July, 1916
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERIXG CHEMISTRY
per cent, was attained. The low oxygen a t the start was probably due to the high content of iron of the water. The weight of carbon dioxide in the effluent air, calculated from the average content of carbon dioxide of the effluent air and the volume of air used, was 46 to j I g. per run. The weight of carbon dioxide originally in the water, calculated from the dissolved carbon dioxide and the capacity of the tank, was 43 to 56 g.; that is, the carbon dioxide content of the effluent agreed reasonably well with the content of dissolved carbon dioxide in the original tap water. The alkalinity in presence of phenolphthalein resulted from the breaking down of bicarbonates naturally in the water. According to the titrations in the experiments the amounts of carbon dioxide set free varied from j z to 64 g. per run. This gas, in accord with the law of Henry, was not liberated because the inflowing air contained carbon dioxide. When sewage was aerated in the presence of sludge the net gain of the effluent air in content of carbon dioxide was IZj g. If 46 to j~ g. are deducted for the carbon dioxide in the effluent air obtained by blowing air through tap water, the 7 j to 80 g. of carbon dioxide that remains may be considered as having been produced by bacterial action during the process of aerating sewage. That this deduction is warranted is shown by the fact that when fresh sewage was aerated without sludge the dissolved carbon dioxide disappeared and the phenolphthalein alkalinity appeared a t nearly the same stage of aeration that it did when tap water was aerated. Sewage allowed to stand after the aeration period loses its alkalinity to phenolphthalein. The sewage was not alkaline in presence of phenolphthalein 17 hrs. after aeration had ceased. When tap water was used, however, the alkalinity in presence of phenolphthalein remained several days, and the total alkalinity remained unchanged. The disappearance of the alkalinity to phenolphthalein of the raw sewage was evidently due to putrefaction of the sewage because of lack of air Strong bacterial action does not take place for the first few hours.
647
the beds were unable to treat adequately the sewage, some of which is I j hrs. old when it gets to the beds. A few years ago the area of the beds was increased by I O acres and three years ago a 0.j acre trickling filter was constructed. The results from the trickling filter have been disappointing; very little nitrification (only 4.55 parts of nitrates per million on an average in 1915) has been secured. The effluent was stable for only about two days. Furthermore, the effluent from the secondary or humus tank is always putrid and the trickling filter unloads not intermittently but continuously throughout the year. Considerable relief has been obtained by resting the filter for two hours daily, but even a t rates as low as j4j,ooo gal. per acre per day the results were unsatisfactory. During passage through the humus tank the nitrates in the trickler effluent were reduced from 4 . 5j t o 3.32 parts per million and the oxygen from 2 ,2 to I , 6 parts per million. This is due to vigorous decomposition of the sludge, the so-called humus, which a t Brockton has all the characteristics of a badly digested septic-tank sludge. When the trickler effluent was applied to sand beds a t a Igo,ooo-gal rate, free ammonia was decreased from 30.6 (which is high for a trickling-filter effluent) to 1 6 . 3 parts per million, and nitrite from 0,732 to 0.069; while, strange to say, nitrate was decreased in the intermittent sand filter from 4 . j j to 2,62 parts per million. The remedy for a sewage so stale and one which cries for oxygen so loudly is obviously more air. In 1915, aeration experiments were made with a cascade aerator 20 ft. long having a slope of I f t . in IO. The devices were so arranged that the sewage could be passed over this aerator ten times, but this was not enough to show the presence of the slightest trace of dissolved oxygen in the effluent. After each aeration the sewage was passed through a bed of cobbles 6 in. in diameter in order to make use of the contact action employed so successfully by Mr. Clark a t Lawrence in his aerated tank but all without sufficient benefit. The effluent from this aerated contact tank passed through an Imhoff tank with little if any change, and after the effluent from the Imhoff tank had been UNIVERSITY OF ILLINOIS, URBANA applied to a