to jump over this space and go thru the other slot, which 
 does not require a change in their direction of movement (see 
 Fig. 9). It appears from the work of the Brooklyn tanks that 
 the design of slots is far more important in its relation to tank 
 operation than has been generally recognized in practice. 
 
 The phenomena which accompany foaming and frothing in 
 the Imhoff tanks are so various that it must be admitted that 
 many explanations are possible. Tanks will foam when no 
 sludge is present, and also when filled with sludge — and when 
 much or little is present. The suspended matters are so light in 
 specific gravity that any cause which gives rise to a rapid gas 
 formation may result in foaming, even changes of barometric 
 pressure in the atmosphere may cause it. The presence of grease 
 in the tank and an acid reaction, seems to be one of the most 
 persistent conditions noted when foaming takes places. But 
 
Sewage Treatment Experiments 31 
 
 foaming takes place when the condition is not acid and there is 
 but little grease present, and when explanation seems impossible. 
 At times during the experiments, and without any apparent rea- 
 son, the entire body of sludge in the digestion chambers of the 
 different tanks would rise to the top, usually without foaming or 
 frothing, and remain for a day or two, then slowly settle. On 
 such occasions, if foaming did not occur gas discharge from the 
 gas vents was at least more than usually active and large eructa- 
 tions of foul gas were noted. At such times there was' considera- 
 ble odor present, and when foaming occurred there was a strong 
 and very persistent odor. Churning the scum with a hoe in the 
 gas vents, or breaking it up with a paddle, usually sufficed to 
 release the entrained gas and permit it to settle. A jet from a 
 hose is probably the most effective method of doing this. At no 
 time was the foaming so serious that it could not be controlled, 
 but while it lasted the tank gave forth bad smells, and its effi- 
 ciency as a remover of suspended matter decreased more than 
 half. The settling chambers were at no time involved in foaming 
 and very seldom showed any scum. 
 
 The regrettable thing about this bad habit of the Imhoff tank 
 is that even if it can be controlled it causes bad odors, which 
 makes it a very uncertain risk, and a bad neighbor, and suggests 
 that such tanks should not be placed near enough to habitations 
 to give rise to damage claims or cause complaint. If it is neces- 
 sary to put such a tank near inhabited structures, where com- 
 plaint is at all probable, it should be covered with a building that 
 would keep in the odors, when these occur. It is a simple matter 
 to provide a building which will insure a neighborhood against 
 nuisance from odors and at the same time permit the use of this 
 tank, which is otherwise such an excellent apparatus for sewage 
 treatment. 
 
 The settling tanks (Dortmund type) never showed the least 
 inclination to foam, and never gave forth odors that could be 
 recognized as coming from sewage a few feet away. This may 
 have been because they were not intended for, or used as digest- 
 ing or septic tanks, the settling matters and sludge being removed 
 daily for other treatment. 
 
 It may be added that the Imhoff tanks very seldom caused 
 odors, and at times when they did, if it was not due to foaming, 
 it was probably due to a septic condition of the sewage entering 
 them. At one time during the experiments all of the tanks caused 
 odors, and it was discovered that the main sewer had become 
 very foul from deposits. After the sewer was cleaned out these 
 odors ceased, and did not occur until the sewer again required 
 cleaning. When this was done, odors again ceased, at least 
 suggesting the connection between bad smells from a disposal 
 plant, and neglect in keeping sewers clean. 
 
 The idea that sewers and disposal plants will run without 
 
32 Brooklyn, N. Y. 
 
 causing trouble if left to take care of themselves, is erroneous 
 as we all know, but it is not sufficiently understood that troubles 
 with the disposal plant are frequently caused by neglect in clean- 
 ing out the sewers. 
 
 For some reason not understood by the writer, the Imhoff 
 tank, while being installed more and more extensively by engi- 
 neers, has fallen into disrepute in public opinion. In rather ex- 
 tensive travels around this country during the last few years, 
 the writer has visited many such tanks, and plans of new installa- 
 tions are constantly appearing in the technical papers', yet, on 
 visiting various plants in operation, many complaints were heard, 
 and the general consensus of opinion was decidedly unfavorable. 
 What is the reason for this attitude on the part of those who are 
 maintaining and operating these tanks ? There is nothing worse 
 than to give a dog a bad name, and many a good dog has been 
 condemned for this reason. 
 
 The Imhoff tank has been closely studied by the writer, and 
 while his judgment may be prejudiced in its favor, and he may be 
 told that his explanations of its failures are nothing more than 
 apologies, yet he feels that they are entitled to a hearing at least. 
 His first studies of the tank were received from Dr. Imhoff, 
 himself, nearly ten years ago, and he had the opportunity of 
 visiting with Dr. Imhoff all of the larger and some of the smaller 
 installations in the Emscher district in 1912, and hearing at first 
 hand the principles upon which the designs were based, and of the 
 various investigations which had been made on various points 
 concerned both in the design and in the operation of these tanks. 
 In 1913 Dr. Imhoff spent some of his valuable time in the writer's 
 office in Brooklyn, discussing the design of the three tanks de- 
 scribed in this paper, he having permitted their use without 
 charge for experimental purposes. Many of the features of the 
 tanks were made to conform especially with his views and in- 
 structions, so that the design in reality was his and not the 
 writer's — and to him alone is due the credit for their success. 
 The writer has, since 1912, visited many American installations 
 of the Imhoff tank, and while he must admit some troubles in 
 operation, and smelling some odors not very bad to a sewage 
 expert, but killing to neighboring residents who had the possi- 
 bility of damages in prospect, he feels that the public has been 
 unjust in its generally unfavorable attitude. There were, how- 
 ever, not a few instances of troubles, for some of which the 
 designers, and not Dr. Imhoff, were to blame, and others which 
 were due to mismanagement, and still others to gross overload- 
 ing, and even to entire neglect. The tank is admitted to be deli- 
 cate to operate, and requires constant care and good manage- 
 ment, but this is true of disposal plants generally. 
 
 It was assumed by the designers of the early American Imhoff 
 tanks, that since American sewage is weaker than German, the 
 
Sewage Treatment Experiments 
 
 33 
 
 digestion chamber of the tank might be made smaller. But the 
 truth probably in most cases is that American sewage as a rule 
 forms a sludge with a larger per cent, of water than is the case 
 abroad, and that its volume is actually greater in consequence, and 
 requires more space for digestion. 
 
 Whatever the various causes of foaming may be, it appears 
 certain that sludge arising above the slot-line will cause it, as 
 well as turn the entire tank above and below into an ordinary 
 septic tank. And whenever this condition does occur an active 
 nuisance from smells is inevitable. In spite of all that can be 
 said against the tank, it must be allowed that its advantages are 
 such that a certain amount of the risk of nuisance may be justi- 
 fied. But one should not hide the truth that such a risk exists, 
 and every tank which is not carefully operated will probably 
 cause local trouble. 
 
 The design of these tanks calls for the highest knowledge of 
 sewage treatment, and of the peculiar qualities of sewage solids, 
 grease, etc., and should never be undertaken without the most 
 careful study. Not a few engineers who have ventured into this 
 field look back with sorrow on their first tanks ; and this has had 
 much to do with public feeling regarding the Imhoff tank. 
 
 TABLE X 
 Tank Data, Imhoff Tanks 
 
 fmhoff Tanks 
 1 2 3 
 
 Elev. water line, ft.: • 31.17 31.17 31.17 
 
 Elev. bottom of tank, ft 0.79 9.29 17.50 
 
 Elev. slope intersection, ft 17.21 21.12 24.17 
 
 Diameter of tank, ft 15.00 15.00 15.00 
 
 Depth, total, ft 30.38 21.88 13.67 
 
 Width of settling chamber, ft 10.67 10.67 10.67 
 
 Depth of settling chamber to slots, ft 13.97 10.05 7.00 
 
 Depth of settling chamber to slopes, ft 9.22 5.30 2.42 
 
 Slopes inclined from level, degs 41° 43' 41© 43' 410 43' 
 
 Depth of scum boards, ft 1.54 1.54 1.54 
 
 Depth of baffles, ft 9.50 7.79 5.54 
 
 Capacity below w. 1., cu. ft 1,610.00 1,113.00 557.00 
 
 Capacity below w. 1., gals 12,050.00 8,325.00 4,166.00 
 
 Capacity 1 hr. retention, gals 290,000.00 200,000.00 100,000.00 
 
 Capacity 2 hr. retention, gals 145,000.00 100,000.00 50,000.00 
 
 Digesting chamber diameter, ft 15.00 15.00 15.00 
 
 Digesting chamber depth, ft 16.42 11.83 6.67 
 
 Capacity below slot-line, cu. ft 2,128.00 1,317.00 469.00 
 
 TABLE XI 
 Tank Data, Settling, Digestion and Humus Tanks 
 
 Settling Digestion 
 Tanks 1-4 Tank 
 
 Elev. water line, ft 7.80 -|-6.50 
 
 Elev. top of hopper bottom, ft 3.80 ■ — 4.50 
 
 Elev. bottom hopper bottom, ft 0.20 — 7.00 
 
 Depth below w. 1., ft 7.60 13.50 
 
 Size water surface, ft 8x8 D.5.00 
 
 Capacity, cu. ft 285.00 265.00 
 
 Capacity, gals 2,132.00 1,988.00 
 
 Capacity 1 hr. retention, gals 51,168.00 
 
 Capacity 2 hr. retention, gals 25,584.00 
 
 Humus 
 Tanks 
 
 6.02 
 
 —3.90 
 
 9.92 
 
 5.6x5.6 
 
 127.00 
 
 950.00 
 22,800.00 
 11,400.00 
 
34 Brooklyn, N. Y. 
 
 Theoretic and Observed Retention 
 
 The term "theoretic retention" in this study means the time 
 required to fill a tank at the rate of flow under consideration. 
 It is well known that the true retention is always less than the 
 theoretic, and to what extent the tanks at the station differed from 
 the theoretic was of much interest in the interpretation of phe- 
 nomena. Observation led to the conclusion in advance of study, 
 that the difference between the theoretic and the actual must be 
 considerable. 
 
 Different means of determining this point were tried, and in 
 fact none proved entirely satisfactory. That which was consid- 
 ered the best was determination by means of dyes. Ammonium 
 chloride and sodium chloride were previously tried, but were 
 rejected because of the high specific gravity of their solutions, 
 which caused them to fall to the bottom and pass out very slowly, 
 giving too long a period. 
 
 Fuchsin dye, by reason of its great strength and the ease of 
 comparing its solutions with standards, gave fair results. The 
 dye was applied at the entrance end of the tank, and samples were 
 taken from the effluent at short intervals until it had disappeared. 
 The intensities of color expressed in per cent, of maximum were 
 plotted as ordinates with time intervals from the application as 
 abscissae, and a curve adjusted to them. The area of this curve 
 should represent the total quantity of dye used. 
 1 The portion of the area to the left of any ordinate line should 
 represent the dye which has passed out, and that to the right, the 
 dye still retained in the tank. 
 
 The abscissa of that ordinate that bisects the area should, 
 therefore, be the time of retention. This represents the time of 
 the passage of the sewage, and should be the actual retention 
 period. There are, however, a number of difficulties with the 
 method. To recognize the actual first appearance of color in the 
 effluent is not easy, nor is it easy to recognize the last trace, and 
 the extent to which fine colloidal matter in the sewage and set- 
 tling solids may carry down the color is also an unknown factor. 
 The results of the tests, however, are of much interest and give 
 valuable comparative information. The accuracy of the tests and 
 the conclusions founded on them are questionable, and the results 
 should not be taken as demonstrations of the actual, or anything 
 near the actually true retention, under average conditions of 
 operation. 
 
 A difficulty that appears to be fundamental is, that such ob- 
 servations hold good only for the retention period from which 
 they are obtained. Thus a tank may show the true retention to 
 be 80 per cent, of the theoretic for a 3-hour retention period, 
 while it may also be found only 40 per cent, of the theoretic 
 for a 2-hour retention period. In other words, the tank is not 
 
Sewage Treatment Experiments 
 
 35 
 
 standardized by this method for all periods, and cannot be, for its 
 efficiency is» different with every rate of retention. Each tank, 
 however, has its best retention period. 
 
 As the 2-hour period appeared to be the best suited for tank 
 operation at this location, most of the experimental work was 
 done with this period. 
 
 The following statement shows the result of the Fuchsin dye 
 test on the different tanks : 
 
 TABLE XII 
 
 Imhoff Theoretic Sewage Observed Loss of Suspensa 
 
 Tank Retention Per Day Retention Retention Removed 
 
 No. hours gals. hours % % 
 
 1 2 145,000 1.00 50 61 
 
 2 2 100,000 1.00 SO 56 
 
 3 1 100,000 0.33 68 51 
 
 Settling 
 Tank No. 
 
 3 2 50,000 L77 10 77 
 
 As observed by use of Fuchsin dye. 
 Observations in June-July, 1915. 
 Tanks all working poorly except settling tank 3. 
 Suspensa removed determined by Imhoff cone. 
 
 Effect of Baffling on Imhoff Tanks 
 
 As originally constructed the tanks were not provided with 
 baffles, as it was intended to determine by observation whether 
 these were necessary, and if so, the proper depth and location 
 for them. 
 
 When all the tanks were operated on a 2-hour theoretic reten- 
 tion, during a 3-months period, their comparative efficiency in 
 the removal of settling matter and suspended solids was noted, 
 revealing marked differences between them. The data led to the 
 conclusion that the flow in all three tanks reached down only to a 
 limited depth, regardless of the depth of the settling chamber. 
 The dimensions of these chambers are shown in Table X. The 
 length and breadth of flow is equal in each settling chamber, but 
 the depth and cubic capacity of each is different from the others. 
 To agree with theoretic retention, with the same quantity of 
 sewage flow passed into each tank, the retention period should 
 agree directly with the cubic capacity. To test this theory the 
 three tanks were each put in operation at the rate of 75,000 
 gallons per day, giving the following theoretic retention to each, 
 with the result shown : 
 
 Imhoff 
 Tank No. 
 
 Theoretic 
 Retention 
 
 Sewage 
 Per Day 
 
 Suspensa 
 Removal 
 
 1 
 
 2 
 
 3 
 
 hours 
 
 3.8 
 
 2.7 
 
 1.3 
 
 gals. 
 
 75,000 
 75,000 
 75,000 
 
 % 
 
 29 
 30 
 15 
 
 Rate continued April 1 to June 30, 1914. 
 
 The average results revealed nearly equal removals in tanks 
 1 and 2, and an apparent loss in tank 3. 
 
36 
 
 Brooklyn, N. Y. 
 
 After considerable study on small models and a good deal of 
 discussion on the hydraulics of the tanks, baffles were placed 6 
 ft. from the entrance weir of the chambers, which, being 15 ft. 
 long, left 9 ft. from the baffle to the outlet weir. This was found 
 to be the best location, at least for these tanks, and the baffles 
 remained in this position until 1917, when some further experi- 
 ments were made with the baffles in various situations. For 
 depth of baffles see Table X. 
 
 With the baffles thus located the following data show the re- 
 sults obtained. 
 
 Imhoff 
 Tank No. 
 
 Theoretic 
 Retention 
 
 Sewage 
 Per Day 
 
 Suspensa 
 Removed 
 
 1 
 
 2 
 
 3 
 
 hours 
 
 2 
 
 2 
 
 2 
 
 gals. 
 145,000 
 100,000 
 
 50,000 
 
 % 
 34 
 35 
 44 
 
 July and August, 1914. 
 
 The demonstration was quite convincing that at least for 
 these tanks baffles were required. And it will be noted that the 
 improvement was greatest in the shallowest of the three tanks, 
 which was rather an unexpected result. 
 
 Floating Scum on Imhoff Tanks 
 
 Floating scum was but seldom present in the settling cham- 
 bers, and when present usually consisted mainly of grease which 
 separated out from the sewage matters. The scum boards were 
 quite efficient in preventing floating matters. Such scum and 
 other materials as collected behind the scum boards were paddled 
 from time to time, and settled without causing much trouble. 
 When necessary, materials that did not settle were thrown into 
 the gas vents. 
 
