GIFT OF I Fifteen Years Filtration Practice in Indianapolis H. E. JORDAN From Proceedings of the Thirteenth Annual Convention of the Indiana Sanitary and Water Supply Association General Statistics, Indianapolis Water Company, Page 65 ulrT FIFTEEN YEARS FILTRATION PRACTICE IN INDIANAPOLIS. H. E. JORDAN'. At the time that the experiments on the purification of the Ohio River water were being made at Louisville, Ky., the In- dianapolis Water Company called into consultation Mr. George W. Fuller in reference to the development of a filtration sys- tem. A brief report was made by him in which it was sug- gested that studies comparable to the Louisville investigation be carried on at Indianapolis. Allen Hazen made an extended report in July of 1896, recommending slow sand filtration. Between this time and 1902 practically every engineer promi- nent in the water purification field at that time made a more or less extended report on the purification of water from White River. The committee of the Board of Directors of the In- dianapolis Water Company in April, 1902, definitely approved suggestions looking toward the construction of a slow sand filtration plant and the United States Sand Filtration Com- pany was awarded the contract for the construction. It may be worth while in passing to suggest that the pe- riod of 1900-1902 was one when purification projects were also under way in a number of the large cities of the country, notably Philadelphia, Pittsburgh and Washington. The Washington filtration question was investigated by the Corps of Engineers of the United States Army under the direction of Colonel Miller, and while his recommendation was for the construction of a mechanical type plant, such opposition de- veloped in the community led by the Medical Association of the District of Columbia, 'that the committee of the United States Senate appointed to investigate the question overruled Colonel Miller's recommendations and Washington, like Al- bany, Philadelphia, Pittsburgh and Indianapolis built. a filtra- tion plant of the slow sand type. In all cases these plants have been modified upon the basis of experience following the actual operation of the system. In each case it was found that while the system was able to handle very adequately raw water of normal conditions, any condition of overload, espe- cially as referred to the amounts of suspended matter, was reflected in a corresponding variation in the quality of the finished product together with a very decided lessening of 437854 the out^nt'of fh"(^plaiik t i-I rt< ^ OO t --*CO* 2ii7 $1.933 2.236 .196 135 Total '-"(_ of total $ .889 !'. 7 % 1.867 41 5 % 1.537 % .207 4 6 . $4.500 The improvement in operating conditions as evidenced by the reduction in expense and increased output is the result of ability to produce a larger quantity of water per operating day and the handling of a smaller amount of sand per million gallons produced. The operation of the slow sand filter plant consists essentially of one item, that is, the maintenance of the sand layer in a condition that will produce the maximum amount of purified water per yard of material handled. The data as to filter unit operation is shown in Table I. llsl cl-"! 1 1 * ^ tf a> !l! 3 |1 S rage Number of Days per Run I Number of Filter Runs J0 000i i -^ i i CO CO < -ooo . I- C<5 OO -- SO -^ t~- O OiOOiOCO-<*-OOOt^ i I H iOOC?5CO -H cv T o >o ' rt O 1C 00 t~- CO CO OJ O 00 CO O COO OOOiOCOCOt^-Oi'^' ' O' OI>-l'~-H'-HCOt^CN|! lkCCOOTt* co co oi <* >ra >c co CM' CD co co fl >0 s? r ^rr^t>. PO < < ! o eo t^ * o >c cCCO"5>C5 no ^>> r ^ Tt ~. - TJIOO-* oce eo co t^. c< eo" oo * >o <*'*' so ec i c o us o UO i ^(M-*-^^ . riri t^oso :t re I -r - < Tf 1C 5 1< -- - id >! Tj*O5 ^-^-iC^- -^CO^O S , , - *' & '~r- C5 CO co CD oo - r-3000 -- O "O !S t^ U3 *< sssgssriH--- CO ^ g a . c r^ t^t^ -~ ' co co -*' * co -* o co *' o o od od co -^' us co - onth, 4,176; of time lost ^ S^ s :-2S s ^52o^ 00 )32; 29-day rr ccount made SHJLNO o o od od co od 01 1> *' od' i>i t^.' us t~-' o> *' od ay month, 4 run, nor is a S !* PQ i w 1 COCO COI>-1O^H-H O5 -* CO IO COCO ' oo" ci so' t^ t^ so' oo' ^4 co' o> *' iv 55 od CO r-l r-t T-( rt (Mrtrt co' urs are: 28-d lakes its first o" g 00 * 0 00 CO C5 T-I 00-H 5 5 J} 'e X> 3 55 i i-llCCOCOO5COl^iO'*C^COt>.COCOkOrt ^H 8 S ^ d [CE FOR C] 4 >c o >o od o od co us co os - 00 CM O> i-l ) Tf O4 r-t t ( 2 vice for clear al. No lost 1 re each. 4. 1 ,R WAS OU 1 00 ^ CO SO Oi 00 (MCi od s 1 i I III ERCENT. 3 T-ll>. CO O5 00 rt CCI CO IM US * t- 3 |1| | ^ ^ it-^H CO^OO .' b-' CO 00 t-- t^ 00 CO Tji CO so' O Ti )< "O iff! s ^USCO^OOO.O-CNCO^USCO^OO^ 2 >> bfi CiC305Sc30O5O5O5Ci05O505O5 -a -| ^2 l co T3 i:: tion of the sand dependent upon the relative specific gravity of the various grain components of the total sand layer. In 1914 a modification of the Nichols washing-in place method was adopted and still is used. The changes in sand handling cost may be summarized briefly as follows: The original method with labor costing from 121/2 to 15 cents an hour involved the expenditure of $1.25 per yard of material handled for removal, washing and replacing. The use of the stilling box instead of the wheeling- in reduced the cost to $1.00 per yard. Various improvements in the sand handling capacity of ejectors used and the wash- ing-in method of restoring reduced the cost until in 1911 and 1912 the total expenditure was 40.5 cents per cubic yard for scraping, ejecting, washing, replacing and smoothing sand. Adoption of the washing-in place method eliminated an addi- tional handling of sand outside the filter unit and with labor at 22 V^ cents an hour made the total cost of sand handling 25 cents a cubic yard in 1917. In 1919, with labor at 40 cents