Nine Years’ Operation of the Baltimore Sewage-Works BY THEODORE C. SCHAETZLE Principal Sanitary Chemist REPRINTED FROM THE ENGINEERING NEWS-RECORD, JULY 14 AND 21, 1921 AUGUST E. CHRISTHILF Highways Engineer MILTON J. RUARK Div. Engineer ot Sewers Printed by The Highways Department of the City of Baltimore. ■%a.n In Co NINE YEARS’ OPERATION OF THE BALTIMORE SEWAGE-WORKS BY THEODORE C. SCHAETZLE Principal Sanitary Chemist of Sewage-Works BALTIMORE, MD. W ITH the publication of the 1914 report of the Baltimore Sewerage Commission the last con¬ sistent record of the development of Balti¬ more's vast sewerage system ended. Throughout the years that this report was published some mention was made of the sewage-works, always in reference to its construction features—except for the 1911 re¬ port, which dealt extensively with the experi¬ ments upon the results of which the present system was adopted. Numerous articles on the construc¬ tion of the plant appeared in the engineering periodicals but little as to its operation, and some erroneous ideas regarding the Imhoff tanks complet¬ ed in 1916 have been circulated. These facts com- Imhoff tank I sornot beoh. t&fmhoff tanks. Mgttrhouse. , Outfalh^ \ t sev^r L Bar screens Vtushinc veafer mnk Revolving screens 2setflin^^ Pov^er "ontro! buildina basins \ house Sludqe’ _ .. y digestion fiushtnc^^^ 1 1 m lifer beols,3C 1 1 1 CIC n res Ij {Br •)ke^ S _ 1 1 2 v^oodsta^ pipes extending 2500ft info nver 'Sludge beds ^ <.'1, o' 200'400'600'800' Sludge beds i t i i \ / dudge Sludge beds —-> tllllllllllllllllll digestion tanks Sludge drying plant- >C3 FIG. 1. GENERAL PLAN OF THE BALTIMORE SEWAGE-WORKS. bined with numerous requests for information have led me to prepare the following resum^ of the op¬ eration of the Baltimore sewage-works from their 790427 2 NINE YEARS’ OPERATION OF THE first full year of use in 1912 to the close of 1920. To make the account more readily understandable an outline description of the plant will first be given. On April 7, 1904, the Maryland Legislature passed the sewerage enabling act which stated that the Sewerage Commission of Baltimore had no auth¬ ority to construct any sewerage system which would discharge sewage, as distinguished from storm water or ground drainage, into the Chesapeake Bay or any of its tributaries. Accordingly the commission ^‘resolved, that the effluent proposed to be discharged into the Chesapeake Bay or its tributaries in the system to be recommended by the engineers shall be of the highest practicable de¬ gree of purity.^' To meet these requirements the Back River Dis¬ posal Works were constructed—and put into oper¬ ation on Oct. 30, 1911. Additions have been made from time to time to care for the increasing sew¬ age flow and at the present writing further ex¬ tensions are under contract. Description of Works— The existing plant con¬ sists of preliminary vertical bar screens spaced 1 in. apart; a meter house with five 42 x 21-in. ven¬ turi meters; 3 preliminary plain sedimentation tanks, known as hydrolytic tanks; 28 radial-flow Imhoff tanks, 19 separate sludge digestion tanks and 7.8 acres of sludge drying beds; 4 cylindrical re¬ volving screens, trickling filter control house, 30 acres of trickling filters, 4 small pump-houses, each equipped with an electrically-driven centrifu¬ gal pump, a hydro-electric power house generating power from available head in the treated sewage, and a final effluent conduit discharging into Back River. The hydrolytic tanks are divided into two com¬ partments by a curtain wall, the first compartment being 103 ft. 11 in. long by 101 ft. 9 in. wide and 10 % ft. deep, while the larger or secondary com¬ partment is 313 ft. 9 in. long by 101 ft. 6 in. wide BALTIMORE SEWAGE WORKS 3 and 13^ ft. deep. Both compartments have sloping bottoms. The Imhoff tanks are circular. They were orig¬ inally of the downward and outward radial-flow type, but some have since been converted to a radial horizontal flow. All of them are to be con¬ verted in this manner. These tanks are 39 ft. inside diameter, 13^ ft. deep to the bottom of the sedi¬ mentation chamber, and 27 ft. total depth. The sludge chamber is 23 ft. in diameter, with a conical bottom 5.3 ft. deep. The capacity of the sedimentation chamber is 8,920 cu ft., and of the sludge chamber 3,663 cubic feet. Of the 19 separate sludge digestion tanks three are 140 x 103 ft. 2 in. in plan and 13 ft. in depth and 16 are cylindrical tanks, 38 ft. inside di¬ ameter, 15 ft deep to the conical bottom, which is 9^ ft. deep, making a total depth of 24% ft. Each revolving screen, open at one end only, is 12 ft. in diameter and 10 ft. long, weighs approxi¬ mately 1 % tons and is constructed of well seasoned white oak. The stiffness is provided by ^-in. steel tierods. The surface is covered by Monel metal cloth, 24 meshes to the inch, reinforced with No. 10 gage galvanized wire. Each screen revolves on a bearing at the closed end and on trunnions at the open end, and is driven by a 2-hp. motor. The trickling filters cover an area of 30 acres, divided into 10 beds. They are filled to an average depth of 8% ft. with 1- to 2%-in. hard broken stone. The stones rest upon slotted tile which cov¬ ers the numerous underdrains leading to a larger underdrain and finally to the settling basins. These beds are provided principally with the Merritt square nozzles, although a few Taylor nozzles are still in use. The flow onto these beds is regulated from the control house, which contains a constant-head chamber, valves electrically operated for turning on and off an entire three-acre bed, and butterfly 4 NINE YEARS’ OPERATION OF THE Talves for causing a fluctuation in the spray, which has a cycle of approximately four minutes. The settling basins are 275 x 290 ft. in plan and 10 ft. deep, with sloping bottoms. The power house is equipped with two 150-hp. turbines. A direct connection to two 110-kw. generators is established by horizontal shafts pass¬ ing through stuffing boxes in the wall of the build¬ ing. The generators supply 2,300 volts to be stepped down to 220 and 110 volts. The power house is also equipped with a large switchboard, one two- stage centrifugal pump of 600 gal. per min. ca¬ pacity against a 157-ft. head, and a gas engine which was used for driving one of the generators before the sewage flow was sufficient for operating the turbines continuously. The sand beds are of three distinct types. The orig¬ inal bed consisted of 30 in. of sand upon 6 in. of gravel, but because of poor drainage the sand depth has been decreased from time to time until at pres¬ ent it is 18 in. and drains satisfactorily. The settling- basin sand beds have a 12-in. depth of sand with no gravel except over the underdrains. The beds built in 1914 aggregate 5.17 acres and are constructed with 4 in. of sand on an 11-in. layer of gravel. Normal Operation Methods —Baltimore has sep¬ arate sewers connected to the sewage-works by a single large outfall sewer 30,370 ft. long, 12 ft. 3 in. wide