 Floating scum in the gas vents, taking the period from Jan- 
 uary, 1915, to the end of April as representative, was made up 
 as follows : 
 
 TABLE XIII 
 
 Quantities Given in Per Cent, of Total Constituents 
 
 Imhoff Tank 
 1 2 3 
 
 % 
 
 Moisture 80.2 
 
 Solids 19.8 
 
 Mineral 26.0 
 
 Organic 74.0 
 
 Fats 26.6 
 
 Average depth of scum, inches 3.75 
 
 % 
 
 % 
 
 81.5 
 
 84.6 
 
 18.5 
 
 15.4 
 
 27.7 
 
 30.5 
 
 72.3 
 
 69.5 
 
 25.4 
 
 11.2 
 
 1.50 
 
 oy 2 
 
 Effect of Imhoff Tanks on Bacterial Content of Sewage 
 
 The effect of passing sewage thru tanks on the bacterial con- 
 tent of the sewage is a doubtful matter, as the results are so 
 various and differ so much with the different methods employed. 
 The determination of this for the Brooklyn sewage was rather 
 
Sewage Treatment Experiments 
 
 37 
 
 more interesting than important. It was assumed that should 
 disinfection prove necessary a trickling filter plant would be in- 
 stalled, and the question of bacterial removal would arise only 
 in connection with an oxidized effluent. Moreover, the writer 
 believes that much discrimination is called for in selecting and 
 interpreting the results of a purely biological study, and the scope 
 of the experimental work did not require that this be included as 
 of primary importance, but a means of interpreting other work 
 and checking results. In the warmer months of the year with 
 septic sewage and much daily departure from average biological 
 conditions, the results would be masked by these conditions and 
 of no great value, and might mislead. In the cold months no 
 valuable results bearing on the local problem were to be expected. 
 It therefore appeared that the results obtained in the spring, espe- 
 cially March and April, were about the best for this study, and 
 the city was fortunate in having some tests made by experts in 
 this line of work at the very time when the results would be of the 
 most interest, and not marked by the many misleading factors 
 existing in summer and winter. 
 
 The tests referred to were made by Messrs. Lederle & Provost, 
 at the Lederle laboratories in New York, and at the experiment 
 station laboratory, in March and April, 1915, in connection with 
 a study, made with the city's consent, for a client who had re- 
 tained these experts to experiment with and report on ozone 
 disinfection of sewage and effluents. 
 
 The sewage tested was taken from the ordinary crude sewage, 
 and the Imhoff tank effluent represents tank 2 only. The reten- 
 tion period was 2 hours. 
 
 TABLE XIV 
 
 Average Bacterial Counts, and Chemical Constituents With Per- 
 centages of Reduction 
 
 March 24 to April 13, 1915, Inclusive — Parts Per Million 
 
 Crude 
 Sewage 
 
 Dissolved oxygen consumed 148. 
 
 Total suspended matter 256. 
 
 Mineral, in suspensa 72. 
 
 Organic, in suspensa. 184. 
 
 Total solids in sewage 739. 
 
 Total minerals in sewage 361. 
 
 Total organics in sewage 378. 
 
 Nitrates , 0.38 
 
 Oxygen consumed 91 
 
 Tank retention (theoretic), 2 hours. 
 
 Bacterial Counts Per C. C. 
 
 Crude 
 Sewage 
 
 Agar, 24 hrs. at 37oC 250,000 
 
 Gelatin, 48 hrs. at 20 oC 600,000 
 
 Bacteria of B. Coli type 42,000 
 
 — ■ Indicates reduction. 
 -I- Indicates increase. 
 
 Effluent from % 
 
 Imhoff Tk. 2 Reduction 
 
 89 
 
 39.9 
 
 179. 
 
 30.1 
 
 42. 
 
 41.7 
 
 137 
 
 25.5 
 
 618 
 
 16.4 
 
 326 
 
 9.4 
 
 292 
 
 22.8 
 
 0.25 
 
 34.2 
 
 72 
 
 20.8 
 
 Imhoff 
 Effluent 
 
 % Changed 
 
 214,000 
 988,000 
 107,000 
 
 —14.4 
 
 -1-64.7 
 -1-154.3 
 
38 
 
 Brooklyn, N. Y. 
 
 It should be noted that the sewage during the period of these 
 biological tests was considerably stronger than the averages 
 for the two months, parts of which fell into the test period. This 
 was probably caused by the movement down the mains of the 
 sewage system of the deposits that had been formed during the 
 winter, brought about by melting snows and rains. On account 
 of conditions caused by a heavy rain fall, the sewage and tank 
 effluent results for March 26 were excluded from the above table. 
 As these results are of interest as showing what storm conditions 
 can do at this plant, they are given below : 
 
 TABLE XV 
 
 Bacterial Count and Chemical Constituents of Sewage 
 
 March 26, 1915— Parts Per Million 
 
 Crude Effluent from % 
 
 Sewage Imhoff Tk. 2 Reduction 
 
 Dissolved oxygen consumed 143.5 105. 27. 
 
 Total suspended matter 236. 172. 27. 
 
 Minerals in suspensa 74. 26. 70. 
 
 Organics in suspensa _ 162. 146. 10. 
 
 Total solids in sewage 822. 606. 26. 
 
 Total minerals in sewage 380. 352. 7. 
 
 Total organics in sewage 442. 254. 42. 
 
 Nitrates 0.42 0.14 41. 
 
 Oxygen consumed 77.5 67 13. 
 
 Tank retention (theoretic), 2 hours. 
 
 Bacterial Counts Per C. C. 
 
 ^ 
 
 Crude Imhoff 
 
 Sewage Effluent % Changed 
 
 Agar, 24 hrs. at 37 C 20,000,000 4,752,000 —76. 
 
 Bacteria of B. Coli type 1,000,000 258,000 —74. 
 
 — Indicates reduction. 
 
 It is obvious from the above that the storm-sewage problem 
 in Brooklyn is both important and very difficult of solution. The 
 Imhoff tank showed a very gratifying removal both of bacteria 
 and suspensa under the conditions, but the size of tanks required 
 to treat the storm flow would make the tax payers sit up and 
 take notice. A plain sedimentation tank would probably have 
 done the same work. 
 
 Settling Tank 3, and Separate Digestion 
 
 Settling tank 3 (Dortmund type) was 8 ft. x 8 ft. in plan, 
 with a vertical depth of 4 ft., and a hopper bottom, having its 
 apex line 3.6 ft. making the greatest total depth 7.6 ft. Its water 
 line was elevation 7.80. 
 
 The flow, entering thru a pipe, was carried down near the 
 center, and discharged into the tank 5 ft. below the water line. 
 The effluent was taken off thru V-shaped notches in the sides of 
 the outfall trough, that surrounds the top on all four sides'. Eight 
 notches were provided, two on each side. A 6-in. sludge-dis- 
 
Sewage Treatment Experiments 
 
 i 
 
 39 
 
 E - 
 
 
 charge pipe was placed in the center, ending with a bell at the 
 bottom, and having a clean-out at the top above the water line. 
 Sludge was discharged from this pipe thru a horizontal branch 
 placed about 2 ft. below the water line and, passing thru the 
 south wall of the tank, discharged thru a gate-valve into the 
 flume leading to the sludge-drying beds, etc. 
 
 Before passing thru the wall, this horizontal portion of the 
 sludge pipe gave off a branch, consisting of 4-inch pipe controlled 
 by a gate-valve, which passed to a sludge-measuring tank, into 
 which the daily settlings and sludge deposited in the hopper could 
 be discharged and measured, and from which they could be fed 
 slowly by gravity into the sludge digestion tank. (See Fig. 13). 
 
 The "digestion tank", or digesting tank, was of steel plate, 
 made to be air and water-tight, 5 ft. in diameter and 15 ft. deep, 
 with a conical bottom. It was set vertically in the ground, the 
 
40 
 
 Brooklyn, N. Y. 
 
 top of the tank being at elevation -f- 8.00. The lower portion was 
 provided with an outside shell, with an air space between it and 
 the tank. 
 
 A sludge-discharge pipe, and an overflow pipe for water, also 
 were provided ; a "clean out" for the sludge pipe, and a manhole 
 in the top that could be sealed airtight. To test the gases pro- 
 duced in digestion, a small valve was put in the top. 
 
 Operation began with filling the tank with tap water to eleva- 
 tion -|- 6.50, the overflow level. Sludge having been passed from 
 the settling tank hopper to the measuring tank, where test samples 
 were taken, and the quantity measured, was then permitted to 
 enter the digestion tank, the gate valve on the overflow pipe 
 being opened, allowing the water displaced by the entering sludge 
 to escape. 
 
 The operation of this tank was entirely successful ; never at 
 any time were the gases given off offensive ; the surplus water 
 escaping thru the overflow when the tank received its charge of 
 sludge never carried offensive odors. The gases given off thru 
 the valve at the top were odorless, and burned with a colorless 
 flame when ignited, consisting mostly of methane gas. 
 
 Fig. 14. Sludge Digestion Tank. 
 
 This is a method of doing the work of the Imhoff Tank in two 
 separate tanks and is remarkably promising. The sewage settles in 
 a plain sedimentation tank and the settlings are drawn daily under 
 the hydraulic head of the settling tank and discharged into the diges- 
 tion tank ; supernatant water being let out of the overflow to provide 
 room and draw in the sludge and settlings. The digested sludge is 
 dried on Imhoff drying beds, and is of the same quality as Imhoff 
 Tank sludge. 
 
Sewage Treatment Experiments 41 
 
 Sludge ripened in this tank perfectly, and could not have been 
 distinguished from Imhoff-tank ripe sludge. It was full of gas 
 bubbles and dried readily on sludge beds. 
 
 The operating depth of the tank was 13.5 ft. The net cubic 
 capacity for digesting sludge was 265 cu. ft. 
 
 Settling tank 3 had a net capacity 285 cu. ft., and working at 
 a theoretical 2-hours retention treated 25,584 gallons of sewage 
 per day, equal to the sewage of 250 people, the digestion tank 
 thus affording only a space of 1.06 cu. ft. per capita, which with 
 this arrangement of tank appeared to be sufficient for producing 
 a good, ripe sludge. This is much less space than proved neces- 
 sary in the Imhoff tanks themselves, and did not allow much 
 extra room for sludge storage. As soon as ripe sludge appeared 
 in the hopper it was discharged and dried. It is very doubtful 
 if a large tank of this kind would be successful with less sludge 
 capacity per capita than the Imhoff tanks require. In operating 
 a large tank, sludge drying would be much more troublesome on 
 account of the quantity of sludge. A small amount of ripe sludge 
 takes little room to dry, a large amount takes a great deal of 
 room, and drying can only be done in favorable weather. 
 
 It is possible that in the combination of a settling tank with a 
 separate digestion tank, less space is required for digestion than 
 is the case in the digesting chamber of a two story tank because 
 of the fact that raw sludge is discharged into the digestion tank 
 at intervals of 12 to 24 hours, instead of continually ; the settling 
 tank in this case doing part of the work of the digesting chamber-; 
 this at least seems to be a fair inference from the writer's ex- 
 perience. 
 
 Ripening of sludge in the separate digestion tank was much 
 more rapid than in the Imhoff tanks. On being started the tank 
 was seeded with a few gallons of ripe Imhoff sludge, which may 
 have helped. Operation began late in March, and on July 27, 
 four months after starting, 26 cu. ft. of ripe sludge was with- 
 drawn. On September 17 all the ripe sludge was withdrawn, and 
 no further sludge added for digesting, for several months. Dur- 
 ing this period it produced in ripe sludge up to the time of the 
 shut-off, 78 cu ft, and later the unripe sludge remaining after the 
 shut-down digested completely, giving 20 cu. ft. more, in all 98 
 cu. ft. The ripe sludge contained before drying 94 per cent 
 moisture and dried to 64 per cent, moisture ; it averaged about 37 
 per cent, volatile, as compared with the Imhoff sludge which 
 average more than 45 per cent, volatile. 
 
 Effect of Depth of Tank and Capacity of Digestion Chamber on 
 Quality of Imhoff Sludge 
 
 Accurate information on the best depth and the capacity of 
 digestion chamber, altho of extreme importance in the design of 
 
42 Brooklyn, N. Y. 
 
 a permanent sewage-treatment plant, is unfortunately very diffi- 
 cult to obtain, but we believe that the results secured in our work 
 cover our design requirements. 
 
 There appeared to be three ways of securing the above desired 
 information. First, by sounding for the sludge level ; second, by 
 securing samples at different depths, and third, by observations on 
 the behavior of the tanks, when run without sludge withdrawal, 
 thruout a period believed to represent our sludge-storage require- 
 ments, considered in connection with average data on the char- 
 acter of the sludge subsequently withdrawn. 
 
 Determination of the sludge level by means of soundings was 
 at all times subject to uncertainty. The same may be said of 
 the method of obtaining samples at different depths. Sounding 
 and sampling could only be performed thru the gas vents, the 
 shape of which, together with that of the tank bottom, interfered 
 greatly with these methods. Even if the design had not inter- 
 fered with these methods, the phenomena observed in the tanks 
 rendered the observations nearly useless. The ebullition of gas 
 was so active, and caused so much circulation in the contents, 
 that the upper portion of the sludge was disturbed and kept very 
 fluid. The more successful the tank, the more difficult was the 
 operation of finding the sludge level. Even the ripe sludge 
 ready for withdrawal was so fluid that a sounding appliance 
 readily sank thru it. The most that can be said from these ob- 
 servations is that there is a gradual increase in the density of the 
 sludge from the level of the slots to the bottom of the tank. This, 
 at least, was the experience at Brooklyn. 
 
 This condition has been ' described by other observers. At 
 Fitchburg, it was found that the best way to ascertain the amount 
 of sludge present was to use a pump with a hose for a suction 
 line, which was lowered slowly, and the contents of the tank, thus 
 removed from different depths, gave the information sought. It 
 is stated in the report describing the experiments at Philadelphia 
 concerning the Imhoff tank that "one of the mechanical difficul- 
 ties was the inability to determine at what level the sludge stood 
 in the digestion chamber, and altho sludge was withdrawn in 
 small quantities at frequent intervals, it is now believed that at 
 times it was allowed to reach too high a level, so that the ebulli- 
 tion of gas forced it into the settling portion of the tank to the 
 serious detriment of the effluent." 
 
 The third method gave more acceptable data than either of 
 the foregoing. A long run was made for the purpose of deter- 
 mining the effect of depth of tank and capacity of sludge cham- 
 ber on quality of sludge. Beginning on October 22, 1914, the 
 tanks were operated at 2-hour retention without withdrawing 
 sludge until June 22, 1915, when sludge had reached the level of 
 the slot in tank 2, and the effluent showed marked deterioration. 
 The tank had received sewage amounting to 100,000 gallons per 
 
Sewage Treatment Experiments 43 
 
 day thruout the period of observation. The inhabitants contribut- 
 ing sewage were estimated at 1,000. As this period covered about 
 one month more than the assumed non-drying period, it was con- 
 cluded that the storage capacity without reaching the danger 
 point, would have been about correct for the contributing popu- 
 lation. 
 
 As the capacity of the digestion chamber below the slots is 
 1,317 cubic feet, this gives 1.317 cubic feet per capita. Tank 1 
 received 145,000, and tank 3 received 50,000 gallons of sewage 
 per day during the same period. At the time that the sludge in 
 Tank 2 had reached the danger point, that in Tank 1 was ap- 
 parently far below this point, thus showing the sludge storage 
 capacity of Tank 1 was greater than necessary. Tank 3 had not 
 reached the danger point at this time, but this tank was not con- 
 sidered as reliable as Tank 2 for the determination of the best 
 capacity of digestion chamber. Tank 3 was subject to the diffi- 
 culty that unripe sludge was at times discharge by "puncture." 
 This term refers to the following phenomenon : the ripe sludge 
 occupied so small a depth usually, that the overlying unripe sludge 
 would at times be drawn into the sludge-pipe together with the 
 ripe. While the proportions of the settling compartment of this 
 tank were the most favorable for sedimentation of any of the 
 tanks, those of the digestion compartment were the least so, and 
 the reason appeared to be because the cross-sectional area was evi- 
 dently too great for the depth. This rendered the observations 
 for sludge-chamber capacity of less value than those on Tanks 
 1 and 2. 
 
 It will be seen by reference to Table XXII, that the sludge in 
 the three Imhoff tanks had, on the average, the following solid 
 content: Tank 1, 8.9 per cent; Tank 2, 8.9 per cent; Tank 3, 
 6.7 per cent. As related to the sludge-storage capacity, it will be 
 noted that the volume occupied by any particular weight of dry 
 solids in sludge is inversely in proportion to the percentage of 
 solids, so that the per capita yield of dry solids in Tank 3 will 
 occupy a volume, 8.9 divided by 6, equal to 1.3 times the volume 
 occupied by the per capita yield in the other two tanks. Thus if 
 1.33 cu. ft. per capita be the allowance for a tank 20 ft. deep, 
 1.73 would have to be allowed for one 14 ft. deep, owing to the 
 greater water content of the sludge. 
 
 This may be studied in a slightly different manner. If the 
 average solid content be obtained as the result of combining in- 
 dividual figures weighted with the amount of sludge at the par- 
 ticular draft, the solid contents are as follows: Tank 1, 8.8 per 
 cent.; Tank 2, 7.5 per cent.; Tank 3, 5.6 per cent. Calculating 
 the volumes as above, a per capita sludge-chamber allowance of 
 1.33 cu. ft. in a 20-ft. tank would be equivalent to 1.12 cu. ft. in a 
 30-ft. tank, and to 1.86 cu. ft. in a 14-ft. tank. It should be ob- 
 served, however, that the average moisture of sludge drawn from 
 
H 
 
 Brooklyn, N. Y. 
 
 tanks of different depths may be to a considerable degree due 
 to the by-passing of uncompacted but fairly ripe sludge having a 
 larger water content than the sludge at the bottom. 
 