instead of 22 V, and decreased efficiency of the laborers, the cost increased to 55 cents, which is, however, as will be re- membered, lower than the conditions under which the plant operated originally. Pwemembering the fundamental proposition that sand handling is the key to the successful operation of the slow sand filtration plant, it becomes increasingly a matter of dis- pleasure to the writer to confess the relatively small mechan- ical improvements which have been made in this operation. It will be remembered that the Pittsburgh plant installed equipment built by the Blaisdell Manufacturing Co. for re- moving and restoring sand. Likewise Wilmington, Del., con- structed its filtration plant in such a way as to accommodate the Blaisdell washing-in place machine. Later developments under the direction of Nichols at Philadelphia have resulted in certain improvements in the method of removing soiled sand from the filter. It still remains necessary, however, under the present condition of labor shortage and inefficiency to attempt to increase in every way possible the mechanical methods of handling sand, and the chief thing to be desired in the operation of a slow sand plant is a piece of equipment relatively light and easily movable which will remove soiled sand, wash it and replace it in the sand layer. 14 PRELIMINARY COAGULATION. The application of the coagulant is not necessary at all times. The preliminary coagulation of White River water begins ordinarily when the turbidity is between 30 and 40 parts per million. The range of turbidity of the raw water throughout the life of the plant is such that -56.1% of the time the turbidity is less than 30. During the summer months, however, coagulant is used to assist in algae reduction. The false information obtained by averages is no more clearly shown than in an analysis of the range of raw water turbidity at any filtration plant. (See Tables V and VI.) For ex- ample, White River water at Indianapolis averages through- out all the years under record 40 parts per million turbidity. Analyzing these figures more closely, during only 18.6% of the time does the turbidity exceed 50 parts per million, while on the other hand during 39.5% of the time the turbidity is less than 20. On only 2.7% of the days during the entire life of the plant has the turbidity exceeded 200 parts per million, yet the operating data referred to previously indicates beyond question the cumulative effect of handling without preliminary treatment the excess turbidity in so small a percentage of the total number of days. The days of the year during which coagulant was applied to the raw water have varied from 149 to 225, and the average pounds per million of coagulant used ranged from 118 to 275. (See Table VII.) It was the opinion when pre-treatment was first decided upon that lime and iron treatment would be applicable to the local situation. This opinion was furthered by the conspicu- ous success of two plants nearby which operated with a raw water having high turbidity. Experience showed that White River water on very few days in the year carries such char- acter and quantity of suspended material as to make this method satisfactory, and on a great many days of the year, a relatively slightly turbid water (and this turbidity largely colloidal) is not satisfactorily treated except with Sulfate of Alumina. Tables V and VI detail the turbidity of the raw and settled water and Table VII the summary as to use of coagulant. 15 1! 41 '-- CD iO CO Tt< ~ o - i ^l ... 3 CO ^ : *: : .cooog -~" CO 1 *3*\ .-=, 2 IM >OCO ciooso 2 ,= 1 - - =1 1 r-----l2 M:i\iiiiiiin M iniiiiiini \ v erage \o S:ini|)lcs : s = 71" "^. * fh =.= r -' J - 18 o !S2S CO-HO5-HOOCOOO>< r *" c^r o" oo" icT oo* >o* o" oo~ < J^^fMOiCOO'^OCOOOc CO **< GO CO o- CCOO OO(M t^ ^H O> 01 C35 & OS O O5 O O5 19 In very condensed form the range of raw water turbidity and rate of coagulant required may be expressed as follows: II ANY WATER TURBIDITY. Pounds per m. g. Range Total Test Days 1904-1919 ' , of Time Alum Used 0- 10 1154 20.7 11- 20 1044 18.8 21-30 926 16.6 31-40 898 16.1 60 41- 50 514 9.2 95 51-100 618 11.1 180 101-150 170 3.0 285 151-200 98 1.8 370 Over 200 150 2.7 450 By the inclusion of this table in this place it is not meant to indicate that these amounts are not necessarily varied from time to time. The temperature of the raw water, together with the fineness of the suspended matter therein contained and the proportion of Hying vegetable material all produce different effects upon coagulant which make its use one not capable of being carried out by following the same set table but depending altogether upon the intelligent and constant observation of the effect of actual use of coagulant upon the particular water being treated at the time. CHLORINATION. The experimental studies on the use of hypochlorite of lime at Boonton, N. J., and the Bubbly Creek Plant at Chicago were investigated by this company, and beginning in July of 1909, hypochlorite of lime was applied to the Indianapolis water supply. The use of hypochlorite of lime continued until May of 1916 when the Wallace-Tiernan dry feed chlorinator was put in service and in January of 1920 the dry feed chlor- inators were modified to apply the chlorine in solution form. During the first three years of the use of the hypochlorite of lime an average of 3 pounds per million gallons was used, ex- pressed as chlorine. This was later reduced to an average of 1% pounds per million gallons and the same amount was used when the shift was made to the use of chlorine gas. The average quantity has increased during the last two years to approximately 2 pounds per million gallons due to a more stringent requirement within the organization as to the qual- ity of the final effluent. Table VIII summarizes the use of chlorine products. 