trickling filter the results were not enough better than those obtained by the municipal filter to warrant further continuance of the experiments. The colloidal matter was not SEWAGE DISPOSAL EXPERIMENTS AT BROCKTON, appreciably decreased in amount but was carried through the MASSACHUSETTS whole row of devices. By ROBERT SPCRRWESTOS Along with the other 1915 experiments a small activaConsulting Sanitary Engineer, Boston ted-sludge tank was operated on the fill-and-draw plan. This The city of Brockton, which has a present population of 62,000, tank was aerated for 4-hr. periods and contained Z j per cent of is located in southeastern Massachusetts, and is sewered on the sludge. The effluent was discharged on an intermittent sand separate system. The sewage drains to a low point in the city, filter 5 ft. thick a t a rate of 500,ooo gal. per acre per day. T h e where it collects in a subsiding basin. It is then fine-screened results were excellent and in striking contrast with the results and pumped to the disposal area about three miles away. obtained by the municipal sand beds receiving the trickling The disposal works consist of 37 acres of sand beds and a filter effluent. The effluent from the activated sludge tank contrickling filter having an area of acre. The average daily flow of sewage is 2,100,ooo gal. Approximately two-thirds tained 254 parts of suspended matter, 3 7 , 6 parts of free ammonia, of the screened sewage is applied directly to 30 acres of sand no nitrite, only a trace of nitrate, and 3.28 parts per million beds a t an average rate of 52,000 gal. per acre per day; the re- of dissolved oxygen. Nevertheless, when this effluent was apmainder is first applied to the o . j acre of trickling filter and then, plied to the experimental sand filter filled with sand from one of the beds which had been receiving trickling filter effluent, after passing through a secondary subsiding basin, to 7 acres of excellent results were obtained; i. e., free ammonia was decreased sand beds. The rate of application on the sand beds is about to 6.62 parts and nitrate was increased to 13.89 parts per mil130,000 gal. per acre per day. According to the last report of the Sewerage Commission lion on the average, while the effluent was brilliant in appearance and stable even a t the high rate of j o 0 , O o ~gal. per acre the sewage has the average composition shown in Table I: per day. The activated sludge experiments were so promising TABLEI-AVERAGE COMPOSITION(P. P. M.) OF SCREENED BROCKTON that they have been continued during 1916 on the continuous S E W A G E DURING 1914 Total Suspended Free ALBUMINOIDAMMONIA CHLORINE plan with a large aerating tank I O it. square and 8 ft. deep Solids Solids Ammonia T o t a l Dissolved Suspended below the flow line, discharging into a settling tank and from it 778 204 55.8 12.6 6.9 5.7 138.6 on a sand bed I O ft. square and 2.5 f t . thick that operates a t a half-million rate. The original plant, built in 1894, consisted of 27 acres of sand beds. These gave excellent results for many years after 4 s the object of these experiments was to devise a process their installation, but with increase in the amount of sewage t o relieve the present plant as much as possible, no attempt and addition of an increasing amount of shoe-factory waste, was made to obtain complete nitrification but only a clarified
T H E J O U R N A L O F I N D U S T R I A L I N 0 ENGIAVEERING C H E M I S T R Y
648
Yol. 8, S o . 7
effluent which contained some free dissolved oxygen and a sludge which would not putrefy in the tank. The experiments began with an aeration period of less than 2 hrs. and with varying volumes o€ air. This aeration period has been gradually increased until a period of 4 hrs. is now used. This seems t o be the one which gives the optimum results, when about 2 5 per cent of sludge i s used. \Then a shorter period is used the sludge becomes septic and loses its efficiency regardless of the amount of air used. Xl'hen the period is greatly lengthened, as shown beyond, not only is no nitrification produced-regardless of the amount of air-but the accumulated suspended matter passes again into t.he colloidal state and the effluent from the settling tank becomes highly colored. The sewage with which the experiments have been made had the composition given in Table 11, and the results obtained by the experiment, as shown by the analyses of weekly composite samples, are given in Table 111, in parts per million except as otherwise designated.