 It would appear that the water content of ripe sludge is largely 
 affected by the proportion of the horizontal cross section of the 
 tank to its depth. In a shallow tank with a wide bottom, the layer 
 of ripe sludge is comparatively thinner than the layer of similar 
 sludge in a deep tank. 
 
 The required drying area of sludge drying beds for local con- 
 ditions was found to be 0.38 square feet per capita. This is 
 subject to assumptions as follows: That all drying shall be done 
 between June 1 and October 1, and that time is required for 
 removal of dried sludge and the preparation of the bed for its 
 next sludge application ; that time should be allowed for stormy 
 periods not infrequently met during the summer. The depth of 
 the wet sludge applied at a time is taken as not to exceed 9 inches. 
 Observations lead us to conclude that there should not be less 
 than ^2 sq. ft. per capita. The best medium for the sludge dry- 
 ing beds was found to be steam ashes, a sand covering being if 
 not undesirable, at least unnecessary. 
 
 Fig. 15. Sludge Drying Beds, from top of Settling Tanks. 
 
 The digestion, storage, and drying of sludge was at no time 
 accompanied by nuisance, even at the time when drying was de- 
 layed by storms. The dried sludge was found suitable for the 
 filling-in of the low-lying meadow, and showed no tendency to 
 further putrefaction. The shrinkage in volume occurring during 
 the first day of drying amounted to about 60 per cent. The 
 period of drying was 4 to 7 days in good weather. In stormy 
 
Sewage Treatment Experiments 
 
 45 
 
 weather it amounted to from 10 to 14 days. When ready for 
 removal- from the bed, the dry sludge was friable and porous, and 
 occupied a volume not exceeding 25 per cent, of its wet volume. 
 
 Fig. 16. Removing Dry Imhoff Sludge. 
 
 The separate digestion tank proved entirely practicable ; the 
 ripe sludge secured from it had all the best qualities of ripe 
 Imhoff-tank sludge. It had, however, one disadvantage over the 
 ordinary Imhoff tank, namely : the necessity of providing a man 
 to discharge the sludge from the settling tank into the digestion 
 tank. For satisfactory operation, this should be done at least 
 twice daily, and in a large plant would add considerably to the 
 expense of operation. As this is not required in the ordinary 
 Imhoff tank, maintenance charges are more in favor of the latter. 
 
 In studying tank performance and the results of these obser- 
 vations on the required capacity of digestion chambers, the writer 
 and his assistants came to the conclusion that the space per cap- 
 ita as determined would be too small for general use in the de- 
 sign of tanks for disposal plants to be operated in the ordinary 
 manner, and that it should be made larger to avoid the danger of 
 foaming. The larger the digestion chamber, the less danger of 
 this appeared to exist in these tanks. Therefore, as a factor of 
 safety it was concluded that 50 per cent, should be added to the 
 figures obtained, when used for purposes of design, giving 2 
 cu. ft. per capita for a tank 20 ft. deep. The smaller the settling 
 chamber the less danger there is of foaming. Shortness of reten- 
 tion, and high rate of operation are more important than high per- 
 centage of removal. The following tables give the results obtained 
 in very short retention periods : 
 
46 Brooklyn, N. Y. 
 
 TABLE XVI 
 Sewage Supplied Experimental Plant 1914-1915 
 
 CRUDE SEWAGE Oct. Nov. ~ " Dec. Jan. Feb. „ Mch, 
 
 Temp. degs. C 19.4 15.9 12.0 Ff6 11.0 13.3 
 
 Settled in cone c.c 1.4 1.6 2.3 2.0 2.4 2.0 
 
 solids: p. p.m. 
 
 Total suspended p.p.m 127 158 201 175 219 203 
 
 Settling in 2 hr. p.p.m 78 97 86 
 
 Colloidal p.p.m 97 122 117 
 
 Volatile p.p.m 101 126 155 142 168 155 
 
 Volatile settling, 2 hr. p.p.m 60 67 59 
 
 Volatile colloidal p.p.m 82 101 96 
 
 oxygen: p.p.m. 
 
 Demand p.p.m 210 278 216 199 213 272 
 
 Dissolved p.p.m 2.2 1.9 2.0 2.8 2.5 1.4 
 
 Saturation % 24 19 18 26 25 13 
 
 Nitrites p.p.m 0.13 0.28 0.37 0.33 0.13 0.31 
 
 Nitrates p.p.m 0.9 0.14 0.13 0.13 0.13 0.16 
 
 oxygen: p.p.m. 
 
 Consumed, 
 Consumed, 
 
 unfiltered p.p.m. 
 filtered p.p.m 
 
 . 51 
 .. 35 
 
 61 
 39 
 
 71 
 44 
 
 74 
 48 
 
 82 
 52 
 
 75 
 50 
 
 
 
 
 Apr. 
 
 May 
 
 June 
 
 July 
 
 Aug. 
 
 Sept. 
 
 Temp. degs. C 15.7 18.0 20.6 22.7 22.5 23.0 
 
 Settled in cone c.c 1.9 1.8 1.6 1.5 1.3 2.3 
 
 solids: p.p.m. 
 
 Total suspended p.p.m :... 197 192 180 147 135 205 
 
 Settling in 2 hrs. p.p.m. 74 69 50 ' 45 43 79 
 
 Colloidal p.p.m 123 123 130 102 92 126 
 
 Volatile p.p.m 149 142 139 103 102 164 
 
 Volatile settling, 2 hrs. p.p.m 48 45 36 26 27 55 
 
 Volatile colloidal p.p.m 101 97 103 77 75 109 
 
 oxygen: p.p.m. 
 
 Demand p.p.m 270 211 276 173 216 305 
 
 Dissolved p.p.m 0.7 1.7 1.0 0.7 0.5 0.1 
 
 Saturation % 7 18 11 8 6 1 
 
 TABLE XVII 
 Performance of Imhoff Tanks, 1914-1915 
 
 IMHOFF EFFLUENT 
 
 TANK 1 Oct. Nov. Dec. Jan. Feb. Mch . 
 
 Temp. degs. C 17.5 16.4 10.1 9.8 11.1 12.1 
 
 Settled in cone c.c 0.5 0.7* 0.9 0.8 0.6 0.6 
 
 solids: p.p.m. 
 
 Total suspended p.p.m 94 118 146 127 142 144 
 
 Settling in 2 hr. p.p.m 31 32 38 
 
 Colloidal p.p.m 96 110 106 
 
 Volatile p.p.m 76 95 117 107 102 110 
 
 Volatile settling, 2 hr. p.p.m 24 16 26 
 
 Volatile colloidal p.p.m 83 86 84 
 
 oxygen: p.p.m. 
 
 Demand p.p.m 113 130 101 165 187 214 
 
 Dissolved p.p.m 0.3 2.1 3.1 1.4 0.6 
 
 Saturation % 3 19 27 13 6 
 
 Consumed: p.p.m. 
 
 Unfiltered p.p.m 45 58 62 60 68 68 
 
 Filtered p.p.m 31 40 40 42 47 48 
 
 Retention hr 2 2 2 2 2 2 
 
 Apr. May June July Aug. Sept. 
 
 Temp. degs. C 15.9 18.1 21.0 23.0 23.9 24.0 
 
 Settled in cone c.c 0.7 0.8 0.6 0.5 0.5 0.7 
 
 solids: p.p.m. 
 
 Total suspended p.p.m 157 154 130 112 102 122 
 
 Settling in 2 hr. p.p.m 40 40 26 26 28 31 
 
 Colloidal p.p.m 117 114 104 84 74 91 
 
 Volatile p.p.m 123 118 104 82 78 95 
 
 Volatile settling, 2 hr. p.p.m 27 25 19 17 15 17 
 
 Volatile colloidal p.p.m 96 93 85 65 63 78 
 
 oxygen: p.p.m. 
 
 Demand p.p.m 225 197 160 135 146 244 
 
 Dissolved p.p.m 0.3 0.4 0.1 
 
 Saturation % 3 4 10 
 
 Retention hr 2 2 2 2 2 2 
 
Sewage Treatment Experiments 47 
 
 TABLE XVIII 
 
 Performance of Imhoff Tanks, 1914-1915 
 
 IMHOFF EFFLUENT 
 
 TANK 2 Oct. Nov. Dec. Jan. Feb. Mch. 
 
 Temp. degs. C 17.5 16.1 9.8 9.0 10.9 12.1 
 
 Settled in cone c.c 0.5 0.5 0.7 0.7 0.5 0.5 
 
 solids: p. p.m. 
 
 Total suspended p.p.m 91 112 132 118 134 134 
 
 Settling in 2 hrs. p.p.m.. 30 30 33 
 
 Colloidal p.p.m 88 104 , 101 
 
 Volatile p.p.m 75 95 110 98 108 105 
 
 Volatile settling, 2 hrs. p.p.m. 24 22 • 21 
 
 Volatile colloidal p.p.m 74 86 84 
 
 OXYGEN : 
 
 Demand p.p.m 66 90 80 136 156 207 
 
 Dissolved p.p.m 0.3 1.5 2.7 1.0 0.3 
 
 Saturation % 3 13 24 10 3 
 
 Consumed: p.p.m. 
 
 Unfiltered p.p.m 47 57 57 58 63 64 
 
 Filtered o.p.m 33 41 38 42 46 48 
 
 Retention hrs 2 2 2 2 2 2_ 
 
 Apr. May June July Aug. Sept. 
 
 Temp. degs. C 15.7 17.7 20.7 23.0 23.6 23.8 
 
 Settled in cone c.c 0.6 0.5 0.5 1.2 0.8 0.9 
 
 solids: p.p.m. 
 
 Total suspended p.p.m .".. 134 125 153 179 113 123 
 
 Settling in 2 hrs. p.p.m 30 24 28 54 35 34 
 
 Colloidal p.p.m 104 101 125 125 7i 89 
 
 Volatile p.p.m 106 98 123 100 106 82 
 
 Volatile settling, 2 hrs. p.p.m 24 19 14 22 26 15 
 
 Volatile colloidal p.p.m 87 84 105 78 80 67 
 
 OXYGEN : 
 
 Demand p.p.m 211 200 223 199 129 222 
 
 Dissolved p.p.m 0.2 0.2 0.2 
 
 Saturation % 2 2 2 
 
 Retention hrs 2 2 2 2 2 2__ 
 
 TABLE XIX 
 
 Performance of Imhoff Tanks, 1914-1915 
 
 IMHOFF EFFLUENT 
 TANK 3 Oct Nov. Dec . Jan . Feb. __ Mch . 
 
 Temp. degs. C 17.6 16.0 10.0 8.6 11.1 11.7 
 
 Settled in cone c.c 0.4 0.4 0.5 0.5 0.3 0.4 
 
 solids: p.p.m. 
 
 Total suspended p.p.m 82 96 112 109 116 118 
 
 Settling in 2 hr. p.p.m 26 23 20 
 
 Colloidal p.p.m 83 92 98 
 
 Volatile p.p.m 69 80 ' 94 88 93 94 
 
 Volatile settling, 2 hr. p.p.m 18 12 12 
 
 Volatile colloidal p.p.m 70 81 82 
 
 OXYGEN : 
 
 Demand p.p.m 106 149 109 135 159 171 
 
 Dissolved p.p.m 0.3 1.3 
 
 Saturation % 3 11 
 
 Consumed: 
 
 Unfiltered p.p.m 45 54 51 50 56 60 
 
 Filtered p.p.m 35 39 34 37 41 44 
 
 Retention hr 2 2 2 2 2 2 
 
 Apr. May June July Aug. Sept. 
 
 Temp. degs. C 15.7 18.8 20.7 22.6 23.6 23.2 
 
 Settled in cone c.c 1.4 1.1 0.9 0.6 0.7 0.4 
 
 solids: p.p.m. 
 
 Total suspended p.p.m 142 149 139 115 106 96 
 
 Settling in 2 hr. p.p.m 61 41 40 28 28 23 
 
 Colloidal p.p.m 121 108 99 87 78 73 
 
 Volatile p.p.m 139 112 111 82 82 75 
 
 Volatile settling, 2 hr. p.p.m 40 25 28 15 18 12 
 
 Volatile colloidal p.p.m 99 87 83 . 67 64 63. 
 
 oxygen : 
 
 Demand p.p.m 265 200 180 134 128 213 
 
 Dissolved p.p.m 0.6 0.4 0.3 
 
 Saturation % 6 4 3 
 
 Retention hr y 2 .y A 1 1 3 3 
 
48 
 
 Brooklyn, N. Y. 
 
 TABLE XX 
 
 Performance of Settling Tank 3, 1914-1915. 
 
 
 
 
 
 
 
 Dec. 
 
 EFFLUENT FROM 
 
 
 
 
 
 
 to 
 
 SETTLING TANK 3 
 
 July 
 
 Aug. 
 
 Sept. 
 
 Oct. 
 
 Nov. 
 
 Apr. 
 
 
 
 
 
 
 
 Closed 
 
 Temp. degs. C. 
 
 Settled in cone c.c 0.7 
 
 solids: p.p.m. 
 
 Total suspended p.p.m 103 
 
 Settling in 2 hr. p.p.m 
 
 Colloidal p.p.m 
 
 Volatile p.p.m 81 
 
 Volatile settling in 2 hr. p.p.m 
 
 Volatile colloidal p.p.m 
 
 oxygen consumed: 
 
 Unfiltered p.p.m SO 
 
 Filtered p.p.m 40 
 
 Retention hr 2 
 
 Apr. 
 
 Temp. degs. C 
 
 Settled in cone c.c 0.5 
 
 solids: p.p.m. 
 
 Total suspended p.p.m 131 
 
 Settling in 2 hr. p.p.m 29 
 
 Colloidal p.p.m 102 
 
 Volatile p.p.m 108 
 
 Volatile settling, 2 hr. p.p.m 23 
 
 Volatile colloidal p.p.m 85 
 
 Oxygen demand p.p.m 
 
 Retention hr 2 
 
 69 
 
 51 
 
 41 
 2 
 
 0.3 
 93 
 
 "76" 
 
 0.4 
 92 
 
 May 
 
 June 
 
 July 
 
 0.5 
 118 
 .„„... 
 
 48 50 64 
 
 35 35 43 
 
 2 2 2 
 
 Aug. 
 
 Sept. 
 
 0.4 
 
 0.2 
 
 0.2 
 
 0.2 
 
 0.2 
 
 113 
 
 108 
 
 100 
 
 96 
 
 103 
 
 22 
 
 14 
 
 18 
 
 13 
 
 12 
 
 91 
 
 94 
 
 82 
 
 83 
 
 91 
 
 93 
 
 88 
 
 77 
 
 78 
 
 84 
 
 16 
 
 11 
 
 12 
 
 8 
 
 6 
 
 77 
 
 77 
 200 
 
 65 
 
 70 
 
 78 
 180 
 
 2 
 
 2 
 
 2 
 
 2 
 
 2 
 
 TABLE XXI 
 
 Removals Effected by Tanks, 1914-1915. 
 
 Per Cent, of Reduction Effected 
 
 Imhoff Tanks 
 1 2 
 
 Settling 
 Tank 3 
 
 volumetric: 
 
 Solids settleable in Imhoff cone 62 
 
 GRAVIMETRIC : 
 
 Total suspended solids 27 
 
 Total settleable solids 53 
 
 Total colloidal solids 26 
 
 Total volatile solids 31 
 
 Oxygen demand 29 
 
 Retention in tanks, hr 2 
 
 64 
 
 78 
 
 80 
 
 27 
 
 40 
 
 40 
 
 62 
 
 66 
 
 61 
 
 12 
 
 21 
 
 21 
 
 31 
 
 46 
 
 40 
 
 34 
 
 41 
 
 20 
 
 2 
 
 2 
 
 2 
 
 TABLE XXII 
 Summary of Sludge Drying Data 
 
 l 
 
 % 
 
 Moisture before drying 91.1 
 
 Moisture after drying 67.8 
 
 Volatile solids 45. S 
 
 Average depth on beds, inches 9. 
 
 Average number of days drying 7.1 
 
 Imhoff Tanks 
 
 
 Separate 
 
 2 
 
 3 
 
 Dig. Tank 
 
 % 
 
 % 
 
 % 
 
 91.1 
 
 93.3 
 
 94.6 
 
 66.2 
 
 69.7 . 
 
 64.1 
 
 46.5 
 
 46.7 
 
 36.6 
 
 9. 
 
 9 
 
 10 
 
 7.3 
 
 7.1 
 
 7.3 
 
Sewage Treatment Experiments 49 
 
 The Riensch-Wurl Screens. 
 
 The fine screens from which the data following were obtained 
 are two in number, of the Riensch-Wurl design, installed at the 
 26th ward treatment plant as a permanent part of that plant. 
 They were completed and commenced operation early in 1916 
 and since then hav Q been in service. 
 