359182 20 TABLE VIII CHLORINATION SUMMARY Date Million Gallons Treated Total Pounds Hypo-Chlorite Used Per Cent. Available Chlorine Pounds Chlorine or Equivalent Used Pounds Chlorine per Million Gallons 1909 1910 3.4 3 4 1911 1912 1913 1914 1915 1916 1916 6,528.772 6,700.000 7,044.600 7,306.959 6,148.793 2,145.769 5 134 191 51,853 50,250 33,000 38,880 37,861 9,945 33.1 33.3 34.4 34.7 34.3 35.2 17,170 16,750 11,352 13,481 12,986 3,501 8 619 2.64 2.5 .61 .85 .11 .64* 68** 1916 7,279.960 12,120 67*** 1917 7 489 287 14 548 94 1918 8,082.482 15,519 9 1919 8,718.511 15,228 1.75 *Hypo to May. **Chlorine from May. ***Total for year. COPPER TREATMENT. During the summer months the growth of micro-organisms in the raw water, if not specially treated, would produce of- fensive conditions in the settling basin and taste in the filtered water. The well known copper sulfate treatment has become a part of the summer routine and Table IX shows the amount and period during each year that it is used. It is recognized that copper sulfate is no more than a sedative and that algae growth cannot be stopped in open reservoirs without com- plete removal of the half-bound carbonic acid upon which the organisms thrive. The theory upon which the treatment is carried on is simply that of holding at a low figure the micro- organic growth until the water reaches the covered filters and reservoirs, when the absence of sunlight reduces the difficulty to a minimum. TABLE IX GENERAL SUMMARY USE OF COPPER SULFATE Date Mil. Gals. Treated Pounds Used Lbs. per Mil. Gal. Number of days during months of 1911 903.071 3,763 4.14 42 May, June, July, August. 1912 95.263 494 5.18 5 May. 1913 1,285.948 3,010 2.34 60 April, May, June, July, August. 1914 570.560 1,443 2.78 26 May, June. 1915 998.210 1,480 1.49 50 August, September, October. 1916 986.730 871 .9 45 August, September, October, November. 1917 1,588.300 2,046 1.29 67 July, August, September. 1918 2,455.7 4,832 1.97 104 June, July, August, September. 1919 2,268.20 4,837 2.14 91 June, July, August, September. 21 BACTERIOLOGICAL EXAMINATIONS SUMMARIZED. The Indianapolis Water Company established one of the very first privately owned laboratories in connection with a water system in November, 1903, and the investigations of the quality of the supply as well as various technical details of laboratory work have been carried on continuously since that date. From 10 to 15 thousand samples of water are han- dled annually. It is not possible to go into a complete discus- sion of details of the bacteriological content of the supply. (Refer to close of paper for complete tables of bacteriological findings.) The number of organisms growing at 37 C. in the plant effluent and the B. Coli content are sufficiently in- dicative of the condition of the water. Daily examinations of the plant effluent with incubation at 20 were carried on from the beginning of the operation of the laboratory. 37 counts were not made continuously until the year 1912. From that date until the present time an analysis of the figures indicates that 43.5% of the time the 37 count is less than 5 per cubic centimeter; 33.5% of the time from 6 to 10; 17.3% of the time from 11 to 20; 2.9% of the time from 21 to 30; 3.8% of the time from 30 to 100 ; with one day since the beginning of the 37 counts a bacteriological content of the filter plant effluent in excess of 100 per cubic centimeter. The average has ranged from 5 per c. c. in 1916 to 12 in 1912. Studying the quality of the finished water as referred to the presence of the Bacillus Coli, during 74.5% of the time no B. Coli are found in 100 c. c. of the effluent, 17.9% of the time 1 or 2 B. Coli per 100 c. c. are present, 3.5% of the time 3, 4 or 5 per 100 c. c., 3.3% of the time from 6 to 10 inclusive, and 0.52% of the time more than 10. The average B. Coli content per 100 c. c. of the filter plant effluent is 0.85. QUALITY OF RAW WATER. In the studies of the total number of organisms growing at 37 C. in the water in the various stages of the purification process in combination with the studies as to B. Coli content, there are certain striking characteristics of the figures from season to season and year to year which are worthy of com- ment. The first refers to the condition of the White River water reaching the local plant. It will be remembered that in 22 1914 certain standards of water purification and sewage treat- ment were laid down by engineers at the request of the In- ternational Joint Commission, in its investigation of the pol- lution of the Great Lakes. In paragraph 4 the statement was made, "While present information does not permit a defi- nite limit of safe loading of a water purification plant to be established, it is our judgment that this limit is exceeded if the annual average number of B. Coli in the water delivered to the plant is higher than about 500 per 100 c. c." This state- ment has a great many possibilities of interpretation, not alone in the language used but with reference to the viewpoint from which this Board of Engineers was looking at the broad ques- tion. They were not alone considering the operation of purifi- cation plants but the degree of efficiency to which sewage puri- fication plants should operate in discharging their effluent into streams which later might be used for water supply. It is the impression of the writer that they, as well as a great many other persons at the same time, had not had access to a large volume of figures referring to the actual conditions which water purification plants had to meet. It will be remembered that in previous discussions in this