and 11 ft. high. The sewage first passes through the bar screens, depositing an average of 35 wheelbarrow loads of rags, fruit peelings, sticks, etc., per day or 30 lb. of dry screenings per 1,000- 000 gal. This material is removed by hand and sold to neighboring farmers. The liquid then con¬ tinues on through the venturi meters, where it is divided, part gomg to the hydrolytic and part to the Imhoff tanks. The hydrolytic tanks are kept in operation until they show signs of undesirable septic action or until the laboratory analyses show the settling efficiency to be below normal. The su¬ pernatant liquid is pumped to one of the other BALTIMORE SEWAGE WORKS § hydrolytic tanks and the residual sludge is pumped into any one of the 19 separate sludge tanks. The Imhoff tanks, on the other hand, are run continuously except for a short interval every two weeks when about 600 cu.ft. of sludge is run out of each tank producing good sludge. The effluents of the hydrolytic and Imhoff tanks dis¬ charge into the same channel and from there pass through the revolving screens. Here the finer particles which adhere to the under side of the screens are washed off by a continuous spray of pur¬ ified effluent which has been pumped from the hy¬ dro-electric power house to a tower 128 ft. high. The return water from these screens is then run through two of the Imhoff tanks used temporarily for this purpose. A special tank to care for this re¬ turn water will be constructed. The sewage then passes on through the control house to the trickling filter beds and from there through the final settling basins to the power house, thence through the land conduit, and is finally discharged into the river 2,500 ft. from the shore line. The sludge from the hydrolytic tanks having been pumped into digestion tanks is removed from these when in the proper condition, spread upon the sludge drying beds to a depth of 12 in. and, when dry enough, is removed by cars to the Heineken Reduction Co.^s drying plant, where the material is dried still more, bagged and shipped for fer¬ tilizing purposes. The Imhoff tank sludge is re- m.oved from the drying beds by farmers who pur¬ chase it for 25c per load. There being two settling basins, one is cleaned while the other operates, the cleaning time being determined under conditions similar to those for the preliminary sedimentation tanks. These tanks are not normally cleaned as frequently as the pri¬ mary tanks. When neither is cleaning both are in operation. The sludge from these tanks is either disposed of directly upon a sand bed provided for this purpose or, when not in the proper condition, 6 NINE YEARS’ OPERATION OF THE it is pumped to any available sludge digestion tank. OPERATION RESULTS FOR NINE FULL YEARS Actual Operation 1911 to 1920 Inclusive —Al¬ though the above outline is the method of opera¬ tion to which it is well to adhere, actual conditions have not allowed this procedure at all times. There¬ fore an attempt will be made to show just what the more important changes have been and at the same time figures will be presented to show the results of such operation. On Nov. 6, 1911, the first quantity of sewage recorded as having reached the plant was 2,411,000 gal., the same being a seven days' flow. Prom this time on until February, 1912, no consistent records were kept but after the latter date both operation and laboratory analyses records were made regu¬ larly. During these early days the plain sedimentation tanks were used as a dosing chamber, the trickling filters being operated but once a day and then only for a short interval. As more laterals were con¬ nected the fill-and-draw method of operation was abandoned and the continuous flow for one bed adopted. The 1912 records show an average daily flow of 11,800,000 gal., whereas the 1920 flow was 52,120,000 gal. per day. Owing to the fact that the outfall sewer had been designed to care for a 1,000,000 people at 150 gal. per day the first few years of operation with a small flow saw the deposition of considerable solid matter on the bottom of the outfall sewer, as well as a distinctly septic sewage to be handled. The former difficulty could only be overcome by cleaning the lower end of the outfall until such a time as the flow should become large enough to produce a scouring velocity. The cleanings were made about twice yearly, the last in 1916. At the present time, with a normal peak load of 69,970, 000 gal., the deposit from the outlet end to a point two miles up stream varies from 4 to 28 in. BALTIMORE SEWAGE WORKS 7 The cleaning was accomplished in the following manner: An adjustable dam, in cross-section the same shape as the lower part of the sewer but 1 ft. less in diameter, was hinged to the rear of a scow {See Engmeering News, July 22, 1915, p. 178). To take the scow up stream a rope was attached to it and a man in a rowboat carried the free end to the first manhole, where he tied it to the iron steps. Then the men in the scow pulled themselves up to this point. The same operation was repeated from manhole to manhole, the dam floating in a horizon¬ tal position. When ready to start cleaning the dam was placed in a vertical position and braced with guy ropes which could be quickly released if neces¬ sary. The rapid flow of water beneath the scow and dam scoured the deposit and as the water backed up the dam would be suddenly released, thus carrying the sediment along. This procedure was not abandoned because the velocity had be¬ come sufficient to scour, although there is consid¬ erable improvement in this matter, but was dis¬ continued because the volume of sewage is at all times of such an amount as to keep the flow line too high to allow sufficient head room for the scow and workmen. HYDROLYTIC AND IMHOFF TANKS Preliminary Sedimentation or Hydrolytic Tanks— During the early years of operation the small vol¬ ume of sewage received at the plant made it pos¬ sible to clean the hydrolytic tanks every ten days to three weeks in the summer, and six weeks in the winter. When cleaned this often very little water had to be added to cause the sludge to run to the pump well from which it was delivered to the separate digestion tanks. This condition held be¬ cause the scum had not had sufficient time to be¬ come thick and dry. At this time there was always one hydrolytic tank in service, one being cleaned and one in reserve. In June, 1916, the flow had become too great for one tank and a second one 8 NINE YEARS’ OPERATION OF THE had to be placed in service for the peak load, making two in parallel. Although the Imhoff tanks began operating in August, 1915, it was neces¬ sary to continue the operation of two hydrolytic tanks, and eventually to run the two continuously in parallel. The reason for this will be mentioned when the Imhoff tank operation is considered. A gradual accumulation of sludge has put to use all the available storage capacity provided by the sep¬ arate digestion tanks. This has resulted in less frequent cleaning