 The Riensch-Wurl screen consists principally of a disc of 
 perforated metal plate, built in sections on a steel frame, with 
 the frustum of a cone built up from its center, the whole being 
 mounted on a shaft whose inclination from the vertical causes the 
 disc to tilt at an angle with the horizon. A close seal is provided 
 between the outer edge of the rotating disc and the stationary 
 setting in which it is placed. To secure this a Z-bar is embedded 
 in the masonry with one of its flanges exactly in the plane of the 
 top of the rim of the frame or setting. The Z-bar carries on its 
 top a large number of small segments, which can be moved in 
 and out, so that the distance between their edges and that of the 
 disc is not greater than the width of the slits or apertures in the 
 disc. When these segments are adjusted, they are held in posi- 
 tion by bolts. 
 
 The plane of the disc as designed in the Brooklyn screens is 
 inclined 25 degrees with the horizon, and about 2/3 of its surface 
 is submerged in the sewage, while the other 1/3 is above the 
 sewage surface. As the disc revolves, the screenings are brought 
 above the sewage and remove by means of revolving brushes 
 carried at the ends of the arms of a large spider. The brushes 
 revolve, as well as the arms and the disc, and the combined result 
 is that every portion of the surface of the screen is brushed over 
 at least twice during each revolution of the disc ; except that the 
 conical portion, which. is cleaned by means of a vertical brush, 
 gets but one brushing. 
 
 The design of the plant provides for a by-pass between the 
 screens so that they may be operated individually or in series, and 
 double screening tried out. Double screening did not prove to be 
 advantageous. The screens are fourteen feet in diameter and 
 operated by steam-engine drive. 
 
 The contract provided that the apertures or slits thru which 
 the screening is effected, should be made of such size as might 
 be found on trial to give the best results with the average flow of 
 sewage at this location. Screening practice and experience else- 
 where indicated that this should be determined experimentally, 
 and with this end in view the specifications provided that four 
 complete sets of screen-surface plates be' supplied, as follows : 
 The first cut with apertures 5/64-inch wide, the second with 
 apertures 1/16-inch wide, the third with apertures 3/64-inch 
 wide, and the fourth with apertures 1/64-inch wide. In each set 
 the apertures were two inches long and staggered in the bronze 
 
50 Brooklyn, N. Y. 
 
 plate, which is 3^-inch thick, each aperture having a counter- 
 sunk cross-section, with the narrow part of the opening on the 
 face of the screen. The 1/64-inch proved too fine and was soon 
 recut to 1/32 inch. 
 
 These sets of screen plates were mounted on the screen frames 
 successively, and tried out, and the aperture dimension 1/16-inch, 
 which was decided to be the most suitable for screening the sew- 
 age at this plant, was selected for permanent installation. The 
 plates not selected remained the property of the contractor. 
 
 It was provided that the size and number of apertures must 
 be such that 6,000,000 gallons of sewage would pass thru the 
 screen in 24 hours, the difference of head of sewage entering and 
 leaving being not more than 12 inches, and that the screen must 
 remove from the sewage practically all particles of suspended 
 matter with a diameter 50 per cent, greater than the cross-section 
 of the aperture. 
 
 The screens were supplied by the Sanitation Corporation of 
 New York, and the design for placing them was made by the 
 engineers of that corporation. In designing the plant, the Bureau 
 of Sewers made provision for the installation of the screens in 
 accordance with the design supplied. The screens were guaran- 
 teed for one year of service, at the end of which they should 
 be in perfect repair, when taken over by the city. No require- 
 ment was made as to the amount, or per cent., of removal which 
 should be made from the sewage, as it was believed that how- 
 ever little or much this might prove to be, the screens were neces- 
 sary to help out the existing sewage treatment plant, then at- 
 tempting to treat four times more sewage than it could fairly 
 take care of. As it was not known what fineness of screens was 
 best for local conditions, this was made the subject of experiment 
 as already mentioned. The experimental work was placed in 
 charge of the experiment station. 
 
 In designing sewage-screening plants, one has but little in- 
 formation to go on. Data are as yet very meager. Sewages 
 differ very widely in the character of suspended solids, and 
 naturally the kind of screens which will serve best with a given 
 sewage, and the required fineness of the screening surface, vary 
 widely. In Dresden, Germany, where there are four large screens 
 of the Riensch-Wurl design, the breadth of the apertures is 2 
 m.m., which is found at that place to give a sufficient removal of 
 suspended matter, and this is the only treatment found necessary 
 for the sewage of that city, which, after screening, is discharged 
 thru multiple outlets directly into the Elbe River, apparently 
 affording a satisfactory disposal of sewage. Dresden has a pop- 
 ulation of between 400,000 and 500,000, and at its lowest stages 
 the river is comparatively a small stream. The writer was told 
 in Dresden that slits or apertures of less width than 2 m.m. 
 reduce the capacity of the screens without much added advantage. 
 
Sewage Treatment Experiments • 51 
 
 This is especially the case where, from the nature of the sus- 
 pended matter, a mat tends to form over the screen surface and 
 produce a straining effect. In Dresden the difference of head 
 thru the screens varies from 2 to 6 c. m., under ordinary condi- 
 tions, but at times is much greater than the larger figure — possi- 
 bly four times as much. 
 
 The two Brooklyn screens were installed early in April, 1916. 
 The north screen was equipped at this time with screen plates 
 perforated with slits 2 inches long and 5/64-inch wide. The 
 south screen was equipped with plates with slits 2 inches long 
 and 1/16-inch wide. 
 
 The following tables give the results obtained at this time on 
 dry weather flow sewage on the official test : 
 
 TABLE XXIII 
 Operation of Riensch-Wurl Screens 
 
 April-May, 1916 
 Capacity and Head Developed 
 
 Screen 
 
 Slit 
 inches 
 
 
 Sewage Screened Average Loss of Head 
 in 24 hrs. — gals. at Screen — inches 
 
 North 
 
 South 
 
 5/64 
 1/16 
 
 
 8,077,851 0.118 
 8,204,113 0.121 
 
 Removal Effected 
 
 as Shown 
 
 by 
 
 Crude Sewage and Effluent 
 
 Crude Sewage Screen Effluent Removal 
 
 North Screen — p. p.m. p.p.m. p.p.m. % 
 
 Suspended solids 159 137 22 14.0 
 
 Settling solids 75 60 15 20. 
 
 Non-settling 84 77 7 8.7 
 
 South Screen — 
 
 Suspended solids 193 158 35 18.0 
 
 Settling solids 109.5 79.5 30 27.4 
 
 Non-settling 83.5 78.5 5 6.0 
 
 The screens continued in operation with 5/64-inch and 4/64- 
 inch slits until the spring of 1917. 
 
 The following table gives a fair picture of the operation of 
 the screens in comparison with the Imhoff tanks. The data are 
 averages from all of the tanks taken together, and of both 
 screens averaged. The crude-sewage influents are also averaged. 
 The table includes dry weather flow only, storm water data as 
 far as possible are excluded, and also all abnormal data. Screen 
 removal was at times subject to wide variations and to extremes, 
 these are excluded as far as possible. 
 
52 • Brooklyn, N. Y. 
 
 TABLE XXIV 
 
 Sewage Imhoff Tanks R. W. Screens 
 
 Inf't. Eff't. % Red. Eff't. % Red. 
 
 Solids, by cone, c.c.l. l.S 
 
 Solids, susp. tot. p.p.m 171. 
 
 Solids, settlings p.p.m 61 
 
 Solids, colloid p.p.m 110 
 
 Solids, volatile p.p.m 133 
 
 Solids, vol. settling p.p.m 41 
 
 Solids, vol. colloid p.p.m 92 
 
 Oxygen demand p.p.m 259 
 
 0.6 
 
 66 
 
 0.9 
 
 50 
 
 103. 
 
 40 
 
 137 
 
 20 
 
 29 
 
 52 
 
 45 
 
 26 
 
 81 
 
 26 
 
 98 
 
 11 
 
 82 
 
 39 
 
 107 
 
 19 
 
 17 
 
 59 
 
 
 
 73 
 
 21 
 
 
 
 152 
 
 41 
 
 182 
 
 30 
 
 Theoretic retention, all tanks, 2 hours. 
 
 Sewage passing tanks per day: 
 
 Tank 1 145,000 gallons 
 
 Tank 2 100,000 
 
 Tank 3 50,000 
 
 295,000 gallons 
 
 Sewage screened (both screens) 12 Mgd. 
 
 North screen 5/64 in. aperture 
 
 South screen 4/64 in. aperture 
 
 Screenings removed (both screens) 14.75 cu. ft. per Mg. 
 
 Weight per cu. ft 72 lbs. 
 
 Moisture 81 % 
 
 In preparing the above table much discrimination has been 
 called for, that the presentation might be fair. Storm removals 
 often go above 50 per cent of the suspensa. Dry weather removal 
 sometimes falls nearly to 0. These extremes mislead. Average 
 performance has been sought for. To obtain, as well as select 
 these data has been very difficult. 
 
 To include the abnormal is to render all conclusions useless. 
 It should be remembered that screens always remove the screena- 
 ble solids, — if these are absent they remove 100 per cent, of 
 nothing. 
 
 In 1917 the following changes were made in the screens: The 
 north screen was changed to a screening surface with apertures 
 3/64-inch wide, and the south screen to a surface with apertures 
 1/32-inch wide. In both cases the apertures were 2 inches long. 
 
 The weather was very unfavorable for making observations 
 on screening dry weather flow, on account of the prevalence of 
 frequent rainy days and showers. The screens were designed for 
 receiving both dry-weather flow and storm water. It was difficult 
 to obtain satisfactory data on dry-weather flow, and many ob- 
 tained were useless on account of storm interference. 
 
 A remarkable difference was observed in the dry-weather 
 sewage discharged into the quieting tank, from which the Imhoff 
 tanks were supplied, and that which entered the screens. To 
 study this phenomenon occupied several weeks of valuable time. 
 Meanwhile the screens operated on a very weak sewage. Un- 
 fortunately the sewer and grit chamber could not be properly 
 examined for obstructions without cutting off the sewage supply 
 to the experimental plant, and as the flow to the Imhoff tanks 
 was normal in quality it was decided not to do this until the 
 July-August-September series of tests were compiled. In the 
 
Sewage Treatment Experiments 
 
 53 
 
 latter part of September the pumps were stopped and the sewer 
 and grit chamber cleaned ; an obstruction was then found, made 
 of tin cans, rags, sticks, stones, silt, etc., formed diagonally across 
 the chamber above the entrances to the screens, where it had 
 acted as a submerged weir, interfering greatly with the dry- 
 weather flow to the screens, but not at all with the flow to the 
 by-pass from which sewage was obtained for the tanks. This 
 did not much interfere with the storm water flow, which readily 
 passed over it. 
 
 The screen data for this period are of value only as showing 
 what removal might be expected from a very weak sewage, — 
 from which much of the suspensa had settled. Under the condi- 
 tions much more effect was produced on the sewage than would 
 have been anticipated in advance. 
 
 TABLE XXV 
 
 Operation of Imhoff Tank 2, on Fairly Strong Sewage, and R. 
 
 W . Screens on Abnormally Weak Sezuage, Compared 
 
 Influent to Influents to 
 
 Crude Sewa ge _ Imhoff Tank 2 South Screen North Screen 
 
 Temp. deg. C 26. 26\ 26\ 
 
 Solids by cone c.c.l 1.9 1.4 0.64 
 
 Solids susp. total p.p.m 183 136 119 
 
 Solids volatile p.p.m 143 106 92 
 
 Solids mineral p.p.m 40 30 27 
 
 Dissolved oxygen p.p.m 0.5 1.0 1.0 
 
 Oxygen demand p.p.m 245 137 110 
 
 Filtered p.p.m. 144. 112 88 
 
 Effluent Effluent R. W. Screens 
 
 Effluents Imhoff Tank 2 South 1/32" North 3/64" 
 
 Tank retention, 2 hr. % Red % Red % Red 
 
 Solids by cone c.c.l 1.1 44 1.2 14 0.5 22 
 
 Solids susp. total p.p.m 138. 25 133 2 117 2 
 
 Solids volatile p.p.m 109 24 108 .... 92 
 
 Solids mineral p.p.m 30 25 25 18 25 4 
 
 Dissolved oxygen p.p.m .... 10 10 
 
 Oxygen demand p.p.m 198 15 126 8 90 18 
 
 Filtered p.p.m 119 17 77 31 65 26 
 
 South Screen North Screen 
 
 Sewage screened per diem 6.4 Mgd. 8.4 Mgd. 
 
 Average amount of screenings per diem 13.2 cu. ft. 13.9 cu. ft. 
 
 Average amount of screenings per mg 2.06 cu. ft. 1.65 cu. ft. 
 
 Average weight of screenings per cu. ft 72 lbs. 72 lbs. 
 
 Computed from averages 
 
 It will be noted that the sewage entering the screens was in 
 nearly every item of its constituents weaker than the effluent 
 leaving the Imhoff tank. It may be presumed, perhaps, that had 
 the effluent from the tank passed thru screens, it would have been 
 improved. The volume of screenings removed from such a weak 
 sewage is worthy of remark. 
 
 Unfortunately the screenings produced were not weighed. The 
 weights given in the table are an average weight for similar 
 screenings determined by a number of experiments. 
 
 After the sewers were cleaned out, the following data were 
 obtained, the screens operating as above on dry weather sewage. 
 
54 
 
 Brooklyn, N. Y. 
 
 Imhoff tank 2 being also supplied with the same sewage, which 
 was not as strong as the yearly average by about 10 per cent., 
 but was not to be considered abnormal. It will be observed that 
 while the screen improved its removal, the tank showed the 
 effect of a weaker sewage : 
 
 TABLE XXVI 
 
 Oct.-Nov.-Dec, 1917 
 
 Imhoff Tank 2 
 Sewage 2 Hrs. Retention 
 
 Inf't. Eff't. Rem'l. % Red 
 
 Solids by cone c.c.1 1.4 0.8 0.6 44 
 
 Solids susp. total p.p.m 141 116 25 18 
 
 Solids volatile p.p.m 118 98 20 17 
 
 Solids mineral p.p.m 23 18 5 22 
 
 R. W. Screens 
 
 Sewage given above South Screen 1/32" North Screen 3/64" 
 
 Efft. Rem'l. % Red Efft. Rem'l. % Red 
 
 Solids by cone c.c.l 0.5 0.9 64 0.6 0.8 54 
 
 Solids susp. total p.p.m 116 25 18 117 24 17 
 
 Solids volatile p.p.m 93 25 21 94 24 20 
 
 Solids mineral p.p.m 21 2 9 22 1 4 
 
 Sewage screened per diem 5 Mgd. 5 Mgd. 
 
 Screenings per diem 54 cu. ft. 52 cu. ft. 
 
 Weight pe r cu. ft 73.7 lbs. 74.2 lbs. 
 
 Screened sewage was applied to the trickling-filter beds during 
 the period beginning July 1 and ending December 31, 1916, with 
 tl\e object of comparing the results obtained with those from Im- 
 hoff -tank effluent. The filter beds had been operated with Imhoff- 
 tank effluent previously to receiving the screened sewage. They 
 had been out of service during February and March, but at the 
 time were in good condition tho they had not fully regained their 
 purifying power. For this reason and in order that the effect of 
 the Imhoff effluent might have time to pass off, the results of the 
 first month of operation are not included in the table which fol- 
 lows : 
 
 TABLE XXVII 
 
 Trickling Filter No. 3, Operated on Riensch-Warl Screen Effluent 
 
 Aug. -Sept. -Oct., 1916 Apertures 1/16 and 5/64-inch 
 
 Sewage Effluent from Filter No. 3* 
 Screened Depth of Stone Medium 
 Average 4 ft. 6 ft. 7'3" 8'6" 
 
 Temp. degs. C 19.4 18.7 18.5 17.7 
 
 Suspended solids, 
 
 Total p.p.m 153 92 86 84 114 
 
 Volatile p.p.m 110 85 61 61 79 
 
 Diss, oxygen, p.p.m 3.1 3.7 3.6 4.1 
 
 Per cent, of sat 33 39 38 43 
 
 Relative stability, 
 
 Unfiltered per cent 37 95 93 97 
 
 Filtered per cent 70 100 100 100 
 
 Oxygen dem. bioch. 
 
 Unfiltered p.p.m 129 72 40 44 40 
 
 Filtered 70 30 19 18 16 
 
 Nitrite nitrogen p.p.m 1.6 2.6 3.0 3.4 
 
 Nitrate nitrogen p.p.m 1.1 4.1 3.8 4.6 
 
 Oxidized nitrogen p.p.m 2.7 6.7 6.8 8.0 
 
 * Rate of application 4 Mgd. per acre. 
 Size of filter medium 1)4 to 2]/ 2 inches. 
 
 10' 
 
 From 
 
 Humus 
 
 Tank 
 
 18.0 
 
 114 
 
 26 
 
 79 
 
 22 
 
 4.3 
 
 2.8 
 
 45 
 
 30 
 
 99 
 
 100 
 
 100 
 
 
 43 
 
 23 
 
 18 
 
 00 
 
 3.0 
 
 3.2 
 
 5.4 
 
 5.7 
 
 8.4 
 
 8.9 
 
Sewage Treatment Experiments 55 
 
 The settled effluent from the 10-foot depth was uniformly 
 stable, and the results on filtered samples from all other depths 
 except the 4-foot indicated that after settling out the humus the 
 effluent would be satisfactory. 
 