Association with reference to the standards for water used on interstate carriers, regret has been expressed at the real lack of a mass of information as to the quality of filtered water in municipalities where no question is raised as to quality, and even at the present time the lack in uniformity in expression of results and in volume of work done is so great as to make real comparisons difficult. In a great many well operated plants (using the term well operated with reference to the mechanical conditions and the actual quality of the finished product) the volume of labora- tory studies is not sufficient to be used as a basis of broad con- clusions. Studies on the B. Coli content of White River have been summarized in such a way as to cover the operations dur- ing the last five years (see page 25) and during that period the ranges in the number of B. Coli per 100 c. c. have varied from 695 as a minimum in September, to 9,076 in March. In other words, the minimum B. Coli content of White River at Indianapolis is higher than the expressed safe limit of the International Joint Commission. This is not an abnormal figure for streams in the northern part of the United States. On the other hand, it appears very probable 23 from such studies as can be referred to on the waters purified in the part of the country east of St. Louis and north of Wash- ington, that the B. Coli content averages as high or higher than this amount. One of the principal objects in adding such a mass of ma- terial to this discussion has been to show as thoroughly as possible what results can be attained in a carefully controlled filtration system. Limitations such as have been set by the Joint Com- mission are undoubtedly proper, if in their promulgation, suf- ficient consultation has been made with actual results attained, and a check placed upon theoretical considerations. The operation of purification plants in this country has been attended with striking reductions in typhoid death rates. The results for Indianapolis are attached to this paper. If the reduction in this type of illness is to be attributed to im- provement in water supply, as has been done, then it must fol- low correlatively that the load these plants have had to bear is not too heavy for their successful operation. Engineers cannot point to the success in reducing water-borne diseases as an argument for building filter plants, and at the same time, set as the limit for their successful operation, a figure, which if literally enforced, would legislate from use the water which these plants are purifying. This, incidentally, is not said for the purpose of condoning carelessness in the treatment of municipal sewage. Observations make clear the fact that municipal sewage is being discharged without purification to such an extent as to sensibly increase the organic loading of many streams. Without respect to the results that can be attained in purifying such water, it is not equitable that the plain duty of sewage purification should be neglected with the result of increasing the load that water purification plants have to bear. SEASONAL VARIATIONS OF BACTERIAL FLORA DUR- ING FILTRATION PROCESS. Certain variations in bacterial flora have been observed from season to season which may partly be local in their sig- nificance, but others of which may not only apply to surface waters of the Central States but to all such supplies in tem- perate climates. 24 The findings as to 20 bacterial growth and 37 bacterial growth and numbers of organisms of the colon group have been studied in detail over a period of 5 years. This data is summarized in tabular form under the following headings: 1. Basic tables. Bacteria at 20 C. 37 C. and Colon Group. 2. Percentage which 37 Count is of 20 Count. Percentage of 37 organisms which are of Colon Group. Percentage of Colon Group which are fecal type. 3. Effect of various steps of Purification process on dif- ferent types of bacteria. The fundamental seasonal variation as evidenced by the enumeration of the total number of bacteria present, as well as those of the specific colon group, is that of greater concen- tration of bacterial life during the months of low temperatures and low concentration during high temperature months, or briefly, a variation inversely proportional to the temperature. This may be accounted for by the greater proportion of de- struction of micro-organisms during the season of the year when biological activity is at its highest. The bacteria de- rived from the sewage pollution and field washings entering the streams during the warm temperature months are subject to the destructive activities of other forms of bacterial and plant life, whereas these agencies being absent or inactive dur- ing the winter months tends to prolong the existence of bac- teria derived from external polluting sources. Following the water through the various steps of the filtra- tion process there is a lessening of the seasonal variation, the filtered water showing a less variation between the bacterial concentration of the summer and winter months than is the case with the raw water. A slight reversal of this condition occurs in the sterilized effluent, where the variations again be- come somewhat greater. A comparison of the total number of bacteria as evidenced by counts made after incubation at 20 C. and 37 C., ex- pressed in the percentage which the 37 count is of the 20 count, shows that the blood temperature organisms form their lowest proportion of the total bacterial flora during the cold winter months, the minimum being 25% during February. They reach their maximum proportion of the 20 organisms 25 a p p I : s'l ~ ~ ~ 1 1 1 =1 o PS 33 O PQ . _ . _ M -.; - (M O S * | ^r-H'so" ^ o . . ^ * Q ^S5S O 3 % 05 -*0 Tto "O O5 S M SS | co cogq^-o A O5 O5 OQ g 3 00 CO^^ b *..oS h 0* ^ 3 T ^c,c,^ 1 '" 1 CO rt -a a l>- CC * CO . CO ^-t O c g * 1 o^coc3 o> S CM ,d ^ t^ 3 g ^S8 0) ^ | (M 1C CO (M o 'M" , U g ^ w fc, ^ 10 1C IO >j O5 1C CO^H > H ^ - 1 10 T cS CO -H i 05000CO *i ~* 1 * Ei o oo t^.