of the hydrolytic tanks, which in turn has produced a heavy scum, pretty well dried out. This has of necessity increased the cost of cleaning because considerable water must be added to remove the scum and sludge to the digestion tanks. Soundings made from 1915 through 1920 show that the cubic yards of 90 per cent undigested sludge produced per million gallons of sewage treat¬ ed has been 5.70, 5.53, 7.34, 6.06, 5.14 and 6.37 for the respective years during this period. An average detention period of 3^ hr. produces the best efflu¬ ent, and approximately 65 per cent efficiency, based upon settling solids removed, can be maintained under normal conditions. Imhoff Tanks —The operation of the Imhoff tanks began in August, 1915, when 7 of ’ the 28 were put into service. Each tank was ‘‘seeded^’ differently to determine the best method of producing well-di¬ gested sludge in the minimum length of time. Among some of the methods used was the addition of newspapers, lime, digested sludge, etc., but no advantage was found over the usual method of allowing the sludge to accumulate and digest in the natural way. Other tanks were gradually put into service until by March, 1916, all but two, which were reserved for activated-sludge experi¬ ments, were operating. The tanks were operated at a low rate during the ripening period, which was rather long, but as soon BALTIMORE SEWAGE WORKS 9 as proper bacterial action seemed to have set in the rate was increased. Considerable trouble was soon experienced with foaming and excessive scum formation. Sludge was removed as frequently as conditions would warrant, there being times when nothing but water could be drawn off, the sludge having gone to scum. The sloping walls were squeegeed by a regular attendant, but in spite of all this the scum in the sedimentation chamber at times reached a thickness of 2 ft. and more. To prevent and quiet foaming a fire hose was used, but the best method was to shut the tank off from a few days to two or three weeks, during which interval skimming of the sedimentation chamber and spraying of the gas stack were usually resorted to. The tanks were then restarted very slowly. Although these tanks were designed to treat 500,000 gallons each, per day, with a 50 per cent overload, the average maintained for about two years for 26 tanks was 4,800,000 gal. per day. That is to say, no more than this amount could be treated with a proper settling efficiency and with¬ out considerable scum escaping. At this rate, how¬ ever, the average efficiency maintained was 61.6 per cent. 'MHOFF TANKS TAKEN OUT OF SERVICE On June 6, 1917, the Imhoff system was taken out of service because, at this time, the volume of flow had become too great for one sedimentation tank and the Imhoff tanks with their limited capacity, and not enough for two hydrolytic tanks and the Imhoffs. That is, if the latter combination were used the detention period in the hydrolytic tanks would become so long that odors would be materially increased. Another reason for discontinuing the Imhoff tanks was one purely local and strictly of a hydraulic nature. Because of the excessive loss of head through the Imhoff system the water level in the large effluent channel leading to the revolving 10 NINE YEARS’ OPERATION OF THE screens had to be carried a foot lower than when the hydrolytic tanks operated by themselves, there¬ by reducing the available area of the revolving screens very considerably. Furthermore, as soon as the head in the outfall sewer dropped below a certain elevation the Imhoff system would stop op¬ erating and then as the flow increased the water would seek its level through the hydrolytic tanks because of the less loss of head through them. In consequence of this the sewage would enter the Imhoff tanks backward, thereby disturbing the scum, pieces of which would leave the tanks for an hour or more after they had started operating. This condition also tended to overwork the revolv¬ ing screens. When the Imhoff system was discontinued the average daily rate of flow was 44,380,000 gal., but by 1919 this had increased to 55,060,000 gal. This made it possible to restart the Imhoff system by slightly increasing the detention period in the hy¬ drolytic tanks. With this end in view two tanks were remodeled and put in operation in May, 1919, for experimental purposes. The results of a long period of experiment warranted a change being made in all the tanks. Accordingly this work was carried on as rapidly as possible and by January, 1921, 18 of the 28 tanks were in service. The al¬ terations consist of a raise in the flow line so that there can be no back flow into the tanks, reducing the number of weirs in the outlet channel of each tank from 52 to 35, reducing their depth to 1 in. and closing permanently all but one of the ports in the bottom of the influent ring around the gas stack. This one port is closed with a wooden plug to which is attached a long handle so that the same can be withdrawn once or twice a day to flush the sediment out of the influent channel. The changes as made cause the tanks to operate with a radial horizontal flow instead of a radial downward and outward flow. The tanks now in operation give results prac- 11 BALTIMORE SEWAGE WORKS tically as good as the two run experimentally, showing an average efficiency of 72.8 per cent, with a rate of 365,000 gal. per tank per day during 1920, as against 61.6 per cent, with a rate of 285,- 000 gal. under the designed method of operation. FIG. 2. DAILY RATES OP FLOW AND UNIT COST OP TREATMENT, BALTIMORE SEWAGE-WORKS Although they are not up to the designed ca¬ pacity of 500,000 gal. per tank per day, it is the writer’s belief that approximately this quantity could be treated if the hydraulic features were such that the tanks could operate the full 24 hrs. instead of 20. It is interesting to note that with the revised tanks little scum has formed in the sedimentation cham¬ ber except for a short interval when the scum in the gas stack became so thick that it apparently short circuited and came up through the slot at the base of the flare wall. By removing some scum 12 NINE YEARS’ OPERATION OF THE from the gas stack, thereby raising the bottom of the scum layer above the slot, this trouble was overcome. Very little trouble was caused by foam¬ ing, though when it did occur, it was easily con¬ trolled by the use of a hose for a few moments. Revolving Screens —In 1912 two large revolving screens constructed of cast iron and steel, weighing approximately ten tons each and costing $9,645, were put in service. Because of their excessive weight they wore out the trunnions in ten days or less, whereas repairing the damage took about two weeks per screen. Thus they gave actual ser¬ vice less than half the time besides being very costly. After fair trial they were abandoned and the wooden screens designed by J. J. Neal, form¬ erly superintendent of the plant, superseded them. As stated before, they were constructed prin¬ cipally of