after the beginning of aeration the color of the supernatant liquid suddenly increased and the sludge began to settle imperfectly and to decrease in volume. The suspended matter evidently had been partly transformed into the colloidal state. Many spent dyes used in the Brockton shoe factories are discharged with the sewage and a t present it is believed that these are a t first absorbed by the activated sludge but that prolonged aeration causes them t o pass into colloidal solution. The Brocktoil sewage, if allomed to stand, develops a strong reddish color. %'hen further aeration will reprecipitate this color cannot be answered now. As Table IV shows, the free ammonia is beginning tp decrease, after having increased, and nitrite and nitrate are on the increase. Perhaps a slow oxidation is taking place which will end in complete nitrification and reclarification.' Whatever the result may be, nitrification b y aeration seems impracticable a t Brockton. The experiments a t Brockton are not yet finished, although nearly so. At present we are using a 4-hr. period of aeration TABLE 11-COMPOSITIOKOF SCREENED SEWAGE susand about 2 3 per cent of sludge and are applying the effluent Week ending FREE ALBUMIKOID NITROGENAS PEKDED (1916j CHLORIXE A4MMoN1.A AXMOSIA Nitrite Nitrate SOLIDS t o a sand bed a t the rate of 500,000 gal. per acre per day. 33.3 4 . . . . . . 108.5 8.60 0.04 0.19 196 Jan. We are now gradually reducing the air to ascertain the minimum 9.80 0.22 31.6 11. . . . . . . 103.5 0.22 152 18 . . . . . . . 9 2 . 0 8.80 0.24 31.6 0.45 166 required for a 4-hr. period of ,aeration. We feel sure that this 25 . . . . . . . 104.5 1 1 . 00 0.07 38.x 0.16 256 minimum, even for clarification alone, will be considerably Feb. 1 . . . . . . . 110.5 11.20 0.11 40.0 194 0.2; 12.60 0.0; 38.7 8 . , , . . . 119.5 0.20 208 more than the z cu. it. per minute per gallon of sewage or less 1 5 . . , . , 111.5 9.2 0.13 36.4 0.07 190 23 . . . . . . 112.0 12.6 0.09 37.5 0.07 222 determined in other experiments. The Brockton experiments 2i.9 10.2 22,.. . . 88,s 0.28 0.27 I80 have showi that the activated-sludge process is the only one Mar. / . . . . . . . 74.5 7 6 0.41 24.5 0.19 108 27.3 13 . . . . . . . 9 8 . 0 9.6 0.24 0.28 146 Tvhich offers any relief to the overworked filter beds, sick with 35 . :i 21 . . . . . 89.5 9.2 0.17 0.01 126 35.3 0.19 2 8 . . . . . . . 93.0 8.2 0.01 172 stale sewage. On the other hand, they have shown that it would 17.9 Apr. 4 . . . . . . 66.0 4.6 0.46 0.47 54 be impracticable to m e the process alone or to produce a highly 14 . . . . . . . 7 4 . 0 2.5.5 6 2 0.09 0.48 92 nitrified effluent thereby. Kever has complete nitrification T.UJLE III--COMPOSITSOK O S W E E K L Y COMPOSITE S A X P L E S O F ACTIVATED EFFLUENT been obtained or anything approaching it. The activated-sludge Week AIR process ~vculd,however, allow the beds t o be operated a t a rate ending PER :\IIK. .SIR SLUDGE FREE NITROGEK AS (1916) Cu. f t . Volumes Per cent AMMONIA Sitrite Nitrate exceeding zoo,ooo gal. per acre. There is no difficulty in ob-0.0001 3.244 Jan. 25 0.0155 taining a highly nitrified effluent from a bed t o lyhicli the effluent 0.0001 3.871 30 0.0190 0.0001 6 3.636 Feb. 0.0185 from the activated sludge process is applied; furthermore, the 3.529 0,0000 13 0.0000 3.750 0.0000 20 0 0000 beds so operated are free Erom subsurface clogging, which noT10,0029 27 0.OOTO 3.077 greatly interferes viith the operation of the municipal plant. 0,0110 Mar. 5 0.0100 1 . 876 0.0001 12 2.308 0,0090 That this plant needs some assistance is shown by the average 0,0004 2.727 0.0035 19 0,0000 76 3.428 0.0000 analyses in Table \- lrom one representative group of beds. ,
Apr.
.....
2 7
2.143
......
......