^ -ti o a >5 1 I * g I o ^ "* *"* 1 zszz a O .s ^ X 00^ . 1 1 & *o S ^ c - S3 S3 > 5 SS3 S3Sg 3SS ^ oo oo o oo y. r. i - v ^. cs os o> * ooo * I OS OOCC O5 OC OO ( 00 CO 00 ss -X "^ i- : ^< 3 e< t--c 5^!r: S^S CO i -i 5 O i CS t^.t^ O5 ? g -, : "= Is - ' ^5 IT "is "^ '5* '- - sr 5^ 5s T i T= II 2: ^ 28 during the month of August, 88%. The same variation or ratio between 20 and 37 organisms holds true in water after settling, filtration and sterilization, except that as each step improves the sanitary quality of the water the 37 propor- tion is lowered during the cold months and equals or becomes greater than the 20 count during the summer months. Study- ing the 37 organisms and the proportion of these which are organisms of the colon type, it is noted that the percentage of B. Coli is highest during the cold weather months with an average of 3.1% of the 37 organisms conforming to the test for the colon group, the maximum in any month is 8.9% in November and 1.4% minimum in June. The coagulation and settling process alone reduces the average percentage of or- ganisms of the colon type from 3.1% to 1.5% with variations from a minimum of .2% in the month of August to 4% in No- vember. In the filtered water, with a seasonal average of 1.6% of the 37 organisms B. Coli, the variation is from a minimum of .07% in August to 4.4% in January. The steril- ization process produces a remarkable reduction of the or- ganisms of the colon type as expressed in this percentage ratio. With an average of .1% of the 37 organisms B. Coli through- out the year, the maximum is .17% in December and the mini- mum .05 for February and March and .06 in July and Sep- tember. The minimum of .05 for February and March in- dicates either a failure of conformance to any seasonal varia- tion on the part of the efficiency of chlorine or an elimination due to the use of a large amount of coagulant during these months. The latter is more probably the reason. The reduc- tion of organisms of the colon group by the sterilization process is so great as to indicate a practical selective action against this type of bacteria. Another series of examinations of the filtered and sterilized water has shown that approximately 25% of the filtered water organisms are spore formers, whereas 75% of the sterilized water organisms are spore formers. Taking this in combina- tion with the number of B. Coli present, the selective action due to sterilization may be shown in the following manner: Of 1,000 organisms in the filtered water 250 will be spore formers and 16 organisms of the colon type. Of 1,000 or- ganisms in the sterilized water 750 will be spore formers and one will be of the colon type. Of the nonspore forming or- 29 ganisms then in the filter effluent 2.16% are Coli type, while in the sterilized water only . I f '? are such, a 5 to 1 reduction. A study has been made over a period of three years of the organisms of the Colon Aerogenes Group which are of the fecal type, that is, having a positive reaction to Methyl Red. In the raw water there seems to be no sea- sonal variation, the minimum being 46% in August and the maximum 79 % in May, that is, 46% of the total number of completed B. Coli were of the fecal type. The steps in the filtration process, however, develop an elimination which by the time that the sterilization process has been completed has a de- cidedly seasonal variation, the Methyl Red positive organisms reaching their minimum percentage during the warm months and their maximum in the cold months. In other words, the survival of fecal B. Coli is less likely during the season of the year when biological activity is at its highest. Studying the reduction by the various steps in the purifica- tion process it will be noted that the reduction by settling and partial coagulation is lowest during the winter months and highest in summer. The same is true of the filtration step ex- cept that the variations from season to season are less. The irregularity of percentage reduction by sterilization process would indicate that there is no seasonal factor in the efficiency of chlorination. Studying the percentage reduction by the entire filtration process, it is beyond question the fact that organisms of the colon type are less likely to survive filtration and sterilization than is either the low temperature group as evidenced by the 20 or the blood temperature group as evi- denced by the 37 count. CONCLUSION. 1. Bacterial concentration in streams and partially puri- fied water is inversely proportional to the temperature. 2. The proportion of all organisms, which are of the gen- eral colon type, is likewise inversely proportional to the tem- perature. 3. Both settling and filtration exercise a selective action against organisms of the Colon type, and sterilization with chlorine products exercises a remarkably increased selective action against these organisms. 