oak, were built with the plant force and weigh about 1^ tons each. There are four of these, the first of which was installed in January, 1914. The others have been built as needed and in consequence have cost different sums which may be stated as approximately $1,100, $1,500, $1,800 and $2,500, or a total of $6,900. They have given excellent satisfaction, no one screen having been out of service, on an average, more than two weeks per year because of mechanical difficulties. The Monel metal cloth is renewed from time to time and the screens are rewound with reinforcing wire. All told, each screen operates about 90 per cent of the time. The only important change in construc¬ tion is the substitution of a 24-mesh screen for the 26-mesh since the latter has given trouble by clog¬ ging. An idea of the efficiency of the screens is obtained by noting that when in service from 5 to 15 pei cent of the sprinkling nozzles are cleaned daily as against 37 to 141 per cent when not running. As a result two men working eight hours daily are able to keep 30 acres of sprinkling filters looking well practically all the time. Furthermore, a series BALTIMORE SEWAGE WORKS 12 ©f analyses made in 1919 show that the return water from these screens contains 186 p. p. m. of susp>ended solids, consisting principally of sur¬ face floating material, such as grease, lumps of scum, etc. Approximately 100,000 gal. of purified sew¬ age per screen is used daily for washing purposes. TREATED EFFLUENT USED TO WASH SCREENS The purified effluent used as a spray on the re¬ volving screens is conveyed to them by means of 2^/^-in. pipes placed horizontally over the upper¬ most part of the screens. As the screens revolve with the suspended matter adhering to the under side of the wire mesh the spray of purified effluent washes this material off into a trough beneath the spray, which returns it to the Imhoff tanks pre¬ viously referred to. The pipes were formerly equipped with a number of nipples having 3-16-in. orifices and acted as a reciprocating arm by means of a cam attachment. The orifices clogged readily and were hard to clean and the reciprocating arm gave mechanical troubles. Because of these facts the arrangement described was replaced by sta¬ tionary pipes with an 0.008-in. width slot cut on the under side. Any clogging is now easily removed by sliding back and forth through the pipe and on the under side of the slot a small V-shaped piece of metal attached to the end of a long pole. The ma¬ terial broken loose in this manner is then removed from the pipe by flushing. The Trickling Filter Control House —This has op¬ erated continuously, as designed, except for one or two changes. The automatic float switches for operating the filter beds have been abandoned because the operator was not always present to record the changes as they took place. Now the switches are operated by hand whenever the water level in the channel leading to the control house rises above or falls below a given elevation. This arrangement also allows a better operation of the 14 NINE YEARS’ OPERATION OF THE Imhoff and hydrolytic tanks in parallel. The loss of head through the operating gates leading to the control house constant-head chamber proved to be so great that one of these gates had to be opened permanently. The butterfly valves, driven by friction disks, have given excellent satisfaction although the blades have been broken or washed away occasionally, which has resulted in minor, changes of reinforcement. The Trickling Filters —In 1912 the Alter beds operated 30.69 acre-days per week as against 125.08 acre-days per week at present. As additional beds were required a new bed was ripened by using it for a short interval each day at the start, then gradually increasing the daily period until, after about six weeks, it would be found normal. After being ripened the beds would be so rotated that the new one ordinarily would operate at least one hour daily and was never idle more than three days at a time. They have sluffed regularly and apparently to such an extent that from all indications it will not be necessary to remove the filter stone for cleaning at the end of ten years, as predicted by the advisory engineers, and it is doubtful that they will ever need to be removed for cleaning purposes. The pipe line installed in the galleries, carrying purified sewage for flushing the underdrains, has not been used for this pur¬ pose. However, the lateral distributing pipes have to be flushed occasionally. This was originally done by opening the valves in the galleries at the end of each distributor and allowing the sediment to drain into a wooden trough which emptied into the filter channel leading to the settling basins. This method was not entirely satisfactory because of the splashing over the trough and into the small underdrains. Now the laterals are flushed by removing the nozzle at the end of each distribut¬ ing line, leaving it off during two cycles of the butterfly valve and then replacing the nozzle. Thus all sewage passes through the filter beds. Occasion- BALTIMORE SEWAGE WORKS 15 ally it is necessary to flush out this section of each bed with a hose. The laterals on each bed are flushed about once every ten days. FILTER OPERATION The cleaning of nozzles referred to when speaking of the revolving screens is accomplished by remov¬ ing the nozzles and knocking the sediment off while the beds are running at the cleaning point, or at ap¬ proximately one-third the quantity of regular ser¬ vice. At times of excessive flow, the butterfly valves are stopped entirely, so that the flow through the beds is constant and at its maximum rate, this being 1.39 times the normal rate. Although the advisory engineers recommended a rate of 2,500,000 gal. per acre per day the rate has varied considerably at times, but the yearly averages shown in Table I have been maintained with generally good nitrification. TABLE I. AVERAGE YEARLY OPERATION DATA BALTIMORE TRICKLING FILTERS. Rate of Filtration Nitrates Relative Stability Year M. G. per Acre Daily P. P. M. Number 1912 2.73 8.3 93 1913 2.64 8.2 91 1914 2.72 7.2 78 1915 2.88 6.2 87 1916 2.84 4.4 93 1917 2.96 7.4 92 1918 3.00 8.1 96 1919 2.79 7.7 89 1920 2.91 7.0 92 In spite of the regular cleaning of the nozzles thin scale forms on the under side of the deflector plate, altering the radius of the spray to such an extent that the scale must be removed. This has been accomplished in two ways, one of which can be done in warm or windy weather only. That is, the bed is allowed to be idle for about two days, when this scale becomes thoroughly dried. The flow is then turned on and the scale flushes off. This method is objectionable because of the strong odor arising from the bed treated in this manner. The other method is efficient but laborious. The nozzle is bciled in lye water, scrubbed with a steel brush and then replaced. 