 The rates of operation on filters Nos. 1, 2, 3 were set at 4 
 mgd. per acre, filter No. 4 being kept at the rate of 2 mgd. These 
 were the same rates as had been employed in 1915, and previously 
 to applying the screened sewage, except that filter bed No. 1 was 
 increased in an effort to test it to destruction as regards clogging. 
 
 There seemed no especial tendency to clog the beds. It was 
 threatened on bed No. 1, but did not develop. The sewage ap- 
 plied to the filter beds was at all times fresher than Imhoff-tank 
 effluent would have been. No unpleasant smells were at any 
 time in evidence. 
 
 In the results obtained from the operation of the Riensch- 
 Wurl screens at this station, it is to be observed that the average 
 percentage removal of settling matter is low. When the sewage 
 enters the screen chamber it comes in with a rush strong enough 
 to force many of the finer particles thru the screen. These par- 
 ticles tend to settle behind the screen, but do not accumulate 
 there, as they are carried away by the effluent. The samples for 
 analysis taken from the effluent were obtained as it flows out into 
 the channel leading to the measuring weir, and may have con- 
 tained an undue proportion of the particles above mentioned. 
 It is difficult to obtain a sample from the effluent which shall 
 be as representative of the work done as a sample taken from 
 the Imhoff-tank effluent would be. A good deal of solid matter 
 which really is screenable as well as settleable and would be re- 
 moved under proper conditions, is forced thru the screens by the 
 high rate of approach. 
 
 A serious difficulty was experienced in obtaining reliable re- 
 sults on the work of the fine screens, due to the fineness of the 
 analytical methods employed. It is obvious that a test for sus- 
 pended solids by the Gooch-crucible method, or for settling mat- 
 ter by the Imhoff cone, cannot take into account the grosser par- 
 ticles of matter that may be present in the sewage, lest the results 
 be totally incongruous. This is partly remedied, of course, by 
 noting the time required to fill a bucket with screenings, and the 
 volume of sewage observed. The bucket takes into account both 
 the fine material and the coarser matter, but unfortunately there 
 seems to be no rational way of crediting the screenings removed 
 to the removal of total suspended solids. It sometimes happened 
 that by the determinations according to Standard Methods, A. P. 
 H. A., the effluent contained more suspended matter than the 
 influent, notwithstanding the fact that the screen was removing 
 350 to 500 lbs. of screenings per million gallons. One is tempted 
 to ask the question, after this experience, are our methods of 
 determining suspended solids above improvement or criticism? 
 
56 Brooklyn, N. Y. 
 
 If we do not obtain correct data on the influent to the screens, 
 obviously we do not for the tanks either. Some discussion on this 
 point seems desirable. 
 
 Cleaning of the Screens 
 
 Twice a day, a jet of water from a hose was played on the 
 revolving disc. The water struck the screen with considerable 
 force, and the slits were thoroly cleaned thereby. This treat- 
 ment was followed by pouring about four ounces of kerosene 
 on each of the four revolving brushes. This application not only 
 removed the grease from the brushes, but also from the screens. 
 In this way it was possible to keep the screen clean and bright at 
 all times. 
 
 Removal of Screenings 
 
 When a bucket was filled with screenings, it was removed and 
 an empty bucket replaced. The time was recorded. In this way the 
 time required for filling each bucket was obtained. The full 
 bucket was hoisted by a pulley and emptied into a special car. 
 The car was rolled to the dump about 50 yards away and emptied 
 on steam ashes. During the summer this dump was sometimes 
 offensive, and became infested with flies, but this could have been 
 prevented at all times by applying a slight amount of lime and 
 covering with a thin cover of soil which was purposely omitted 
 at this time in order that the tendency to cause nuisance might 
 be observed. With the coming of cooler weather, tendency to 
 cause nuisance ended. When the screenings are dry, nuisance is 
 practically absent. It is desirable to treat the screenings shortly 
 after their removal, so that all nuisances will be avoided. Com- 
 posting with ordinary earth has proved sufficient to prevent all 
 nuisance. Artificial drying and incineration are recommended as 
 a method of disposal. 
 
 Samples of the screenings were examined for moisture con- 
 tent, volatile matter and ash. The volatile matter and ash were 
 determined on the dry specimen. Samples for analysis were 
 collected in an 8j^-inch porcelain dish, and the test for moisture 
 content was made on a portion of this sample in a 4-inch evap- 
 orating dish. Care was taken to obtain as representative a sam- 
 ple as possible. The moisture determination was made by evap- 
 orating the water in a weighed sample on a water bath, for 
 twenty hours. The sample was then further dried in a hot-air 
 oven at 100 deg. C, for four hours more. The loss in weight 
 was determined and the percentage of water calculated. The 
 volatile matter was determined gravimetrically by igniting a care- 
 fully weighed portion, in a Wiesnegg muffle furnace, until the con- 
 tents were at a red heat. The loss in weight gave the volatile 
 matter, and the difference from the original weight gave the ash. 
 
Sewage Treatment Experiments 
 
 57 
 
 The following table gives the results of screenings from the 
 south screen. Apertures 1/32-inch. 
 
 Analyses of Screenings 
 
 Moisture Volatile Matter Ash 
 
 Date Per Cent. Per Cent. Per Cent. 
 
 July 27, 1917 88.7 87.8 12.2 
 
 July 30, 1917 86.0 81.6 18.4 
 
 Aug. 1, 1917 (a) S9.0 67.0 33.0 
 
 Aug. 1, 1917 (b) 76.8 68.3 31.7 
 
 Aug. 7, 1917 (c) 66.4 82.0 18.0 
 
 Aug. 10, 1917 71.6 82.0 18.0 
 
 Aug. IS, 1917 87.3 74.8 25.2 
 
 Averages 83.4 81.6 18.4 
 
 : f 
 
 The averages exclude the three special samples. 
 
 (a) Sample of storm screenings. 
 
 (b) Sample of screenings taken near the end of the storm of August 1. 
 
 (c) Sample taken after water in screenings had been permitted to drain off. 
 
 Time required for one complete revolution of screen 2.4 minutes. Time required for one 
 complete revolution of brushes 7.5 seconds. 
 
 Results from the north screen follow. Apertures 3/64-inch. 
 
 Moisture Volatile Matter Ash 
 
 Date Per Cent. Per Cent. Per Cent. 
 
 Aug. 21, 1917 82.0 80.9 19.1 
 
 Aug. 23, 1917 80.0 77.4 22.6 
 
 Aug. 24, 1917 86.8 93.5 6.5 
 
 Aug. 27, 1917. 89.5 95.6 4.4 
 
 Aug. 31,1917 85.9 78.8 21.2 
 
 Sept. 5, 1917 89.0 90.5 9.5 
 
 Sept. 12, 1917 90.4 94.0 6.0 
 
 Averages 86.3 87.2 12.8 
 
 Time required for one complete revolution of screen 2.3 minutes. 
 Time required for one complete revolution of brushes 7.8 seconds. 
 
 It will be seen from this table that the water content of screen- 
 ings is generally high. Another important fact is the relatively 
 large amount of organic matter. The weight of the screenings 
 varied from 64 to 80 lbs. per cu. ft. the average being about 72 
 lbs., which is rather high ; this is probably due to the age and 
 condition of the sewage, and considerable grit. 
 
 Storm Water Operation of R. W . Screens 
 
 It was evident on studying the data obtained that storms pre- 
 sent such varied phenomena that averages were highly mislead- 
 ing. It was also obvious to the observers that the ordinary meth- 
 ods of sampling and examination by Gooch crucible, and even 
 by the Imhoff cone, were of little or no value, as these could not 
 possibly give any correct result when used on matters of such 
 large size as the storm water brought down — such as tin cans, 
 old shoes, rags in quantity, as well as all grades and sizes of 
 suspended solids. It was concluded, and we believe correctly, 
 that the best means of exhibiting storm-screen performance was 
 to give the amount of screenings removed by the screen, and to 
 state how many gallons of storm sewage carried one cubic foot of 
 screenings. 
 
58 Brooklyn, N. Y. 
 
 The following tables are constructed on the principle just 
 stated. 
 
 The first column gives the time at which a bucket holding 
 8.6 cu. ft. of wet screenings was filled, and replaced by an empty 
 bucket. 
 
 Operation on dry-weather flow following the storm, is in- 
 cluded for comparison. 
 
 The second column gives the time in minutes required to fill 
 the bucket with screenings. 
 
 The third column gives the head lost by the flow in passing 
 the screen, and the size of the screen apertures. 
 
 The fourth column gives the cubic feet of flow which has 
 passed thru the screen and over the knife-edge weir, while the 
 bucket was filling. 
 
 The fifth column gives the gallons of sewage needed to produce 
 1 cu. ft. of screenings. 
 
 The sixth column gives the volume of wet screenings in parts 
 per million of the sewage out of which it was taken by the 
 screens, regardless of its water content. 
 
 Other data are added below the table giving general results. 
 
 TABLE XXVIII 
 R. W . Screen, Typical Operation 
 
 North Screen, Apertures 5/64-inch 
 June 7, 1916, storm began 8:00 A. M., ended 11:00 P. M. 
 
 
 
 
 5/64" 
 Slits 
 
 
 Wet 
 
 
 Time 
 
 
 Quantity Screened 
 
 Screenings 
 
 Bucket 
 
 
 Required 
 
 Head 
 
 Q. by to 1 cu. ft. 
 
 to Volume 
 
 Filled 
 
 
 to Fill 
 
 Lost 
 
 Weir Screenings 
 
 of Sewage 
 
 hour min. in. cu. ft. gal. p.p.m. 
 
 10:00 a. m 120 0J4 62,640 54,450 137 
 
 11:00 60 54 31,440 27,340 274 
 
 11:30 30 54 15,780 13,720 544 
 
 12:00 30 54 15,840 13,770 543 
 
 12:40 p. m 40 y 2 20,600 17,910 412 
 
 1:30 50 y 2 25,100 21,830 334 
 
 2:10 40 54 19,040 16,560 452 
 
 3:30 80 V 2 36,400 31,670 236 
 
 4:45 75 54 33,750 29,380 255 
 
 6:35 110 )/ 2 49,973 43,500 172 
 
 8:00 85 y 2 39,474 43,050 215 
 
 9:50 110 y 2 53,570 46,550 161 
 
 11:00 end 70 y 2 37,310 32,480 230 
 
 Dry weather flow June 8 
 
 6:00 a. m 420 J4 194,880 169,546 44 
 
 8:15 p. m 135 54 63,640 55,366 135 
 
 Contents of bucket 8.60 cu. ft. 
 
 1. Storm operation, 15 hours: 
 
 Storm-sewage screened, 440,917 cu. ft. equals 3,306,900 gals. 
 
 Screenings removed, 13 buckets equals 111.8 cu. ft. equals 33.6 cu. ft. per mg. 
 
 Minimum flow screened per cu. ft. of screenings, 13,720 gals. 
 
 Average moisture of screenings during storm, 60 per cent. 
 
 Volatile matter in dry solids, 70 per cent.; mineral, 30 per cent. 
 
 2. Dry weather operation, 954 hours: 
 
 Dry-weather flow screened, 258,820 cu. ft. equals 1,940,000 gals. 
 Screenings from dry weather flow, 17.2 cu. ft. equals 8.8 cu. ft. per mg. 
 Average moisture dry weather flow screenings, 82 per cent. 
 Volatile matter in dry solids, 83 per cent.; mineral, 17 per cent. 
 
Sewage Treatment Experiments 
 
 59 
 
 TABLE XXIX 
 R. IV. Screen, Typical Operation 
 
 South Screen, Apertures 4/64-inch July 25, 26, 1916. See data below 
 
 4/64" Wit 
 
 Time Slits Quantity Screened Screenings 
 
 Bucket Required Head Q. by to 1 cu. ft. to Volume 
 
 Filled to Fill Lost Weir Screenings of Sewage 
 
 hour min. in. cu. ft. gal. p. p.m. 
 July 25 7:00 a. m. Storm flow 
 
 9:00 120 0.y 2 89,100 77,600 97 
 
 1:00 p. m 240 yi 168,000 146,000 51 
 
 3:30 p. m 150 y 2 105,450 91,740 82 
 
 4:00 30 M 22,560 19,600 380 
 
 4:30 30 H 19,500 17,000 441 
 
 6:15 end 105 H 65,100 56,800 132 
 
 Dry weather flow 
 
 11:00 p. m 285 T / 2 213,750 186,000 40 
 
 July 26 4:20 a. m 320 y 2 208,000 181,000 41 
 
 8:30 250 y 2 160,100 139,200 54 
 
 11:30 end 180 y 2 108,000 94,000 80 
 
 Storm flow 
 
 1:00 p. m 90 y 2 50,200 43,700 171 
 
 1:15 p. m 15 3/ 4 8,360 7,265 1,030 
 
 1:30 end 15 U 8,360 7,265 1,030 
 
 Dry weather flow 
 
 9:00 p. m 45C yi 312,200 271,700 28 
 
 11:00 p. m 120 y 2 99,600 86,600 86 
 
 July 27 1:20 a. m 140 yi 116,200 101,200 74 
 
 3:00 a. m 100 y 2 55,800 49,550 154 
 
 8:30 a. m 270 y 175,500 152,6 00 49 
 
 Contents of bucket 8.60 cu. ft. 
 
 1. Storm operation, 13 y% hours: 
 
 Storm-sewage screened, 436,699 cu. ft. equals 4,025,000 gals. 
 
 Screenings removed, 9 buckets equals 77.4 cu. ft. equals 19.2 cu. ft. per mg. 
 
 Minimum flow screened per cu. ft. of screenings, 7,265 gals. 
 
 Average moisture of screenings during storm, 63 per cent. 
 
 Volatile matter in dry solids, 73 per cent.; mineral, 27 per cent. 
 
 2. Dry weather operation, 35 y hours: 
 
 Dry weather flow screened, 1,449,150 cu. ft. equals 10,868,000 gals. 
 Screenings removed, 9 buckets equals 77.4 cu. ft. equals 7.5 cu. ft. per mg. 
 Average moisture, dry weather screenings, 80 per cent. 
 Volatile matter in dry solids, 81 per cent.; mineral, 19 per cent. 
 
 TABLE XXX 
 R. IV. Screen, Typical Operation 
 
 North Screen, Apertures 3/64-inch 
 August 24, 1917, storm began 1:20 P. M., ended 8:24 P. M. 
 
 3/64" Wet 
 
 Time Slits Quantity Screened Screenings 
 
 Bucket Required Head Q. by to 1 cu. ft. to Volume 
 
 Filled to Fill Lost Weir Screenings of Sewage 
 
 hour min. in. cu. ft. gal. p. p.m. 
 
 1:35 p. m 15 ltf. 6,264 5,450 1,372 
 
 2:00 25 1 10,440 9,080 826 
 
 5:10 190 H 79,344 69,000 108 
 
 6:10 60 % 24,056 20,900 358 
 
 6:50 40 34 16,724 14,550 514 
 
 7:40 50 y 20,880 18,150 414 
 
 8:25 end 45 y 18,792 16,350 457 
 
 Aug. 25 Dry weather flow 
 
 12:25 a. m 240 y 100,224 87,300 86 
 
 9:00 515 y 215,064 187,100 40 
 
 8:30 p. m 690 y 288,144 250,9 00 30 
 
 Contents of bucket 8.60 cu. ft. 
 
 1. Storm operation, 7 hours, 5 minutes: 
 
 Storm-sewage screened, 176,500 cu. ft. equals 1,323,700 gals. 
 
 Screenings removed, 7 buckets equals 60.2 cu. ft. equals 46.3 cu. ft. per mg. 
 
 Minimum flow screened per cu. ft. of screenings, 5,450 gals. 
 
 Average moisture of screenings during storm, 83 per cent. 
 
 Volatile matter in dry solids, 87.5 per cent.; mineral, 12.5 per cent. 
 
 2. Dry weather operation, 24 hours, 5 minutes: 
 
 Dry weather flow screened, 603,432 cu. ft. equals 4,525,700 gals. 
 Screenings removed, 3 buckets equals 25.8 cu. ft. equals 5.7 cu. ft. per mg. 
 Average moisture of dry weather screenings, 86.8 per cent. 
 Volatile matter in dry solids, 93.5 per cent.- mineral, 1.5 per cent. 
 
60 Brooklyn, N. Y. 
 
 TABLE XXXI 
 R. W . Screen, Typical Operation 
 
 South Screen, Apertures 2/64-inch 
 August 1, 1917, storm began 11:35 A. M., ended 8:10 P. M 
 
 2/64" Wet 
 
 Time Slits Quantity Screened Screenings. 
 