4. Of the total number of Coli type organisms present, the 30 Methyl Red positive or so-called fecal type survive the purifi- cation processes, step by step, in increasingly less proportion as the temperature rises. REDUCTION OF TYPHOID FEVER RATE. The installation of water purification plants assists in re- ducing materially both general and typhoid death rates. As a matter of fact, the Mills-Reincke phenomenon, so-called, indi- cates that elimination of intestinal disorders results in mate- rial reduction in various other seemingly dissociated death rates. As a corollary to the laboratory data presented in this dis- cussion, it is worth while to note the data as to general and typhoid death rates in Indianapolis since 1891. Previous to the installation of the filtration plant, the ty- phoid death rate was high, reaching epidemic amounts in 1893, 1895 and 1904. The average for all years was 51.8. After the plant was placed in operation, the reduction in rates proceeded slowly, but finally to a very satisfactory rate in 1918 and 1919. In 1916 there was a decided rise in typhoid deaths, summer typhoid associated with swimming in polluted streams. The general average typhoid rate since 1905 has been 22, a reduction to 43% of the pre-filtration days. In 1904, when there was an investigation of the water supply and general sanitary conditions in Indianapolis, in addi- tion to recognizing the necessity of completing the filtration plant already under construction, it was recommended that the large number of private wells and unsanitary privies be eliminated. Repeatedly since that time various individuals have urged, that the same action be taken, but no result has obtained. Studies of typhoid cases over a period of years locates over 80% of the total as occurring where a private well or privy or both are used. If Indianapolis performed its duty in improv- ing sanitary conditions as well as the Indianapolis Water Com- pany has fulfilled its duty to the public the typhoid death rate would be less than 1 per 100,000. The factors which appear to have operated to reduce ty- phoid are in order: 31 1. Purification of the city water supply. 2. Houses in newly built up areas equipped with city water and sanitary plumbing. 3. Anti-typhoid vaccination. SUMMARY GENERAL ANT) TYPHOID DEATH KATES 18911919 Inclusi%-e Total Deaths Death Rate Year Population All Causes Typhoid Fever All Causes per 1,000 Typhoid Fever per 100,000 Deaths Which are Typhoid 1890 105 436 1891 111,800 2,128 34 19 30.4 1.6 1892 118,200 1,985 54 16.8 45.6 2.7 1893 124,500 2,070 110 16.6 88.4 5.3 1894 130,900 1,834 56 14 42.7 3.1 1895 137,300 2,237 140 16.3 101.8 6.2 1896 143,700 2,057 75 14.3 52.1 3.7 1897 150,000 2,111 62 14.1 41.3 2.9 1893 156,400 14.4 35.2 2.4 1899 162,800 2,388 74 14.7 45.5 3.1 1900 169,164 2,626 74 15.5 43.8 2.8 1901 175,700 59 14.2 33.6 2.4 1902 182,200 2,492 76 13.7 41.7 3. 1903 188,600 2,790 93 14.8 49.2 3.3 1904 195,000 3,194 143 16.4 73.3 4.6 1905 201,300 3,081 71 15.2 35.3 2.3 1906 207,900 2,975 70 14.3 33.7 2.3 1907 214,400 3,163 62 14.8 28.9 2. 1908 220,700 2,907 60 13.2 27.2 2.1 1909 227,200 3,041 47 13.4 20.7 .5 1910 233,650 4,039 67 17.3 28.6 .7 1911 241,750 3,920 63 16.3 26.2 .6 1912 248,700 3,739 45 15.1 18.2 .2 1913 255,000 3,906 61 15.4 23.9 .6 1914 262,500 4,136 69 15.7 26.3 .7 1915 270,000 3,907 37 14.4 13.7 .9 1916 278,000 4,323 71 15.5 25.5 1.6 1917 286,250 4,587 31 16.0 10.8 .7 1918 295,000 5,273 19 17.8 6.4 .36 1919 304,000 4,137 14 13.6 4.6 .34 Average. 18911904 before filtration, 51.8. Average. 19051919 after filtration, 22 . . These figures as to typhoid death rates are presented as associate studies with the laboratory findings on the public water supply. The data as to B. Coli and total bacterial content, as well as the discussions on the quality of the raw water and the ability to purify it, are presented and made on the basis that the water supply, as a source of water-borne diseases, was eliminated when the filter plant was placed in operation. GENERAL SUMMARY. The results of operation from a standpoint of quality of supply, and the studies in the cost of operation make it possi- ble to summarize the experiences with a modified slow sand filtration plant briefly as follows : 32 The prime requirement of successful operation of slow sand filters is a proper condition of the sand layer. Operating in favor of this is the increased size of the particles applied in a pre-treated water, as well as reduced total suspended mat- ter. Operating against the condition of the sand layer are two main factors. The first is the summer increase in micro-org- anisms which is becoming a far more important factor in water purification in the Central West than is generally recog- nized. This difficulty is also a means of interfering with sat- isfactory operation of rapid sand filters. The second is "air-binding." In slow sand filters during the winter months there may be a very considerable reduction of output due to the occlusion of air within the sand layer, commonly termed air-binding." In slow sand filters during the winter months minimum temperature the capacity for solution of oxygen is at its highest. Very small changes in temperature or physical condition seem to throw a portion of this out of solution and this is particularly manifested in filter sand layers where at times during the cold months a very material restriction to the actual flow capacity of the filter unit may be occasioned by the inclusion of air bubbles between the grains of sand. This may manifest itself in a spongy consistency of the sand layer quite comparable to quicksand when the water is drained off, and is also evidenced by small craters or rosettes, as it may be termed, of sand, scattered indiscriminately over the surface where the air has gathered in large particles and forced its way to the surface. This problem has been the source of very serious difficulties, notably in the case of the Wilmington, Del., plant, and represents at the present time one factor in the operation of all filtration plants, but notably slow sand units, likely to give difficulty during the cold weather. The operation of slow sand filter plants, while it has ex- tended over a great many years, has not been the subject of such careful study in this country as have the operations of mechanical filter plants. The best summary of the basic rules of slow sand filtration of water was made by George W. Fuller at the Lawrence Experiment Station at Lawrence, Mass., in 1894. These fundamentals may be briefly re-stated as follows : 1. Bacterial efficiency of slow sand filters increases with age, other conditions being equal. 