16 NINE YEARS’OPERATION OP THE Two alterations have been made in the nozzles proper. The threads of the cast-iron nozzle plates had gradually deteriorated until the nozzle would no longer hold in place. Instead of using new plates a threaded brass bush¬ ing has been placed in the plates needing this re¬ vision and their life seems to be longer than under the original condition. Approximately 4,000 of the 5,830 nozzle plates have been altered in this manner. The other change has been to cut the threads off of the deflector plate stem as they became worn and loose and to rivet the stem into the nozzle base. TWO BASINS RUN IN PARALLEL The Final Settling Basins —These basins, of which there are two, were recommended for a 3-hr. deten¬ tion period, whereas actually the average for all the years of operation has been 6.04 hrs. Originally it was possible to operate one basin while the other was being cleaned or held in reserve. As the flow grew larger it became necessary to run both basins in parallel. Then the loss of head through the mix¬ ing chamber and inlet valves to the settling basins proved to be so great that it was necessary to raise the walls of the channel leading to the mixing cham¬ ber, to build a bypass around it (to the settling basins) and to make additional inlet openings into the basins. It might be said, in passing, that the mixing chamber, originally installed for the addition of hypochlorite of lime, has never been used except experimentally, and then only once. The Hydro-Electric Power House —This part of the plant has been operated as designed except for minor mechanical and electrical changes. When the plant first started it was necessary to use the gaso¬ line engine for generating electricity part of each day, but there has been sufficient sewage at all times since early in 1916 to allow the permanent closing down of this engine. BALTIMORE SEWAGE WORKS 17 SLUDGE DISPOSAL Here at Baltimore, as well as at practically all disposal works, sludge disposal is of primary im¬ portance. As first designed three square digestion tanks of 609,000 cu. ft. capacity were provided. In 1914 sixteen additional tanks of the circular type, with a volume of 330,000 cu. ft.,, were built. A con¬ siderable addition to the sand bed area was likewise provided. However, the sludge has continued to ac¬ cumulate at a faster rate than it has been disposed of, so that an addition to the digestion tanks is now under construction, while more beds will be provided in the near future. Although troublesome at times, our sludge oper¬ ation experience has been interesting. In the first years of plant operation there was plenty of storage capacity so that sufficient time could be allowed to produce well-digested sludge for disposal upon the sand beds. This sludge drained readily, dried and cracked well. During digestion the scum on the tanks formed from a few inches to approximately 2 ft. in thickness, but at no time prevented the nor¬ mal operation of these tanks, which was as follows: The undigested sludge was pumped from the hy¬ drolytic tanks into the digestion tanks, being de¬ livered at one side and just beneath the surface. The digestion tank receiving the sludge would then lie undisturbed over night, during which period the water, which had been added to the undigested sludge during the pumping, would separate from the sludge and rise to a place just beneath the scum. From here the water was drawn off by means of a channel near the top of these digestion tanks and was disposed of in the idle hydrolytic tank or the one which had been in service longest. This tank would then be ready to receive more undigested sludge or for the removal of some of the digested. The digested sludge was removed by opening one of the three well gates on the side opposite the incoming sludge. The sludge 18 NINE YEARS’ OPERATION OF THE was drawn from the sloping bottom through any one of the three drain channels to a collecting well and from there through a drain to the pump well, from which place it was delivered to the sand beds. It was noticed that certain parts of these tanks remained dead because there were only three draw¬ off points. To remedy this two of the tanks were emptied entirely and the drain channels covered with concrete slabs. Four lift gates were placed across the tank on each channel, making 12 in all for each tank. The idea was to raise by a chain that gate serving the part of the tank from which it was de¬ sired to draw the sludge. After the tanks were full and in operation unfortunately the gates could not be lifted by less than two men and, when once raised, it is doubtful whether they closed again. In 1914 the scum had started to form from 2 to 2^ ft. in thickness and a hose was played upon the surface to break it up. This was unsuccessful as the area was so large that the force of the water was spent by the time it reached the center of the tanks. Sludge was continuing to form at an in¬ creasing rate and weather conditions prevented rapid drying on the limited sand bed area, so that when the new circular tanks and sand beds were completed in 1915 they were immediately pressed into service. In spite of this the sludge continued to accumulate and by the later part of 1916 there was again need of more storage capacity. This seemed preferable to more sand bed area since the weather conditions regulate to a great extent the rate at which the beds can be cleaned and because the area then in use was about sufficient for the available laboring force even when the sludge was drying satisfactorily. Table II gives an idea of the sludge situation as it shows the total sewage flow for each year, the digested sludge produced and dis¬ posed of and the quantity in the tanks at the end of each year. The column of sludge disposed of gives BALTIMORE SEWAGE WORKS 19 the total, which includes that placed on sand beds, sold to farmers and placed on other areas. TABLE II. DIGESTED SLUDGE PRODUCTION AND DISPOSAL AT BALTIMORE SEWAGE-WORKS Cubic yards of 90 per cent Moisture Digested Sludge Sewage Flow Produced Pro¬ M. G. per M. G. duced Disposed In Tanks per (Based on per of per At End of Year Year Soundings) Year Year Year 1912 4,320.85 1.52 6,568 No record No record 1913 6,597.48 1.72 11,348 No record No record 1914 8,372.09 3.85 32,233 24,981 15,168 1915 11,607.41 4.05 47,010 49,856 12,322 1916 14.374.00 3.69 53,040 45,760 19,602 1917 16,199.83 3.15 51,029 29,616 41,015 1918 19,284.78 3.32 64,025 39,103 65,937 1919 20,096.08 2.98 59,886 65,617 60,206 1920 19,074.23 3.00 57,223 57,214 60,215 More sludge was produced during the years 1916 to 1918 than was disposed of, and in consequence the quantity remaining in the tanks at the end of each year grew larger. The ultimate result has been a gradual change in the moisture content of the sludge. As previously stated, at first well-digest¬ ed sludge was sufficiently watery to run freely from the tanks, while the scum formation varied from a few inches to approximately 2 ft. By 1914 the scum had reached 2 to 2 V 2 ft. in thickness but it was still possible to remove well-digested sludge without