 Bucket Required Head Q. by to 1 cu. ft. to Volume 
 
 Filled to Fill Lost Weir Screenings of Sewage 
 
 hour min. in. cu. ft. gal. p.p.m. 
 
 11:50 a. m 17 2. 6,925 6,020 1,243 
 
 12:50 p. m 60 0.J4 20,904 18,200 412 
 
 3-00 130 y 2 36,270 ■ 31,550 237 
 
 4:10 70 y 2 3,255 2,930 2,641 
 
 4-45 35 H 3,255 2,930 2,641 
 
 5:30 45 H 6,300 5,440 1,365 
 
 8:10 end 150 y 27,900 24,300 309 
 
 Aug. 2 Dry weather flow 
 
 7:45 a. m 705 y 2 278,400 242,000 31 
 
 11:45 a. m 240 y 94,800 82,430 91 
 
 11:45 p. m 720 y 2 284,500 247,500 30 
 
 Contents of bucket 8.60 cu. ft. 
 
 1. Storm operation, % l / 2 hours: 
 
 Storm-sewage screened, 104,809 cu. ft. equals 786,000 gals. 
 
 Screenings removed, 7 buckets equals 60.2 cu. ft. equals 76.6 cu. ft. per mg. 
 
 Minimum flow screened per cu. ft. of screenings, 2,930 gals. 
 
 Average moisture of screenings during storm, 59 per cent. 
 
 Volatile matter in dry solids, 67 per cent.; mineral, 33 per cent. 
 
 2. Dry weather operation, 27J4 hours: 
 
 Dry weather flow screened, 657,700 cu. ft. equals 4,932,700 gals. 
 Screenings from dry weather flow, 25.8 cu. ft. equals 5.2 cu. ft. per mg. 
 Average moisture, dry weather screenings, 77 per cent. 
 Volatile matter in dry solids, 68 per cent.; mineral, 32 per cent. 
 
 Sezvage Treatment by Oxidation 
 
 So much space and time have been taken in the presentation 
 of sewage data, and methods of removing suspended solids from 
 sewage, that the subject of oxidation of sewage must be elim- 
 inated altogether, or given very little space. To cut it out en- 
 tirely would be to leave the other data presented incomplete, as 
 it is very desirable to observe the effect of tank treatment and 
 fine-screen treatment, from the standpoint of the trickling filter 
 bed. All that can be attempted, however, will be to give some 
 space to this subject for that purpose. The treatment of sewage 
 by oxidation was by far the most extensive and, perhaps, the 
 most important work done at the Brooklyn experiment station. 
 The results of these studies cannot be presented in a limited 
 space or at this time. But if the official report of the work is not 
 published before the next annual convention of this Society, the 
 writer will arrange, if possible, to make it the subject of a pa- 
 per. Under this head must also be included the treatment of 
 sewage by aeration and activation, both having been given ex- 
 tensive study at the station, but which on account of lack of space 
 cannot be included here. 
 
 The term "trickling filter" instead of "sprinkling filter" was 
 adopted in conformity with the recommendation of the Commit- 
 tee on Nomenclature of the American Public Health Association. 
 
Sewage Treatment Experiments 
 
 61 
 
 PLAN 
 
 TRICKLING FILTERS GROUP A 
 
 12 3 4 5 6 
 
 Fig. 17. 
 
 This is based on the essential features of the filter rather than on 
 the method of application of the sewage, and therefore seems a 
 logical term. 
 
 The trickling-filter beds were all supplied with sewage by 
 means of dosing tanks. 
 
62 
 
 Brooklyn, N. Y. 
 
 Each dosing tank was provided with a 5-inch Miller siphon 
 which discharged the dose into an inverted pyramidal feeding 
 tank, from the bottom of which it was carried, by a pipe em- 
 bedded in the medium, to the sprinkling nozzle, by which it was 
 sprayed over the bed. These tanks and the inverted pyramidal 
 feeding tanks into which they discharge were constructed of yel- 
 low pine. 
 
 Fig. 18. Trickling Filters. 
 
 The elevation of the dosing-tank water line at the instant of 
 siphon discharge was 26.74 feet. The water-line in the inverted 
 pyramidal feeding tanks was controlled by the amount of sewage 
 discharged from the dosing tanks, and its maximum elevation 
 with the largest dose that could be discharged was 23.50 feet, 
 which was 6 ft. 2 in. above the surface of the beds. This head 
 could be varied by a movable bulkhead placed in the dosing tank, 
 which increased the number of doses per hour. 
 
 As the bulkhead was moved toward the siphon the effective 
 capacity of the tank was diminished. In the upper part of the 
 bulkhead were placed circular orifices cut in thin sheet copper, 
 each of which delivered half a million gallons per acre daily on 
 the beds, under a head held constant by means of an adjustable 
 outlet pipe. By stopping an appropriate number of these orifices 
 with corks the rate on the filters was maintained as desired. These 
 orifices were subject to hourly inspection and did not show any 
 tendency to clog with Imhoff effluent. Their accuracy could at 
 any time be checked by calibration in the dosing tank, and this 
 was frequently done. 
 
Sewage Treatment Experiments 
 
 63 
 
 Description of Filters 
 
 There were two groups of trickling filter beds which have 
 been indicated as Group A and Group B. Each will be described 
 separately. 
 
 Group A consisted of the ordinary type of trickling filters. 
 There were four beds. The foundations were carried on piles, 
 and the bottoms of the beds were 6 feet above the level of the 
 marshland over which they were built. High water at unusually 
 high tides washed underneath, at times covering the marsh to a 
 depth of 6 or 8 inches. Each bed was square and had an effective 
 area of .005 of an acre. 
 
 Partitions, 4 inches thick, were carried from the floor to the 
 surface of the medium between adjacent beds. 
 
 The outer walls of the beds were formed by means of rein- 
 forced concrete piers, carrying a slab coping of reinforced con- 
 crete at the tip. The piers were cast with slots for receiving 
 3-inch yellow-pine slabs or shutters, which were set at an angle 
 of 45 degrees, sloping inward, spaced 2 feet apart, against which 
 the medium rested, affording a maximum admission for air and 
 preventing the escape of sewage. 
 
 The floor was formed of concrete with a slight slope to the 
 outlet of the underdrains. Half-tile, 6 inches in diameter, were 
 laid with the convex sides up on the concrete floor to afford 
 drainage ; the effluent flowing to gutters or troughs placed around 
 the bottom of the beds, outside of the wall piers. Each bed had 
 its individual gutter, which discharged by means of a pipe or 
 flume to its individual secondary settling tank. 
 
 Fig. 19. Trickling Filter Bed, after Two Years' Service. 
 
64 
 
 Brooklyn, X. Y. 
 
 ■ ■ . H-f . 
 
 L jfe 38 1 
 
 
 6 
 
 a 
 
 CD 
 
 mgm ~ H 
 
 1 
 
 
 
 Fig. 20. Trickling Filter Bed after Two Years' Service. 
 
 The medium was 10 feet in depth over the top of the under- 
 drains, and of very carefully selected broken trap-rock, many 
 runs thru screens having been necessary to obtain the result re- 
 quired. 
 
 In order to prevent the effects of wind on the sewage dis- 
 tribution, a shield was provided, consisting of a board fence at 
 the surface carried between the beds and around them. 
 
 In order that the effect of the depth of filter medium under 
 similar conditions of operation might be obtained, test trays were 
 placed in the filter beds at different depths. The trays were so 
 placed that samples could be taken at depths from the surface of 
 4 feet, 6 feet, 7 feet 3 inches, and 8 feet 6 inches. Samples 
 taken from the bottom of the bed give the result of 10 feet 
 depth. 
 
 For the distribution of sewage to the surface of the filter, sev- 
 eral types of nozzles were tried, but for the majority of the ex- 
 periments, which were at high rates, the Taylor square spray and 
 the Worcester nozzles were employed. 
 
 Sice of Medium in Filter Beds, Group A 
 
 Bed No. 1, stone passing ring \y 2 inches in diameter, retained 
 by ^4-inch ring. 
 
 Bed No. 2, stone passing ring 2 inches in diameter, retained 
 by 1-inch ring. 
 
 Bed No. 3, stone passing ring 2 l / 2 inches in diameter, retained 
 by \ l /4 -inch ring. 
 
 Bed No. 4, stone passing ring 2 l / 2 inches in diameter, retained 
 by 1 ^4-inch ring. 
 
Sewage Treatment Experiments 
 
 65 
 
 Fig. 21. Trickling Filter Bed after Two Years' Service. 
 
 Fig. 22. Trickling Filter Bed after Two Years' Service. 
 
 Group B was a tank 12 feet in diameter and 16 feet high in 
 which were placed two beds. A partition wall divided the tank 
 into two equal parts and each side was filled with stone filtering 
 medium to the same depths. The bottom of each side was under- 
 drained with 6-inch half-pipe tile, on a concrete bed, and closed 
 from external air by the tank walls. Each side was kept entirely 
 from the other, and drained independently to a secondary tank. 
 Each side was provided with a grid for supplying compressed 
 air, placed within the medium near the bottom, formed of j^-inch 
 
66 
 
 Brooklyn, N. Y. 
 
 iron pipe, perforated every 6 inches with 3^-inch holes, thru 
 which compressed air could be supplied. 
 
 In operation the sewage was sprayed upon the surface by 
 a single nozzle placed at the center of the two beds, over the 
 dividing wall between them. Two designs of Taylor circular- 
 spray nozzles, and the Worcester nozzle, were tried during the 
 course of the experiments. Both beds could be operated as or- 
 dinary trickling filter beds, in which case air entered the bed 
 from the surface only. Compressed air could be supplied to both 
 beds at the same time, or one side could be operating as an ordi- 
 nary trickling filter, while the other was operated as a trickling 
 filter with compressed air added in the bed, in order that the 
 effluents might be compared, and the effect of the added air ob- 
 served. The filtering medium was of the best selected broken 
 trap-rock. The depth of the medium was 10 feet, and might be 
 increased up to 16 feet, if desired. 
 
 Fig. 23. Surface of Trickling Filter 3, after Six Months' Service on 
 Screened Sewage. 
 
 Size of Medium in Filter Beds, Group B 
 
 Bed No. 5, stone passing ring 2 l / 2 inches in diameter, retained 
 by 1^4 -inch ring. 
 
 Bed No. 6, stone passing ring 2y 2 inches in diameter, retained 
 by l*4-inch ring. 
 
 Size and Character of Stone for Trickling Filters 
 
 Every effort was made to obtain an accurate knowledge of the 
 stone which was put into the filters. To this end, very frequent 
 samples were taken during the delivery of the material at the 
 
Sewage Treatment Experiments' 67 
 
 plant, and all stone which did not conform very closely to the 
 provisions of the specifications were rejected and the contractor 
 was required to re-screen it. 
 
 The specifications for filter No. 4, calling for stone from 2-inch 
 to 2^-inch size, were found too severe, when the effort was 
 made to screen the stone to the specified size ; so, after many un- 
 successful attempts, the contractor was allowed to furnish stone 
 which would pass a 2^2-inch ring and be retained upon l^-inch 
 ring. 
 
 The following table shows the percentage by weight of the 
 stone in each filter which did not lie within the specified limits : 
 
 Filter No. 1 4.7% 
 
 Filter No. 2 6.3% 
 
 Filter No. 3, 5, 6 2.3% 
 
 Filter No. 4 3.7% 
 
 The data obtained by analyses of samples were as f^Uows : 
 (1) median size, (2) uniformity number, (3) superficial area, 
 and (4) per cent, of voids. 
 
 The "median size" is a number somewhat analogous to the 
 "effective size" of filter sand, but is defined as follows : it is the 
 diameter of a ring of such size that fifty per cent, (by number) 
 of the stones will be retained upon it and fifty per cent, pass it. 
 It is generally not much different from the mean of the diameters 
 of the rings defining the sample in the specifications. The "uni- 
 formity number" is analogous to the "uniformity coefficient" 
 used in water filter sand, and is the ratio of the diameter of the 
 ring thru which 75 per cent, (by number) of the stones will pass 
 to that of the ring thru which 25 per cent, will pass. It ap- 
 proaches 1.00 as the material approaches perfect uniformity, and 
 increases with the non-uniformity of the sample. The per cent, 
 of voids was determined by weighing the water necessary to fill 
 the void spaces in an amount of stone whose gross volume was 
 known. 
 
 TABLE XXXII 
 
 Properties of Broken Trap Rock For Use in Filters 
 
 Filter 1 Filter 2 Filters 3, 5, 6 Filter 4 
 
 Specified size (inches) V/2-H 2-1 2J4-1J4 lYi-Wi 
 
 Median size (inches) .94 1.32 1.83 2.14 
 
 Uniformity number 1.50 1.56 1.42 1.20 
 
 Superficial area per ton (sq. ft.) :.... 1025 711 568 465 
 
 Superficial area per cu. yd. (sq. ft.) 1343 945 741 579 
 
 Per cent, void space _ 46.8 45.0 46.5 47.0 
 
 On examining the above table, it will be noted that the super- 
 ficial area is roughly proportioned to the reciprocal of the median 
 size. This would, of course, be strictly true, if the material 
 consisted of uniform spheres. 
 
 The most uniform sample (filter No. 4) has the highest per- 
 centage of voids, and the least uniform (filter No. 2) has the 
 lowest percentage, which is also what should be expected. 
 
68 
 
 Brooklyn, N. Y. 
 
 Determination of Stone Surfaces 
 
 Since the degree of purification obtained in a trickling filter is 
 dependent, among other factors, upon the number of bacteria 
 which can act upon the sewage, and the number upon the amount 
 of surface upon which they can multiply, a knowledge of the 
 superficial area of the stone, per ton or per cubic yard is desir- 
 able. Hitherto, so far as the writer's knowledge goes, no one 
 has published any data upon this subject, and an attempt was 
 made here to arrive at knowledge of it, and the laws of its 
 variation. 
 
 The method employed to obtain this information was as fol- 
 lows : Samples of broken granite and of broken trap were ob- 
 tained and screened thru sieves having rings of the following 
 diameters: 2y 2 " , \]/ A " , 1", %", J / 2 ". About 100 stones of each 
 size were kept for area and weight determination. The area was 
 determined by tracing with a sharp pencil the shapes of the faces 
 of the stones on a sheet of paper. As the faces of the stones are 
 almost always nearly plane, this presented no serious difficulty, 
 aside from extreme laboriousness. The areas were then deter- 
 mined by planimeter. 
 
 To enable the information to be applied to the various sizes 
 of stone, as put into the filters, the ratio of the mean area to the 
 area of a sphere, of the same material of equal weight (or equiv- 
 alent sphere), seemed the easiest to use. The following table 
 Sfives the results of the studv : 
 
 TABLE XXXIII 
 Analyses of Stone for Trickling Filter Beds 
 
 Weight Surface 
 
 Ratio of 
 Pieces mean surf. 
 
 Sizes in Total Average Total Average to surf, of 
 
 Inches sample grams grams sq. ft. sq. ft. equiv. sphere 
 
 H-V2 
 
 Granite 100 265 2.65 .561 .0056 1.08 
 
 Trap 100 415 4.15 .881 .0088 1.12 
 
 1-H 
 
 Granite 102 630 6.18 1.049 .0103 1.12 
 
 Trap 100 829 8.29 1.423 .0142 1.30 
 
 154-1 , 
 
 Granite 98 1320 13.46 f 2.115 .0216 1.40 
 
 2^-i y A 
 
 Trap 100 1855 18.55 2.428 .0243 1.29 
 
 Granite 108 7289 67.50 6.049 .0560 1.24 
 
 Trap 100 6330 63.30 5.015 .0502 1.18 
 
 The filter beds were made of trap rock exclusively. In the 
 above table, results of analyses from granite samples are included 
 for comparison. 
 
 Observation of the samples shows that among the pieces there 
 are a greater or less number of stones which are ''scales" or 
 "plates", whose thickness is small compared with the other di- 
 
Sewage Treatment Experiments 69 
 
 mensions. The relatively high values of the above ratio- in the 
 sizes from 1/4" to 1*4" indicate a rather high proportion of these 
 thin stones in the above mentioned sizes. 
 
 As the stone for the filters was delivered, samples were taken 
 and screened by sieves of plate containing circular holes having 
 diameter varying by %" . The total weight of a size of separation 
 divided by the number of stones, gives the mean weight, from 
 which the area of the equivalent sphere may be found from the 
 following formula : 
 
 A = J 363 -", 16W ' x .0010765 
 
 \ 
 
 in which A is area of equivalent sphere in square feet; W, the 
 mean weight of a single stone is grams ; g, the specific gravity of 
 the rock ; and .0010764 the constant for reducing sq. cm. to sq. ft. 
 A suitable multiplier selected from the following table enabled 
 the true mean superficial area corresponding to the mean weight 
 to be calculated. 
 