2. New filter sand is quite unlike that taken from filters which have been in operation for same time. The grains of the latter are covered with a sticky coating; in the case of grains situated at or just below the upper surface layer of sand this coating is so thick that the grains are considerably dis- colored. Here it is that the applied bacteria are detained in the largest numbers. "3. In new filters, and in old filters which have been out of operation for a considerable period, normal bacterial results do not appear to be obtained until these films are formed. "4. In old filters which are in regular operation, and which yield normal chemical and bacterial results, a marked deterioration in these results occurs when for any reason these is a well-defined mechanical disturbance of the main body of sand, whereby the continuity of the films is broken to a certain degree." * * * "5. Low rates are undoubtedly safer than high rates ; but, nevertheless, up to ascertain limit the rate apparently exerts very little influence, and this limit is different for different filters and varies with other conditions in the case of the same filter." * * * "6. With our present knowledge it may be stated that the factor which causes the effect of the rate of filtration upon bacterial efficiency to become practically nil, under normal conditions, is chiefly the age of the filter." In the operation of the Indianapolis plant, the studies made during the fifteen years make it possible to add to these observations the following: The pre-treatment by coagulant of a water supplied to slow sand filters results in the grouping together of the sus- pended particles in fairly large aggregates which substitute in a measure for the sticky coating of the surface layer. The bacterial content of a sand layer filtering pre-treated water is not so high in total numbers nor so active as in the case of a filter handling untreated influent. The bacterial efficiency of filters operating in this fashion is less than that of filters operating with an untreated influ- ent. The removal of organic material from the water to be filtered lessens the supply of material to be deposited in the filters, and at the same time interferes with certain biological processes which are more active in the plain type. 34 Slow sand filters operating with a pre-treated water are more susceptible to seasonal variations of bacterial flora, both in the influent water and in the sand layer, and may at times unload somewhat in the fashion of sewage filters. This un- loading process has no relation to the quality of raw water, and with a sterilization treatment following is not apparent in the finished product. Shutting off units and allowing them to stand for twenty- four to forty-eight hours does not seem to interfere with efficiency in production of bacterial reduction. Continuation of this, however, for a week or more seems to result in the deposition of material within the upper sand layer which ma- terially reduces the production of the filter unit on the ensuing run. Variations in depth of sand layer from eight to thirty inches have been allowed to exist on the local plant, and the results of operation indicate that the thinner layers give no less satisfactory bacterial purification. In point of ease of handling the filter unit, the thinner layer is preferable. CONCLUSION. The operation of the Indianapolis filtration plant was, in its earlier years, attended with some difficulty, which by the covering and dividing of the filters in 1905 and 1906 and the adoption of the preliminary coagulation and settling process in 1908, has been eliminated, with the result that water of excellent quality is being produced at a normal cost. The in- stallation has justified itself. Also has the amount of techni- cal control justified itself. In November, 1903, the Company established its laboratory, the operation of which has been continuous and increasing in volume. It has been the settled policy of the entire organization to leave nothing undone which would satisfy all persons concerned as to the quality of the supply. It is proper in this connection that appreciation should be expressed to the officers of the organization who have co-operated in all things looking toward the successful operation of the filtration system. It is also proper to express appreciation of the fine spirit of service, of Mr. C. K. Calvert who, since 1908, has been the chemist at the filtration plant. 35 -o * c: ~ - 1 ~> - -. 00 1 o o t- ft -i r-' / 1 c^^eM-i<-Heot^c t~ ^ 1 -- ~ ooos -- 10 t - -x . -ri re -r : r~ ao , - r. - - ~ - - - - - c-.S5S-.ci E | 459182 36 >lO^-HIM.C^O>OiOt>. SJ^^SS' > Tt< ^ CO SO l>- (M CM CS COOOOOO^O 37 ' --- ; | 7, _ ar . =r - . : . 38 Average OS ^H OS 00 l^ t^ 1C s 1 w 1 S Minimum o >oo CO 1 1 C mum S888SSS 1 3 ^ "* Average i f rt< CO *O CO O i-H SJ5SSSSS oo mum m b .2 .S 1 p5 2 3 SoSSooo s 1 M ** J * 9 * f. C CO ** 1C O ^ 00 O> 2 OS OS OS OS OS OS OS OS - - - - - 40 H B 2 H o ^ ^ t^-c^ > 05 I- CO U5 C- < t^cOr^cOC^Ot-H* fOOlOCicMtM-^ OO^jCOOOOlOCOCOCOCO^CO>0 SS^g 1 " 10 " S^V-Teg.*." OOCO SSSSfeSSSSSiS S o 1 |l!ii3 S BSI I : ^ I L = 8 i sgsssss g S l 3 \ i >, 53 S8 s- 3 gis Q 5 - -' _ i a ,g= sg o gSSB g 00 .s r * ^ = ~ ~ f3 CM M i S s. IliSSSSi i ^ i t g a |i|?|ll|l 5 = i 00"5--1000-5ce-0 g -i r= ~ t r - 5. - r i .