the addition of water. By 1917 the surface scum, iy 2 ft. thick, had dried to a moisture con¬ tent of 69.9 per cent., with the next 2% ft. below containing 75.9 per cent water and the remaining depth sufficiently watery to run. However, the best digested sludge seemed to be in these top 4 ft., so that in order to place the best digested sludge upon the sand beds a 2-inch hose working under a 128-ft. head was used to break the surface up and mix it with the more liquid sludge. This gradual loss of wa¬ ter continued until the worst conditions for the large square digestion tanks was reached in October 1918, whereas the circular tanks present their most diffi¬ cult operation conditions at the present time. In fact the scum in some of the circular tanks is as 20 NINE YEARS’ OPERATION OF THE much as 11 ft. thick. In October, 1918, soundings and analyses showed the moisture content of the ♦op 2 ft of the square digestion tanks to be 62.4 per cent, while the balance of the depth, or 11^ ft., except for small pockets of water, contained 78 per cent, of water. To relieve these conditions various available areas have been utilized by filling with sludge cut up with water. It was felt that more rapid progress could be made by removing the sludge without the ad¬ dition of water. With this end in view a paddle conveyor, electrically driven, was put in service in April, 1918. For part of the time this removed the heavy sludge to the area adjacent to the square digestion tanks from which place it was hauled away by the farmers, and for the balance of the time placed it directly into cars for hauling to the rotary drier plant. In 1919 more sludge was disposed of than pro¬ duced when the area between the filter beds and hydrolytic tanks was filled, thereby not only helping the sludge situation but improving the ap¬ pearance of the grounds. The effect of this was noted in the soundings and analyses made in the large digestion tanks in October, 1919, when the scum moisture had increased from 62.4 to 69.8 per cent and the sludge beneath had increased from 78 to 79.9 per cent. During 1920, when the same amount of sludge was disposed of as produced, the moisture content of the scum and of the sludge be¬ neath was further increased, so that on Oct. 1, 1920, the figures were 72.4 and 81.4 per cent respective¬ ly. Since early in 1918 one of the most annoying fea¬ tures of the digestion tank operation has been the appearance of undigested sludge at the outlet end of the tanks. This has been accounted for by the fact that the raw sludge, as pumped into the digestion tanks at one side, found its way under the scum or through the heavy sludge, much as water BALTIMORE SEWAGE WORKS 21 does through a clay bank. However, the well-di¬ gested scum, when broken up and mixed with un¬ digested sludge, has enabled the placing of at least partly digested material upon the sand beds at all times. For this purpose a 3^/^-in. fire-hose has been substituted for the former 2-in. one and i» proving very satisfactory. The operating force feels that when the excess sludge has been disposed of and sufficient storage capacity and sand beds have been provided for the current accumulation, so that the tanks can be kept with a scum not more than 2 ft. thick and the sludge with not less than 89 per cent water, the results will be very satisfactory, because past ex¬ perience has shown that sludge digests well in the separate tanks. One of the interesting features of the well-di¬ gested sludge is that it now contains 57 per cent volatile matter as compared to 44.2 per cent in 1914, when the tanks were operating normally and well. This change is probably due in part to the fact that in the early days of operation considerable mineral matter entered the sewers, due to the constructio* work then in progress. TABLE III. AVERAGE ANALYSES OF SLUDGE FROM DIFFERENT TANKS AT BALTIMORE Per Cent of Dry Residue Total Phos- —Wet Sludge— phoric Per Specific Ni- Acid Potash Cent Grav- Vola- tro- as as Digestion Water ity tile Fats gen PiOs K 2 O tank sludge. Imhoff . 91.86 1.021 66.21 4.02 2.45 0.52 O.IT tank sludge. . 92.38 1.017 62.74 • • • • 2.75 0.58 O.ly Raw sludge* Settling . 79.16 73.84 9.00 2.64 • • • • • • • • basin sludge . 92.37 57.98 • • • • 3.19 • • • • • • * • ♦Analysis as sludge exists in tanks before the addi- tion of water for cleaning. As the raw sludge is being pumped into the digestion tanks the moisture content is 91.56 per cent, and the specific gravity 1.020. The sludge as produced at this plant must be con¬ sidered under four headings; namely, raw, separate 22 NINE YEARS’ OPERATION OF THE digestion tank, Imhoff tank, and final settling basin sludge. An average analysis of these types is pre¬ sented in Table III. Sand Beds —The sludge is run onto the sand beds to a depth of 12 in. and when well digested cracks and dries rapidly in the summer, provided it is not an abnormally wet season. Ordinarily the beds drain well, but when one works to the con¬ trary either of two methods is used to relieve the water. A man walks through the bed stirring up the entire mass or a small pipe is inserted at one or more corners of the bed to drain the water to the adjacent area. Then when the bed is empty the sand is loosened by raking. In the years 1915 to 1920 inclusive 91,369 cu. yd. of dried sludge were taken from the beds and the beds were cleaned and refilled, on an average of S% times a year. Calculating from the average of 3.326 cu. yd. per 1,000,000 gal. of 90 per cent di¬ gested sludge produced and 30,615 cu. yd. of 90% sludge handled on the digestion-tanks sand beds yearly they have cared for the production of sludge from 9,205,000,000 gal. of sewage. At 100 gal. per capita daily, which is approximately our average, a yearly, average of 252,192 people have been served by 5.06 acres of sand bed area during this five year period. In other words, each person has required 0.874 sq. ft. of sand bed area per year. Table III shows that the sludge from the digestion tanks as placed on the sand beds contains an aver¬ age of 91.86 per cent water. The range of moist¬ ure, however, is from 86.6 to 96.4 per cent. That run from the Imhoff tanks ranges from 85.7 to 94.8 per cent moisture. When removed from the beds, the water content of the sludge ranges from 48.6 to 78.3 per cent, with an average of 68.7 per cent. In removing the sludge it is shoveled into cars and dumped in front of the Heineken Reduction Co’s drying plant. From the beginning of operation to April, 1916, this material was placed in piles for 23 BALTIMORE SEWAGE WORKS future use in leveling off the grounds or for sale to the farmers. But on Feb. 1, 1916, a five-year con¬ tract was made for all sand-bed dried sludge so that since then the sludge has been dumped on this company's storage pile. The cars were formerly filled level full, removed from the beds by hand and then drawn by a mule to the dumping ground. The first part of this year a 22-hp. gasoline engine was purchased for hauling. Now the