 TABLE XXXIV 
 
 Ratio of mean area to 
 
 Mean weight area of equiv. sphere 
 
 Grams 
 
 4 1.12 
 
 S 1.15 
 
 6 1.20 
 
 7 : 1.2S 
 
 8 1.29 
 
 9 to 20 1.29 
 
 2S 1.27 
 
 30 1.26 
 
 35 / 1.25 
 
 40 1.24 
 
 45 1.22 
 
 50 1.21 
 
 55 1.20 
 
 60 1.19 
 
 80 1.18 
 
 100 1.17 
 
 120 1.15 
 
 140 1.14 
 
 160 1.13 
 
 180 1.12 
 
 200 1.10 
 
 220 1.09 
 
 This table of multipliers was interpolated from the data in 
 the first table, and from an area determination performed subse- 
 quent to the first studies. The first table carried the knowledge 
 of the ratio only to a mean weight of 60 grams, and the subse- 
 quent determination enabled it to be known when the mean weight 
 was 225 grams. 
 
 Having the mean area of a single stone and the number, of 
 stones of that separation size in the sample, it is simple to com- 
 pute the total area per ton, and after weighing a known volume 
 of the sample, to compute the total area per cubic yard, as put 
 into the filter. 
 
70 
 
 Brooklyn, N. Y. 
 
 > 3 » 
 
 x § a 
 
 •^ •& § 
 
 J a .2 
 
 n 1 h 
 
 2 « 1- 
 
 
 to 
 
 "5* 
 
 > 
 
 < 1 
 
 
 ^^ bC* 
 
 a 
 
 5 
 
 o 
 
 .5 
 
 o 
 
 C<l 
 
 
 CD 
 
 £ 
 
 lab 
 lid 
 
 > 
 
 
 P 1 ^ bb 
 
 
 d 
 
 3 
 
 1-9 
 
 00 
 
 OOiOHMWOONH 
 
 "^ bb 
 
 
 S 1 
 
 f-iOi-H *<r O O CNl <— I 
 
 ^ bb 
 
 
 p. 
 
 Os CO 
 1>. »C i— i CO O O "*f ■•— 1 
 
 Lr : "^ fail 
 
 
 to 
 
 s 
 
 bb 
 ;* 
 
 H 
 
 
 ic-^r^cooC'O'Or— oa 
 
 t— kO i— • CO O O CO i— i 
 
 ^ ^i bb 
 
 
 3 
 
 coiC'— ' co c c ir h 
 
 ^1 bb 
 
 
 3" 
 
 00 CO 
 CO iO i— i Tf* o o CNl i— < 
 
 = "* tb 
 
 
 a 
 < 
 
 COCO 
 
 NOH COO O"* CN| 
 
 ^ ^ bi 
 
 
 s 
 
 bb 
 
 < 
 
 OJffirH CO *>■ OS CO C^l 
 
 ^° bb 
 
 .s 
 
 Q 
 
 
 a 
 
 3 
 
 1-9 
 
 CM CO 
 
 CNieNIOSCO'OOOO'rpcO 
 
 NiOH co oo o io cq 
 
 ^^ bb 
 
 
 >> 
 
 03 OO »— i COOOiCiM 
 
 =>° bb 
 
 
 ft 
 
 < 
 
 oo »o 
 
 OS CO *— 1 cOCOQCf^Ol 
 
 *■:"? bb 
 
 
 _bb 
 
 > 
 
 < 
 
 CO (N O CO *-• N O OO CI 
 I>- iO i-t COOlO^f-H 
 
 *i*~. bi 
 
 .3 
 
 o 
 
 
 a 
 
 3 
 >-9 
 
 
 =5=i bb 
 
 
 >> 
 
 CM l>; 
 
 ■HNtCJcONOOW'-' 
 
 eo-*tii-H co O O oi i— * 
 
 ^^ bb 
 -**1 
 
 
 ft 
 <! 
 
 COCNlCOCOCOCOOOrOS 
 (>. tfr ,-h CO N C O M 
 
 «!"* bb 
 
 
 S 
 
 bb 
 
 > 
 
 •a) 
 
 HHr^n«opooo 
 
 ^ ^ bi. 
 
 13| 
 
 .s 
 
 O 
 
 o 
 
 
 d 
 
 3 
 1-9 
 
 OOCCO'OCON'- • 
 Orf*N CO O C IM h 
 
 ■^^ bb 
 
 
 >> 
 
 Ct) 
 
 CO *0 
 
 tDcomcoiccc?;- - 
 
 (MOO'—' COOCKH 
 
 -^ bb 
 oq 
 
 
 a 
 
 < 
 
 OS t^ 
 
 CCCO^COCDCOCCO 
 NIC'-' COOON 
 
 '" ' l> * bi 
 
 -^ = 2 
 
 
 CN 
 
 <® 
 
 3 
 
 a 
 H 
 
 "g 
 
 -= 
 
 a 
 
 > 
 
 < 
 
 <M CNl CO ' CNl CM 
 
 | -<t-^H ^H O 
 
 
 >< 
 
 X 
 
 y. 
 - 
 
 c 
 
 "c3 
 
 a 
 
 3 
 
 -r 
 
 i 
 
 'c 
 
 CO 
 
 
 d 
 
 3 
 
 1-5 
 
 i co o 
 
 IhcoOOO g 
 
 
 ] 
 
 OS 
 
 1 ^^ 
 
 1 OOON 'CO CO 
 t^ O i— < Cs 
 
 
 
 ft 
 
 -< 
 
 I oqoq 
 
 i lO i*0> U3 ' d CO 
 
 
 
 a) 
 
 | 
 
 co 
 
 3 
 
 o 
 
 > 
 <* 
 
 OS CO CO ^H cm I— 
 
 00 -* ^- w CO 
 
 
 3 
 
 1-9 
 
 1 CO p 
 OOOhh O 
 coco oq rH CO 
 
 
 
 >> 
 
 S3 
 
 CNJ CM CO ^- 00 — * 
 Os -^ i— ' <— i — < 
 — ^- CM 
 
 
 < 
 
 t- os io " r— o 
 
 
 
 
 Os 
 
 Tot. susp. solids 
 Vol. u 
 Temperature 
 Diss.ox. p.p.m. 
 
 " " % sat. 
 Rel. stability 
 
 " (paper-filt.) 
 Diss. ox. demaud 
 
 ■ (filt.) 
 
 Nitrogen as 
 
 nitrites 
 
 nitrates 
 Dailv rate per acre 
 
 
 
 1 
 
 »«cs 
 
 
 
 
 35 O 
 
 
 
 > 
 
 
 . t^ 
 
 — o • 
 
 
 o> 
 
 ■** 1 
 
 CO(N^h NO 
 
 
 
 
 a 
 
 
 
 
 
 
 a 
 
 l>-cO 
 
 
 !M CO - 
 
 
 3 
 
 ^^OIWCOO 
 
 .^tl 
 
 (M <M . 
 
 
 1-9 
 
 CO CNl i-l CM O 
 
 ;*^ 
 
 ^^ 
 
 
 >. 
 
 Tt-*-< 
 
 
 O =o • 
 
 
 a 
 
 
 
 .oo . 
 
 CN OS . 
 
 
 3 
 
 w 
 
 a 
 
 CO CM »—l CO O 
 
 ; ^ 
 
 
 
 ^ 
 
 ^H OS 
 
 
 iO o ■ 
 
 
 
 a, 
 
 
 
 -«00 
 
 
 
 < 
 
 ■*COH CM O 
 
 '. ^ 
 
 
 
 
 
 
 
 
 > 
 
 rHCfilOTjiTfOOOlCirO'H^^ 
 
 O CO i—i ■•# OS O CO i— • — < ^ 
 
 
 
 •41 
 
 
 
 
 
 
 C 
 
 o^ ^^ bb 
 
 
 d 
 
 NCO'tOOONHHi'H 
 O CO CM ^OONh i-*<l 
 
 
 
 
 
 
 
 
 
 
 
 
 >> 
 
 CO iO CO ^ 
 
 
 
 
 Nco«;^«cojoioc-iHrtw 
 
 OS CO i-« 'tCRON-H ^h neq 
 
 
 
 2 
 
 
 u; 
 
 °;^: ^i"? bb 
 
 
 
 D, 
 
 OOW"*MiOONir;^05^ 
 HtDH ^CJOiO- « ^ 
 
 
 
 < 
 
 
 
 
 
 
 
 ^^ ^^ bi 
 
 
 
 
 OCUDCONOOCN'HOq 
 
 
 
 
 
 
 
 
 
 3 
 
 °.°? rt ! l> : bb 
 
 
 
 iCNONNOOOmNlNM 
 N»C(M COOOCO^ ^H^ 
 
 
 o 
 
 
 >. 
 
 «'* ^^ bb 
 
 
 QO 
 
 
 ^i— imit^ocoxHC;^ 
 
 
 ^ 
 
 ■Q. 
 
 
 
 j- 
 
 ots ^- o ^ 
 
 Q 
 
 
 O. 
 
 CMcOt^COCOiOOCOCO'— *00«*H 
 OS CO t-H COOCO^'-i "< 
 
 
 < 
 
 *- 
 
 
 • 
 
 "=2^ ^'^ bb 
 
 
 
 ^(MNCONCOt^C^X 1 ^ 
 
 
 
 
 
 
 
 
 <o 
 
 
 d 
 3 
 
 to ° ^ bb 
 
 g 
 
 
 NHOCOMOOXChhw 
 
 b- io cm co o o co cni ^»«=; 
 
 lO 
 
 &3 
 
 >> 
 
 °°.^ ^^ bb 
 
 
 OJCuO-^rHOOCCOO'-iCO^ 
 cO-**^-l ■^lOO'CO-^ *^ 
 
 - 
 
 *~ 
 
 2 
 
 sr 
 
 
 ^ 
 
 ^oo ^^ bb 
 
 i 
 
 
 D. 
 
 ON'tCONCOi.OW'-iN'H 
 CO uO t-h CO O O CO CM t^* 
 
 t-< 
 
 
 <1 
 
 
 
 
 -!°°. °:^ bb 
 
 
 
 
 OON(MCC(DOOCO'-I00^ 
 NiOH CNl OS O -<f i— i <^i 
 
 53 
 
 
 <!j 
 
 ^ 
 
 c 
 
 3 
 1-9 
 
 ^°. ^:^: bi 
 
 
 CCiOCM M O O CO -H ^h^ 
 
 ,—1 T^ 
 
 S 
 
 SO 
 
 o 
 
 >. 
 
 OS CNl CO OS y, 
 
 
 
 iHO>ccO'-'a30'Hoo^HN | «?) 
 
 I>- iO »— i COCjO^h <5 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 1 iO^TfCOC01CCC--CM 
 O0 CO '— < CO OS o -^ CM i*=; 
 
 ft, 
 
 
 < 
 
 
 
 
 -1-1 ^^ bi 
 
 
 
 > 
 
 lO^CClM'MNClNCOO^ 
 t^- IC »-< CNl i— i CNl lO <* i< 
 
 
 
 < 
 
 Ja 
 
 
 
 •^ 
 
 a 
 
 i "^ f - ■* c 1 bb 
 
 F^ 
 
 
 NHHrt3COiCXC:CC^ 
 COCOM rt^COiOCO ^ 
 
 -* 
 
 <-> 
 
 
 >> 
 
 ^:^ ^^ bb 
 
 
 
 
 NOONOl^'-CNLOOCC O ^ 
 CD^<«H CNl i— < CNl uO CO *£> 
 
 
 
 a 
 
 
 fZ 
 
 «« ^^ bb 
 
 
 
 
 ujfoioiMcoa'^CJcooo'^ 
 
 1>- iO <— I NiHtMi^^ . »^5 
 
 
 
 < 
 
 
 -3 
 
 
 
 
 
 
 
 
 
 
 
 ja 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 Os 
 
 a 
 
 
 
 
 
 a 
 
 ft 
 
 
 
 
 O 
 
 
 d 
 
 
 
 
 
 ft 
 t5 
 
 : Sjai 
 
 
 D. 
 
 
 T3 
 
 i 3 . 
 
 g 3 
 
 
 
 Q 
 
 
 » ft . CO ^ . g 1 >, 
 
 
 
 
 
 
 
 H> 
 
 E- 
 
 n 
 
 « 
 
 
 a 
 
 ^ 
 
 u 
 
 iw 1-^*1 
 
 "S£ H 
 S.25 
 
Sewage Treatment Experiments 
 
 71 
 
 The following numerical example will illustrate. The screen 
 analyses of the stone put into filters No. 3, 5 and 6 gave the fol- 
 low data : 
 
 Passing 
 
 Size 
 
 Retained by 
 
 Total Weight 
 lbs. 
 
 No. Stones 
 
 2/ 2 
 
 2 
 
 154 
 
 I/2 
 
 1 
 
 2 
 
 m 
 
 Hi 
 1 
 
 118.5 
 
 67.5 
 
 23.2 
 
 16.0 
 
 5.0 
 
 .27 
 
 306 
 247 
 145 
 157 
 
 85 
 7 
 
 Beginning with the 2y^ 
 118.5x454 
 
 -2 size, the mean weight of a single 
 
 stone is 
 
 306 
 
 = 176 trains. From the formula for sur- 
 
 face-equivalent sphere, we find this weight corresponds to .0843 
 sq. ft. From the table of ratios, we find the proper multiplier 
 for size 176 grams to be 1.12; hence the true mean area of a 
 single stone is .0943 sq. ft. This multiplied by 306 gives 28.9 
 sq. ft. for the area of the stones of the sample lying between 2 1 /z' r 
 and 2" . Proceeding in the same way, we find the total area of 
 the size 2 — 1^4 to be 18.8 sq. ft.; of the size l-}4 — 1^ to be 
 8.0 sq. ft.; of the size \ l / 2 to 1% to be 6.6 sq. ft.; of the size 
 1^4 to 1 to be 2.9 sq. ft.; and of the size 1 — ^4 to be 2 sq. ft., 
 making a total area for the sample (which weighed 230.47 lbs.) 
 of 65.4 sq. ft., or 568 sq. ft. per ton. It was found that the stone, 
 as placed in the filter, weighed 1.305 tons per cubic yard, so that 
 the area amounted to 741 sq. ft. per cubic yard. 
 
 Operation of Trickling Filters 
 Arrangement of the Beds and Final Settling Tanks : 
 
 Filter No. 1 Discharges thru Tank No. 6 
 Filter No. 2 Discharges thru Tank No. 10 
 Filter No. 3 Discharges thru Tank No. 9 
 Filter No. 4 Discharges thru Tank No. 7 
 Filter No. 5 Discharges thru Tank No. 5 
 Filter No. 6 Discharges thru Tank No. 8 
 Area of beds, Nos. 1 to 4, inclusive, .005 acre 
 Area of beds, Nos. 5 to 6, inclusive, .00127 acre 
 
 During the course of the experiments the sewage distributed 
 on the trickling filters consisted of (1) Imhoff tank effluent; 
 (2) crude sewage which had been subjected to fine screening. 
 The former was experimented with from November 1, 1913, 
 when the filters were first put into operation, to December 15, 
 1915. During the latter part of December, 1915, and during the 
 winter and spring of 1916, the filters were shut down, pending 
 the completion of the Riensch-Wurl screen plant. Crude sewage, 
 
72 
 
 Brooklyn, N. Y. 
 
 > 
 
 -i 
 
 X 
 
 3 
 
 X 
 
 i- 
 
 X 
 
 
 
 
 Ed 
 
 >, 
 
 J 
 
 a 
 
 CQ 
 
 £ 
 
 < 
 
 £ 
 
 H 
 
 V} 
 
 
 co 
 
 > 
 
 < 
 
 CO 
 
 CO l>; 
 
 COCOCOCMCOOCOCO 
 •*}" CM •*** Oi O -** CM 
 
 
 .5 
 
 5 
 
 o 
 
 .s 
 
 o 
 
 = 
 
 a 
 o 
 
 
 to 
 
 s 
 
 bb 
 H 
 
 > 
 
 < 
 
 CO 
 
 CO 
 
 CO GO 
 "OWMMOOCO-* 
 ^ tM ^OOMN 
 
 '".". bb 
 -* 
 
 
 d 
 
 cc 
 
 co r-^ 
 
 f-: 3 ? bb 
 
 
 bb 
 
 3 
 < 
 
 CO 
 
 SS^SSoS^ 
 
 
 
 "a 
 
 1-5 
 
 o 
 
 CO CM 
 
 n .°*. bb 
 
 
 s 
 
 is 
 
 > 
 
 § 
 
 tJ< tM " " -*■ s - 5 *■*< A3 
 
 ^^ fci 
 "2S 
 
 
 d 
 
 CO 
 
 IC 
 
 ■* «' K c c o - o 
 ■**- CM TCO'-t 
 
 a ° , ° bb 
 ~ 2 ! 
 
 
 bb 
 
 3 
 
 <! 
 
 CO 
 
 c: GO 
 
 ^coco^*oooo 
 
 r* CM -rf C O -cf — 
 
 -*"* bb 
 
 Tt 1 
 
 
 . 
 