0.0260
0,0080
The most interesting feature of our experience har been the effect of plain aeration on the process. After the experiments had been in operation for two months and complete nitrification had riot been obtained even lvhen z z volume., of air per i-olume of sewage were used, it was decided to stop the flow of sewage through the activated-sludge tank and continue aeraABLE LE IV--
COLlPOSSTIOX
OF E F F L U E N T FROX P L A I N &RATION
SLCDGEDISSOI,VED FREE Per cent M a r . 24 . . . . . . . . . . . . 29 25 . . . . . . . . . . . . 26 26 . . . . . . . . . . . . 25 27 . . . . . . . . . . . . 22 28 . . . . . . . . . . . . I ? IliitE
29 . . . . . . . . . . . . 18
30 . . . . . . . . . . . . 31 . . . . . . . . . . . 1............ Apr. 2. . . . . . . . . . . . 3............
18 16
15 18
191 4 . . . . . . . . . . . . 24 5 . ......... 25 6 . . . . . . . . . . . . 27
:. . . . . . . . . . . . 25 a , . . . . . . . . . 21
191
l 9o ., ., .. .. .. .. . .. .. .. .. .. 19 11 . . . . . . . . . . .
12 .
16
. . . . . . . . . . . 18
. . , . . , I? 15 . . . . . . . . . . . 19 13..
,,, ,
OXYGEN A X M O N I A 0.00 30.91 0.00 31.88 0,oo 35.17 0 00 -14.34 6.40 '16.36 55.56 . . 70.34 6.7 80.00 85.00 72.86 .... 71.58 ,.., 70.34 . . . 66.91 .... 74.18 i5.56 .. .... , 71.58 . . ~ . 62.77 62.77 .... 57.46 .... 50.37 . l . n
~ - I T R O G G N AS
Sitrite
0,000 0.104
n 074
0.089 0.074 O.li0
0.190 0.195 0.330 0.520
0.570 0.530 0.630 0.560
0,560 0.530 0.590 0.670 0,680 0.960
Nitrate 0.00 0.09 0.06 0.06 0.05 0.06 0.07 0.10 0.10 0.08 0.12 0.13 0.15 0.16 0.17 0.19 0.17
....
....
....
tion until the nitrification was complete. This experiment began March 2 1 and is still in progress. It was conducted in the large tank until April I and then transferred t o the smaller tank. The results of plain aeration are given in Table IV. The clarification was good a t the beginning, but 18 days
TABLE~--AKALYSIS AhT.lnroNra
(P,
P.
11.) OF REPRESEKTATIVE a R O C K T O K FILTER-
BED
EFFLUZNT (1915)
f F r e e . . . . . . . . . . . 25.96 1Albuminoid. . . . 1.70
NITROCEX{
Nitrite. . . . . . 0,039 N i.t r 'a t e . . . . . . 0 . 8 7
The rate of filtration a t which this effluent was produced m-as only 52,000 gal. per acre per day, or about one-tenth of that used experimentally with the activated-sludge effluent with which a highly nitrified bed effluent was produced. Some of the sludge contains 99 per cent of mater and it is believed that tank treatment before applying it to the beds for drying, will be necessary. 14
3EA.cow
STREET, BosTOK
THE ACTIVATED-SLUDGE PROCESS IN TREATMENT OF TANNERY WASTES By HARRISOX P. EDDYAKD ALMOK I,. F.\zEs Of Metcalf &- Eddy, Consulting Engineers, Boston and Chicago GESERAL TREATMENT O F INDUSTRIAL XhSTES
The activated-sludge process of sewage treatment may be particularly well adapted for the treatment of some industrial wastes. The firm with which the writers are connected is now directing activated sludge tests on paper-mill wastes a t the mills of Bird & Son, East 'il'alpole, hlass., on woolen-mill wastes a t the Assabet hIills of the American IToolen Co., hlaynard, Mass., and on tannery wastes a t the factories of \Tinslow Bros. & Smith Co., Xorwood, Mass. The tests on woolen-mill wastes 1 O n June 15th t h e free ammonia had been reduced t o 1 P. p . M . and t h e nitrates had been increased t o more t h a n 1 P. p. &I. T h e sewage was still d a r k colored a n d would n o t clarify b y subsidence.