cars are heaped and the engine hauls them directly from the beds, thereby giving a greater daily sludge removal. Owing to the adherence of sand or gravel to the under side of the sludge when the latter is being re¬ moved the 4-in. beds need a renewal of the sand layer. A small amount of sand has been added from time to time, but it is fair to say that a total of 3 in. has been lost by all beds since 1915. The adhesion of sand or gravel to the sludge is a matter of considerable importance at the Baltimore sewage works because the sand-bed dried sludge is sold under the above-mentioned five-year contract on the basis of its nitrogen content. If the sludge after passing through this company's rotary drier con¬ tains 2 per cent nitrogen as equivalent ammonia on the 10 per cent moisture basis, the city receives 81c. per ton, weighed with the water content rang¬ ing from 10 to 15 per cent. There have been times when the sludge has fallen below the required nitro¬ gen content, the reason for which can be readily seen when it is realized that the mineral matter, principally sand, has run as high as 39.6 per cent. To overcome this difficulty one of the beds was partially covered with a slotted wooden platform. This has been in service since 1916 and is still in good condition. It accomplished the purpose for which it was installed and the sludge disposed of upon it has acted very similar to that placed upon the sand direct. The present contractor for the sand-bed dried sludge does not seem disposed to renew his contract on the basis referred to above 24 NINE YEARS’ OPERATION OF THE because he claims to have lost money during the past five years. Therefore efforts are now being made to form a new contract. As removed from the beds the sludge has a specific gravity of 1.098 per cent, contains 70.67 per cent water, 54.83 per cent volatile matter, and 2.264 per cent nitrogen on the dry basis, whereas if analyzed at the time of removal, avoiding the sand, the figure.s are: Specific gravity, 1.077 per cent; water, 70.42; volatile matter, 61.76; nitrogen, 2.596 per cent on the dry basis. Although the drying plant is not a part of the sew¬ age works proper regular analyses have been made of the moisture and nitrogen content of the sludge after drying. In brief, the process of drying is to feed the wet sludge into that end of the rotary drier at which the furnace is located. The heated air and sludge pass through the drier together and dis¬ charge into a dust box. From here the sludge passes to an elevated revolving screen which removes match sticks, corks, etc., and from here down through a chute, at the end of which it is bagged. An average of all the analyses made upon this material from April, 1916, when the contract started, to Dec. 31, 1920, is, water 15.81 per cent and nitrogen on the dry basis 1.862 per cent. It is of interest to know whether there is any appreciable loss of nitrogen in passing through the drier. For this purpose a run was made in 1916 with the following result: Before passing through the drier 72.50 per cent water and 2.111 per cent nitrogen on dry basis; after passing through drier 18.89 per cent water and 2.075 per cent nitrogen on dry basis. Costs and Revenue— Excluding the cost of grounds, consulting engineers’ charges and a proportion of the chief engineer’s salary, the total construction cost of the Baltimore sewage works was $2,500,000, in round numbers. This does not include the cost of the alterations now being made to the Imhoff tanks. BALTIMORE SEWAGE WORKS 25 The actual expenditures for maintenance and operation, exclusive of interest on investment and depreciation, have shown a gradual decrease in the unit cost of treatment for each year from 1912 to 1917. In the latter year, due to increased costs of material and labor, this price rose. Table IV and Fig. 2 show the average and maximum daily rate of flow for each year as well as the cost per million gallons for treatment. No maximum daily rates are available previous to 1914 and it might be said that the absolute maximum treated has been 106,090,000 gallons. The Imhoff tanks began operating in August, 1915, and continued until June, 1917. They were idle, as previouly stated, until May, 1919, when two were restarted experimentally. These two have been oper¬ ating since and in addition 16 others were added during 1920. The comparative cost of treating sewage in the Imhoff and in the hydrolytic tanks with sludge di¬ gestion in separate tanks is shown by Table V. The large sum of $5,403 for the Imhoff tanks in 1919 is attributed to the fact that but two tanks were in operation and, being run experimentally, had more time applied to them per unit volume of sewage treated than would be the case under normal conditions. The additional increase for 1920 is caused by the increases given the laborers, to falling off in total flow and the necessity of disposing of a good proportion of the sludge both in the tanks and on the sand beds which had resulted from the period of operation in 1916 and 1917. The cost per cubic yard of sludge removed from the sand beds with an average haul of 1,000 ft. was $0,292 in 1915, $0,292 in 1916, $0,374 in 1917, $0,382 in 1918, $0,454 in 1919, and $0,442 in 1920. In . 1919, with the electrically-driven conveyor load¬ ing the cars at the digestion tanks and conveying the sludge direct to the dump, with a haul of 2,400 ft., the cost was $0,643 per cubic yard. 26 NINE YEARS’ OPERATION OP THE TABLE IV. MILLIONS OF GALLONS OF BALTI¬ MORE SEWAGE TREATED DAILY AND COST PER MILLION GALLONS. Year Average Daily Rate Average Maximum Daily Rate Cost of Treatment per Million Gallons 1912 11.80 $4,860 1913 18.10 4.800 1914 22.94 32.96 4.500 1915 31.80 42.90 3.450 1916 39.27 52.10 2.594 1917 44.38 59.15 2.882 1918 52.84 68.48 2.862 1919 55.06 71.83 3.389 1920 52.12^ 69.97 3.902^* ♦The falling off in flow is due to the decrease in water consumption brought about by a campaign for conservation because of the danger of a water short¬ age. ♦♦The increase in cost is due to the increase in labor costs and decrease in sewage flow. TABLE V. COST PER MILLION GALLONS OF TWO METHODS OF TANK TREATMENT AT BALTIMORE. Hydrolytic and Separate Year Imhoff Tanks Sludge Digestion Tanks 1916 $3,630 $2,439 1917 3.933 2.841 1918 Imhoff Tanks Not in Operation 1919 5.403 3.337 1920 6.897 3.738 TABLE VI. REVENUE FROM SEWAGE SLUDGE AT BALTIMORE From Drying Plant Year From Farmers (By Contract) Total 1914 $1,931.33 $1,931.33 1915 2,242.75 2,242.75 1916 1,098.15 $1,591.79 2,689.94 1917 346.40 28.07 374.47 1918 344.80 2,322.51 2,667.31 1919 309.00 1,557.61 1,866.61 1920 233.00 1,275.07 1,508.07 There has always been a small revenue from rentals of land and buildings, which will average ap¬ proximately $900 yearly. Since 1914 the bar screenings and sludge have been sold. The sludge has been sold both wet and dry; that is, in the same condition as when placed on the sand beds and as taken from storage piles after being removed from these beds. The price has