 3 
 l-S 
 
 lO 
 
 GO O 
 COCMCOOCMOC-:CM 
 
 »o cm ^roc^- 
 
 ^ ^ ti 
 
 ^ 5 2S 
 
 
 -* 
 
 s 
 
 bb 
 
 <! 
 
 cC 
 
 ■* c-i co r- a c oo o 
 
 ■*=•* bb 
 
 NOOg 
 
 
 d 
 
 O 
 
 CO 
 
 q cm 
 
 CONMOOOO'H 
 ^ CM WCCO-* 
 
 ^r" bb 
 
 ^2S 
 
 CM 
 
 
 bb 
 
 3 
 
 O 
 
 CO 
 
 Tf cm co a iO -cf — 
 
 ^ ^ bb 
 
 CM 00^ 
 CM 
 
 
 3 
 
 CO 
 
 o 
 
 NWWCCNCNO 
 rP CM CO ct: O •* CM 
 
 
 
 CM 
 
 £ 
 
 bb 
 
 is 
 
 ■5 
 
 iO 
 
 ia ro 
 
 ^»* cm ro a» o ^ cq 
 
 ^== bi 
 
 CMCCg 
 
 
 d 
 co 
 
 □0 
 
 »OCOCO^-0~:0 — 
 
 «»-: bb 
 
 
 bb 
 3 
 
 GO 
 
 CO C3 
 CO CM ^OOtOH 
 
 =°«?bi 
 
 
 *3 
 
 P-S 
 
 CO 
 
 CO --* 
 cOCMCMt^TfOCCCO 
 
 ■* CM CM cn OCO ^ 
 
 
 
 hi 
 
 > 
 
 eo 
 
 X. 7- 
 
 CO CM O 5 O CM — ' 
 
 «.S| 
 
 o 
 
 
 d 
 
 CO 
 
 O 
 
 ic as 
 
 «co«Voo^o 
 
 CM CM ^dCTf CM 
 
 
 
 bb 
 
 3 
 
 -< 
 
 CO 
 
 IQOC 
 
 eoco^'coooSo 
 
 cm cm l;o-'J- 
 
 ^^ bb 
 
 W 2S 
 
 CM 
 
 
 "3 
 
 CO 
 CO 
 
 lO CM 
 
 'fN'^OOOO'^o 
 
 ".ei bb 
 
 "2S 
 
 
 cm 
 
 J2 
 
 c 
 
 ca 
 
 EH 
 
 sc 
 
 o 
 _c 
 
 > 
 
 CM 
 
 coo 
 
 IQCQOO OS 
 OS CM t- 
 
 
 1 
 E-< 
 o" 
 *3 
 
 CQ 
 CO 
 
 
 d 
 cp 
 co 
 
 2 
 
 as o 
 
 t^CO O O co 
 
 .as CM CO 
 
 
 
 3 
 
 ' 2 
 
 00^- 
 CM CO" © — t^ 
 
 GO CM CO 
 
 
 
 *3 
 1-5 
 
 ** 
 
 oo 
 
 N«CC co 
 
 O CM CO 
 
 
 
 M 
 
 a 
 
 % 
 co 
 
 5 
 
 > 
 
 < 
 
 CO 
 CO 
 
 roddc t— 
 
 CM CM CO 
 
 
 
 d 
 
 CO 
 
 o 
 
 CM 
 
 o — 
 
 ^ CO* o «-* »« 
 CO CM O 
 
 
 
 bb 
 
 3 
 
 uO 
 
 CO 
 
 CM CM* O CO CO 
 
 OCN — 
 *H tM 
 
 
 
 "3 
 
 >-5 
 
 ^ 
 
 CO CM* O GO CO 
 
 Ocm t- 
 
 
 
 
 OS 
 
 a 
 
 1 s 
 
 o d 
 
 Eh 
 
 Vol. susp. solids 
 
 p.p.m. 
 Temp. C. 
 Diss. ox. p.p.m. 
 
 " " % sat. 
 Hel. stab. % 
 
 " " (paper filt.) 
 Ox. demand p.p.m. 
 " " (paper filt.) 
 
 Nitrogen as 
 
 nitrites 
 
 nitrates 
 Daily rate per acre. 
 
 CM ^- CM CM O -i-l ■ *-< 
 
 ^h ^ CM CM O 
 
 ^h ^^ CM CM O 
 
 CO — Ol HCOO 
 
 1-tTfOOiOONC 
 
 2^2S 
 
 ! O O CO -- i—«*=! 
 
 iO -^ CM >CCC« 
 
 — CM 0*3 
 
 COrH CO O 
 
 iflOSN^NOOMONNw 
 X'-Ofl 'T- — C: CO — i ^-»< 
 
 M-tNCCiONO^ 
 
 
 •tCrtrf NWC--iflC 
 
 3^H -^-^ 
 
 OOrtMOOOXON- 
 
 2^ = S 
 
 IMCCCCCKONO 
 
 2S 
 
 ic^ro co O o co cm 
 
 '2S 
 
 C^- IO CM GO 5J 
 
 .cgc-g 
 
 iO ^ CM CO O O CO 
 
 ft t-.CMCO*3 
 
 'f W -t K N C O C *1 ' 
 
 'SS 
 
 »C CO CM CO O O CO 
 
 co q 
 
 OOOCDCNN^ 
 
 CO »0 CM CM 
 
 ClCJCCiC^ 1 - 
 
 CO ^« CM -H 
 
 ■CK1«CrHi. 
 
 iCLO'CfOSCOO'-^ 1 " 
 M CM W3 1^3 CO * 
 
 t- *C CM CO CO GC '-0 -M 
 
 CO iT5 
 
 CO t^ 05 O 65 
 
 5 a 
 
 1^ 
 
 S- 3 i 
 
 co p, _ tn 
 
 -g-3 | g^ ~„ 
 
Sewage Treatment Experiments 
 
 73 
 
 which had passed these screens, was applied to the beds of Group 
 A, beginning June 1, 1916, and this method of operation was con- 
 tinued to the middle of January, 1917, when, owing to the failure 
 of the pump, it was necessary to shut down until March 1, when 
 operation began, first on Imhoff tank effluent for a month, then 
 on screen effluent for the rest of the year. 
 
 The results obtained from the application of Imhoff tank 
 effluent, and from that of screened crude sewage, are given in 
 sufficient detail to enable a comparison. See tables 27, 35, 36. 
 
 Experiments With Imhoff 1 Tank Effluent on Trickling Filters 
 Analytical observations date from April 1, 1914. Previous to 
 this the niters were undergoing a period of tuning up, and their 
 mechanical operation was being studied. 
 
 When first put in service, all the niters were operated at the 
 rate of 2,000,000 gallons per acre daily, which was continued 
 during the time of preliminary trial. This rate was maintained 
 on the niters of Group A from April 1, 1914, to October 1, 1914. 
 The filters of Group B were operated at varying rates from two 
 to five million gallons per acre daily. 
 
 On October 1, 1914, filters Nos. 2 and 3 were set operating at 
 4,000,000 gallons per acre daily ; filters 1 and 4 remaining at 
 2,000,000. These rates were carried until the end of 1917. Fil- 
 ters 5 and 6 were operated from the spring of 1915 at the rate 
 of 4,000,000 gallons per acre for the remainder of the experi- 
 ments. 
 
 FINAL SETTLING TANKS (humus) 
 PLAN 
 
 ?-A 
 
 rtrr 
 
 SECTION A-A 
 
 Fiff. 24. 
 
74 
 
 Brooklyn, N. Y. 
 
 Final settling tanks proved to be essential in the treatment 
 of the effluent from the trickling filters. The effluent carried 
 very considerable quantities of floculent, but very readily set- 
 tleable materials in suspension, also the remains of animal life 
 from the beds, in the form of dead worms in masses, as well as 
 living ones. The amount of suspensa carried by the filter effluent 
 were at times surprising, when for any reason an unloading of 
 films derived from the medium appeared. The sludge settled out 
 in the settling tanks at times made all the difference between a 
 putrescible and a non-putrescible effluent from the plant. This 
 sludge was highly putrescible and very difficult to dry. It should 
 be returned as a rule to primary tanks for full treatment. 
 
 The quantity of settlings removed by these tanks seemed to 
 be about the same whether the sewage had passed an Imhoff 
 tank or a fine screen before filtration, and to depend mostly if 
 not altogether on filter conditions and phenomena. 
 
 ABC 
 
 Fig. 25. Trickling Filter Effluent. 
 
 The bottles are about 16 inches high and are stood against a sheet 
 of white cardboard supporting a line of black. 
 
 A — Crude Sewage. Settleable solids about 200 P. P. M. 
 
 B — Imhoff Tank Effluent. 80% removal of settleable solids. 
 
 C — Unsettled Effluent from 7 ft. 3 in. depth of filter. Relation 
 stability 99% to 100% average. 
 
 The settling tank following filter No. 3 removed in 1 hour's 
 retention, from October, 1914, to September, 1915, 62 per cent, 
 of the solids in suspension in the filter effluent. The settling 
 tank following filter No. 1, removed in 2 hours' retention 90 
 per cent of the solids, and the settling tank following filter No. 
 5 removed in 3 hours' retention 70 per cent, of the effluent sus- 
 pensa. With these tanks in operation bacterial removal approx- 
 imated 99 per cent, and the plant effluent was 100 per cent, stable. 
 
Sewage Treatment Experiments 75 
 
 Concluding Remarks 
 
 In concluding this account of his work, the writer feels as 
 tho after much labor he had accomplished but little. At most, 
 he has obtained some fragmentary information as to what tanks, 
 screens and filters will, and especially what they will not do; but 
 these are, after all, rather negative data. The knowledge derived 
 from one plant can only be of partial and unbalanced character. 
 Observations from many plants obtained in accordance with a 
 definite system of investigation employed in each instance, or 
 universally, would be of real value and would advance science to 
 a higher plane. Unfortunately, there is not that uniformity now 
 in practice which will give such a result. Much has been done to 
 advance this object, however, especially by such institutions as 
 the Massachusetts Institute of Technology, the American Public 
 Health Association, and other agencies of progress that might 
 be mentioned. The work of some of our consulting engineers 
 has also been very fruitful along these lines. The books of 
 Messrs. Metcalf and Eddy, and of George W. Fuller have accom- 
 plished much; but more is needed, especially in the way of co- 
 operation among engineers and sanitarians, and more quiet, studi- 
 ous discussion among these men. 
 
 Looking at these problems from the standpoint of a mere muni- 
 cipal engineer, the writer has never felt and does not now feel 
 satisfied with the present status of sewage disposal, — that neces- 
 sary evil that must be compromised with — under the present state 
 of our knowledge. As to tanks of all kinds, they are abominable 
 to many, and so also are screens of all kinds. "It cannot be that 
 these are the only possible methods of treating sewage" ! Nor 
 is the trickling filter a friend to all. As the pot calls the kettle 
 black, so the ancient House of Tanks and the aspiring House of 
 Screens belabor each other with bad names, out with the lot! 
 "A plague on both your houses !' Give us something new and 
 without fault. Possibly the future will produce such a method 
 of sewage treatment that this age will seem the age of barbarism, 
 and the following ideal of such a plant shall close this paper: 
 It shall be a simple electrically operated machine, placed in a 
 fine large hall, with potted palms at convenient places for decora- 
 tion. The sewage will enter from below and will rise up thru 
 the plant, being discharged, after treatment, a pure pellucid 
 stream of water, cascading down to the nearby city reservoir. 
 Meanwhile, from one side of the plant, in neatly done up bundles, 
 will come forth automatically compressed fertilizer, containing 
 nitrogen units enough to pay all expenses. 
 
CONTENTS 
 
 Introduction 3 
 
 A Brief Review of the Work 5 
 
 Plan of the Plant 6 
 
 Mechanical Equipment 8 
 
 Laboratory Control and Sampling 9 
 
 Definitions of Terms 10 
 
 Local Conditions 12 
 
 Measurement of Sewage : 13 
 
 Calibrated Orifices - 14 
 
 Measurement of Compressed Air 15 
 
 Distribution Control 17 
 
 Obtaining Sewage for the Tests 18 
 
 Population and Daily Flow of Sewage 19 
 
 Per Capita Flow for Each Hour and Day, 20 
 
 Storm Water Flow, Character of 21 
 
 Sewage, Character of 22 
 
 Imhoff Tanks and Settling Tanks 25 
 
 Type of Slot Used in Imhoff Tanks 26 
 
 Slope Inclination, Imhoff Tanks 26 
 
 Period of Ripening 30 
 
 Foaming and Odors 30 
 
 Neglect of Sewerage System Cause of Odors 32 
 
 Dimensions of Tanks 33 
 
 Theoretic and Observed Retention 34 
 
 Effects of Baffles in Tanks 35 
 
 Character of Floating Scum 36 
 
 Bacterial Content of Sewage and Effluent 36 
 
 Settling Tank and Separate Digestion Tank 38 
 
 Capacity of Digestion Chamber Per Capita 43-45 
 
 Sludge Drying Bed, Area Per Capita 44 
 
 Sewage and Effluent Data 46 
 
 The Riensch-Wurl Screens 49 
 
 Requirements of Contract for Screens : 50 
 
 Results of Operation of Screens 51 
 
 Screened Sewage on Trickling Filter Beds 54 
 
 Cleaning Screens 56 
 
 Removal of Screenings 56 
 
 Analyses of Screenings 56 
 
 Weight of Screenings 56 
 
 Typical Operation of Screens, Storm, and Dry Weather 57 
 
 Sewage Treatment by Oxidation 60 
 
 The Trickling Filter Beds 61 
 
 Size and Character of Stone in -Filters 66 
 
 Determination of Area of Stone Surfaces 68 
 
 Summary of Trickling Filter Results 70-72 
 
 Final Settling Tanks , 74 
 
 Concluding Remarks 75 
 
TABLES 
 
 Table Page 
 
 I— Population of 26th Ward, Brooklyn, N. Y 19 
 
 II — Daily and Hourly Flow of Sewage 19 
 
 III — Per Capita Flow, for Every Hour, Day, and Week 20 
 
 IV — Storm Water Flow, Character of , 21 
 
 V — Sewage, General Characteristics - 22 
 
 VI — Sewage, Oxygen Relations 22 
 
 VII — Sewage, Nitrogen Content 23 
 
 VIII — Sewage, Cycle of Changes in Strength 23 
 
 IX — Sewage, Volume of Deposit and Time of Settling 25 
 
 X — Tank Data, Imhoff Tank Dimensions 33 
 
 XI — Tank Data, Settling, Digestion and Humus 33 
 
 XII — Tank Data, Theoretic and Observed Retention 35 
 
 XIII — -Floating Scum, Imhoff Tanks 36 
 
 XIV — Effect of Tank on Bacterial Count 37 
 
 XV — Effect of Tank on Bacterial Count Under Storm Conditions 38 
 
 XVI— Crude Sewage Supplied Plant 1914-15 46 
 
 XVII— Effluent, Imhoff Tank 1 46 
 
 XVIII— Effluent, Imhoff Tank 2 47 
 
 XIX— Effluent, Imhoff Tank 3 47 
 
 XX— Effluent Plain Settling Tank 3 48 
 
 XXI — Percentages of Removal Compared 48 
 
 XXII — Sludge Drying Data, Summary of 48 
 
 XXIII— Riensch-Wurl Screens, Official Test 51 
 
 XXIV — Screens and Tanks Compared 52 
 
 XXV — Screens and Tanks Compared. Tank Operating on Strong 
 
 and Screen on Weak Sewage 53 
 
 XXVI — Screens and Tanks Compared. Tank and Screen Operating 
 
 on Same Sewage 54 
 
 XXVII — Screen Effluent on Trickling Filter Beds 54 
 
 XXVIII — Typical Operation of Screen 5/64 Inch Apertures 58 
 
 XXIX- — Typical Operation of Screen 4/64 Inch Apertures 59 
 
 XXX — Typical Operation of Screen 3/64 Inch Apertures 59 
 
 XXXI — Typical Operation of Screen 2/64 Inch Apertures 60 
 
 XXXII — Trickling Filters, Properties of Trap Rock Used 67 
 
 XXXIII— Analysis of Stone Particles for Filters 68 
 
 XXXIV — Mean Weight of Particles and Ratio to Spheres 69 
 
 XXXV — Trickling Filters — Summary of Results April, May, June 70 
 
 XXXVI — Trickling Filters — Summary of Results July, Aug., Sept 72 
 
COLUMBIA UNIVERSITY LIBRARIES 
 
 This book is due on the date indicated below, or at the 
 expiration of a definite period after the date of borrowing, as 
 provided by the library rules or by special arrangement with 
 the Librarian in charge. 
 
 DATE BORROWED 
 
 DATE DUE 
 
 DATE BORROWED 
 
 DATE DUE 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 
 
 
 «* <• --*- ij 
 
 
 n w* 
 
 
 * ^# 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 C28(946)MIOO 
 
 
 
 
Manufactured oy 
 ! 6AYLORD BROS. Ine 
 1 Syracuse, N. Y. 
 
 Sloclcton, Calif. 
 
: . ■ .■■:-■■■.;:■:■'■■: Mhi