been 25c. per load, except for a short interval, when it sold for 35c. This price was again reduced to 25c. per load after a falling off in sales proved that the farmers were unwilling to pay the 10c. increase. BALTIMORE SEWAGE WORKS 27 The wet sludge was removed in tank wagons holding approximately 250 gal., whereas the dried material has been sold by the load, one to two tons consti¬ tuting a load. In 1916 the sale of sludge to farmers was cur¬ tailed because the contract for sand-bed dried sludge began in the spring of this year. In 1917 it was still further curtailed by stopping the sale of wet sludge to those farmers who hauled over the main roads. Table VI shows revenue from the sale of sludge to the farmers and under contract. In 1917 the revenue dropped exceptionally low, due to the fact that the drying plant burned down and practically all of the sludge which was dried ran below the required 2 per cent equivalent ammonia content on account of the adhesion of sand and gravel in removing the sludge from the sand beds. Of the total output of this drying plant approxi¬ mately two-thirds has contained the required amount of nitrogen and has therefore been paid for. TABLE VII to XI. AVERAGE MONTHLY ANALY¬ SES OF BALTIMORE SEWAGE BEFORE, DUR¬ ING AND AFTER TREATMENT, JANUARY, 1912, TO DECEMBER, 1920, INCLUSIVE. VII. RAW SEWAGE Chemical Determinations in •rH m P o d m ^ B o d 5 • c u 0) u -M otal c Agar 'd o 380,000 41,500 850,000 47,000 550,000 37,500 900,000 66,000 1,400,000 190,000 1,800,000 180,000 2,200,000 270,000 2,600,000 290,000 4,100,000 650,000 2,100,000 250,000 1,200,000 220,000 • • 460,000 72,000 28 NINE YEARS’ OPERATION OF THE TABLE VIII. EFFLUENT FROM HYDROLYTIC TANKS January. ... 80 26 95 . . 210,000 35,000 February... 83 33 107 . . 190,000 29,000 March. 67 36 Ill . . 210,000 33,000 April. 69 31 106 . . 350,000 48,000 May. 59 44 120 750,000 100,000 June. 109 60 120 . . 1,100,000 73,000 July. 107 58 124 . . 1,300,000 110,000 August. 104 49 100 . . 1,000,000 150,000 September. . 63 37 109 . . 1,700,000 110,000 October. . . . 70 36 120 . . 2,000,000 100,000 November. . 69 34 116 600,000 87,000 December. . 65 35 97 390,000 45,000 TABLE IX. EFFLUENT FROM IMHOFF TANKS January. 31 105 320,000 61,000 February... 36 115 330,000 56,000 March. 26 134 150,000 51,000 April. 23 144 750,000 100,000 May. 25 146 1,100,000 210,000 June. 25 150 800,000 120.000 July. 23 August. 24 • ••••••• ••••••• September. . 57 • ••••••• • •••••• October.... 20 • ••••••• November. . 23 • • December.. . 41 • • • • r • • • ••••••• • • • • w «> TABLE X. EFFLUENT FROM FILTER BEDS January... . . 60 56 37 5.2 81 53,000 5,800 February. .. 73 46 40 5.4 82 140,000 9,900 March. 63 45 35 6.8 88 49,000 5,600 April. 93 67 34 8.0 94 140,000 9,300 May. 67 40 29 11.3 95 260,000 18,500 June. 96 52 32 8.7 92 700,000 23,000 July-... 65 52 36 8.1 91 440,000 31.000 August. 58 44 26 8.0 96 470,000 35,000 September. . 37 29 30 6.5 93 650,000 75,000 October. 48 32 30 6.7 98 270,000 28,000 November. . 59 44 36 7.6 96 140,000 17,000 December. . 96 54 40 6.1 83 88,000 14,000 TABLE XI. FINAL EFFLUENT January.. . . 51 23 28 5.4 85 59,000 6,700 February.. . 32 22 29 5.4 88 79,000 5,700 March. 33 20 28 6.8 93 59,000 5,000 April. 44 29 23 7.2 90 110,000 8,200 May. 53 33 22 8.9 93 600,000 14,000 June. 49 32 24 7.4 97 900,000 15,000 July. 39 32 21 7.4 94 1,300,000 32,000 August. 38 21 16 7.1 98 900,000 29,000 September. . 26 16 18 6.1 98 850,000 68,000 October. . . . 28 17 20 6.4 99 600,000 20,000 November. . 39 24 24 7.2 98 350,000 15,000 December.. . 40 29 29 5.9 86 200,000 15,000 Plant Performance —Practically no mention has been made thus far of the analytical results of the sewage at its different stages of transit through the plant. While the routine analyses do not in¬ clude total solids, chlorine, alkalinity and fats, an average for these is 773, 161, 153 and 81.7 p. p. m. respectively. The figure for total solids was ob- 29 BALTIMORE SEWAGE WORKS tained in 1916, at which time the suspended solids were 138 p. p. m. A 24-hr. run in February, 1920, gave an average of ,17.8 p. p. m. free ammonia. To acquire some knowledge of the performance of the plant during the various seasons the aver¬ age of all results for the different months from 1912 through 1920 inclusive have been compiled and are presented herewith in Tables VII to XI. The yearly averages are given in Tables XII to XV. Table XVI presents the plant efficiency by showing the per cent reduction in solids, bacteria, and de- oxygenating powers between the raw sewage and final effluent and by presenting the nitrates and relative stability of the final effluent, the raw sew¬ age being practically devoid of oxygen in any form. The final effluent of the plant during 1919 con¬ tained 2.50 p. p. m. dissolved oxygen and had an oxygen consuming value of 25.3 p. p. m. by the permanganate method with a 30-minute digestion period. During 1916 the total solids were 616 p. p. m. and the suspended solids 47 p. p. m. Samples of the river water are taken at fairly regular periods from April to December of each year. There are eleven stations, one approximately one-half mile up stream from the discharge pipe, another at the mouth of the discharge pipe, and the others down stream with a maximum distance of 9% miles below the outlet. The nearest oyster beds are 14 miles below the discharge. At all times since the plant operation began there has been a plentiful supply of oxygen at each station. This summary would not be complete without some mention being made of the two most im¬ portant experiments conducted at the disposal plant since its operations began, on activated sludge and grease recovery. Activated Sludge —Under an agreement between the Sewerage Commission of Baltimore and the U. S. 30 NINE YEAPwS’ OPERATION OF THE Public Health Service to work jointly on the ac¬ tivated sludge process a series of laboratory experi¬ ments was started early in March, 1915, and con¬ tinued until August of the same year. There were six experiments conducted simultaneously, five of which were with the air drawn through the bottle by means of an aspirator, while in the sixth the aerating was done by means of a propeller inclosed in a tube. The continuous operation of the latter experiment was broken up so often that reliable results were not obtained. The other five, run under different conditions, indicated that the ac¬ tivated-sludge process conducted under the fill- and-draw method was applicable to Baltimore sew¬ age; that it was possible to generate sludge more quickly with mixtures containing final settling basin sludge than with raw sewage alone or raw sewage plus raw sewage sludge, and that the absence of light had no noticeable effect upon the activated- sludge organisms. TABLES XII TO XV. AVERAGE YEARLY ANALY¬ SES OP BALTIMORE SEWAGE BEFORE, DUR¬ ING AND AFTER TREATMENT, 1912 TO 1920, INCLUSIVE. TABLE XII. RAW SEWAGE Chemical Determinations, P. P. M. Bacteria per C. C. w >> +-> O m d d -hS (rf d ci bfi d 'd