^ 5 ~ 1- r X. ri -M CM i~- cc c-i CO | igSS55S = ggSl o re ri TI CM" s c S r ~ - i r- . - - i - s_ ti I =. ~. -. ~ -. -. -. ~. - - -. ~. 42 OOC-ooo^fio< cooo>OM*^^oo -Hoowoo .^H 71 71 44 " U5 CO OCOO3lOCOiOCO'*J' i-H 1-1 1-H i-l CO rl 00 1O i^c^SS^I' e^-HOO^c^cct^c^ooortTif--^ 1C CO M iM **< t-i CO Tjl :-t 3 s ^^ssss^s lO^*" " cJ 03 C Z c: r~. -~ 1 j^^ ec'oT 5 1 cc " ' 5 a M 3 ~. ~. . -. -. -. -- ^ d ^ * I ! o Q O = M > s 1 ;i=;|||a >c s j - s ~ s I SS5:SS^^S 1 1 S3Sg32S -H ~HO*CON 5 bi 9 < SS^ ^>, "3 i^ .-5^=2231= > 05 S 2SS? 2 i -^S|S 2 1 r -.s ?J * -^5S 1 l:Q 5 is = 1 SSg ^^^^-^0^0 s if :- . - - I - f ~. M 2 ~. -. - - -. -. -. CO ~4 . oo os bi S tNW.^cO^OOO bi) OJ OJ 05 05 OS OS 05 O3O5O5OJO5O5O5OS ^ 47 M , s2gs = 2 .- - 1 -,-, CO 1 - a- oo 1 s ^^^^. oo 1 3 s 5 3 1 ***'*-*- s SAMPLES 3. C .Growir "a - TAP Bacteria per a -- 00 I a-a- GO i =- - 2 oo I SSS="== ~ ~ ~ s; o * ^ -H I -M ~? ~f -^ - *ie.f~COOr-SS*SSSSSoSSoS ?. s g s 5 ! 3 f * I a ~ gna -i rirci-^iC-HCOOiws*-" r^ I^QOC^O^^OOOO i iiii eo **~ Is i- y. rio^O 00 iOO-*0000-*> cooo 2 5?^ ' S O* CO O * CM o o OM ooc< o QCO r^cooo -: ' is CT rt ro t>- cO O S (NOOWO-HOO M Z~ ' * o i< d m < oo -~. t^. . CO < -H' -H Tj (Mt^lMCC^OS t^ CO CO OfltfS >O ^< OJ tf3 r+ 00 O t - r i CO t~ g i - . - rc n t^ ^< ^* CO ^-4 i-^ i-^ r^ cs oo J5 S-^- 4 ;: P4V400 5 s ss Jg ^^M^OOO-U5 s; g 00 - 1 - s i coej)-* J^ 'f ^^ ; g - s CO "S 0000 ^?? go o o o w e^ -^ ec o o 01 oo t co * o CMI^O CO CO CS r^ see>e> to 3*w oa ^>c s ^^>^koccc*^ooo^ > - > r-< I> CO i-l CO ^ * U5 ^H < 5>O5t^OOOOI>-^H|>.00" o> - si i i-" SSSS8S; o * SI. ^^;^- r - = 1" 1 = -~. SSSSS532 r 1 - - r i i _ .- So s s t^ y - is r - -i y U5 if -y /: - r :;- sS --- n --=. 00 CO is I So gjSSSS-g; -- - 9 g o s o B S. : .: u II og 1 ii sa . - = = C-l 5s II \ -,- S si : _ .- _ CO 1 - -- 00 CO te ' ' E ~ ~ CO g \ " 8SK8 o" 5 ii r -S5S o 5 p - >-S2S 00 JCCO .- ~> -S-S3SS CO god s -sss 1= 8 CO 2 CO ^ ~ g*88 ro i CO f CO - 1 -- -- CN CO CO r i - M I -1 O " - j- ,"*. z a r^ -r - -r t - r r r. r - ~ -. - ' 54 pi -H oo 00 3 S 1 w 2 H H fe O o 1 1 o t^ ws >o o; cq oo CM Tjt 00 O OO CO CO t^ ^H |8 *co sgasss? OS CO 1C CM 00 oo '? ^-HOS-^OO-^CO^ O CO 'O i sssssa- S rt " So co TJI ic co r^ oo OS -. So CMCO^^CO^CCOS . 2 H OS OS OS CS SS OS OS 3^ 2n OSOSOSO50SOSOOS 2fc * H~ * H s I 5 111 l~- !O --T 28 |8 = = = = = = | f o -*oooe>o~ c^ ^ 7 **" = W = re ri n CD 2 ' s s rt = i- -/; -ri ~i OS ?J ! >o ^i = ro r^ co %~ ! ^< o 01 oo w ^< eMMr.\l> Number of days on which there occurred various numbers per 100 c. c. FIIIERKD WATEH Date 1-2 3-5 6-10 11-25 26-50 Over 50 Total Test Days Average B. Coli perlOOc.c. Bacteria per c. c. 37 1915. Jan Feb March.... April May June 3 22 20 10 4 2 3 6 8 9 2 3 2 14 4 2 4 1 2 1 2 5 1 1 7 10 22 24 27 26 26 25 11 37 1 0.35 4 113 62 26 11 9 10 18 Julv Aug Sept Get Nov Dec 2 1 8 2 2 5 15 14 9 2 3 3 2 7 9 16 6 2 10 10 1 2 5 4 3 3 3 7 27 26 26 26 26 26 20 6 4.5 2 31 95 13 . 22 13 9 13 23 Total. . . 72 75 25 68 21 19 27 307 27 19 1916. Jan 3 5 7 11 26 93 21 Feb 1 2 3 16 1 1 24 g 11 March . . . \pril 10 9 6 9 3 2 6 2 2 2 1 27 25 4 4 13 10 May June 13 16 9 7 4 1 2 1 27 26 1.4 2 10 9 July Aug 7 10 9 4 3 2 4 6 1 1 1 2 25 25 6 6.6 24 40 Sept Oct.... Nov Dec 11 16 u 4 6 7 6 4 2 3 3 4 6 ' 4 7 2 3 '"6 i 25 26 25 25 3 15 2.6 12 25 13 11 29 Total.. . 109 69 29 57 15 15 12 306 12 18 1917. Jan Feb 3 2 2 6 9 9 10 7 24 24 1,580 332 136 102 March.... \pril 13 13 12 9 2 2 o 27 24 12 12 23 10 May June 13 16 g g 4 2 1 26 26 56 10 34 61 July Aug 12 18 11 9 2 25 27 12 2 4 107 26 Sent 17 7 o 24 1 8 30 Oct 17 9 1 27 5 45 Nov g 14 4 26 20 23 Dec 5 9 4 25 192 84 Total. 137 104 42 22 305 186 57 1918. Jan 5 4 IS 27 135 158 Feb . . 7 3 14 24 135 95 March ~ 10 9 26 108 28 April g g 4 7 26 4 10 May- 10 7 7 2 26 2 9 June " 11 g 5 j 25 2 58 July 10 5 6 21 1 118 Aue g 10 7 1 27- 2 72 Sept.... Oct 16 6 g 9 2 3 1 1 4 25 27 9 1 59 11 Nov. . 3 2 7 6 7 25 17 28 Dec 2 2 3 3 9 4 25 29 33 Total.. . 90 54 42 41 12 20 45 304 38 57 60 BACILLUS COLI COMMUNIS Number of days on which there occurred various numbers per 100 c. c. FILTERED WATER Cont. Date 1-2 3-5 6-10 11-25 26-50 Over 50 Total Test Days Average B.Coli perlOOc.c. Bacteria per c. c. 37 1919. Jan Feb . . . March.... April May "3" 3 9 7 '"s" 8 8 4 '"*" 4 8 6 5 4 2 4 1 1 3 4 8 ...... 1 12 " i 26 24 25 26 25 127 4 11 2.8 5 6 64* 16 18 9 16 June July 4 15 10 7 1 2 4 2 3 i 23 26 9 1 57 56 Aug Sept. 15 4 7 8 4 6 6 " i' '"i"' 26 26 .8 6 81 40 Oct 12 8 3 4 27 2.1 12 Nov 1 2 7 9 6 25 66 19 Dec 1 1 2 21 26 191 37 Total.. . 74 68 42 34 21 24 42 305 35.5 35 BACILLUS COLI COMMUNIS Number of days on which there occurred various numbers per 100 c. c. FILTER PLANT EFFLUENT Date 1-2 3-5 6-10 Over 10 Total Test Days B. Coli per 100 c.c. Bacteria per c. c. 37 1915. Jan Feb 11 19 10 3 2 1 ........ 23 24 1.2 0.6 19 14 March 26 1 27 04 8 April 23 3 26 0.23 5 May 21 4 1 26 4 4 June. . . 17 8 1 26 0.7 5 July Aug 10 12 7 13 1 1 4 3 25 26 5.0 1 6 4 Sept Oct. 18 12 7 10 1 2 '"2 26 26 0.6 1 7 5 7 Nov Dec 15 10 9 5 2 5 '"4"' "Y" 26 26 1.0 4.5 6 11 Total.. . 194 80 17 11 5 307 1.4 8 1916 Jan Feb March. . . . April May June July. ..... Aug Sept Oct Nov Dec 16 21 26 25 22 25 21 19 26 23 24 25 5 2 1 5 1 3 4 1 1 1 2 2 1 2 2 1 26 24 27 25 27 26 25 25 26 26 25 26 4.2 1 0.07 0.0 0.4 0.08 0.6 0.9 0.0 0.4 0.04 0.04 8 4 7 4 4 4 3 5 3 5 5 11 Total. . . 273 24 3 5 3 308 0.64 5 (II BACILLUS ('"I. I cn.MMr.\i> Number of day* on which there occurred various number* per 100 C. FlI.TKIt 1'l.AM Km.rt\T Cont. 1-2 ^ 3-5 6-10 Over 10 Total H. Coli per 100 c.c. Bacteria per e 37 1917. Jan 11 10 3 | o 27 2 >\ Feb March... \pril 19 18 21 4 6 4 2 o 1 o o 24 ..-> .9 25 10 5 May June July Aug. .. Sept Oct Nov.. Dec 22 21 2\ 22 20 20 18 5 4 5 4 5 3 6 8 6 13 It 26 25 28 24 27 27 25 1.5 .3 .3 .25 .4 .6 5. 5 4 5 4 5 5 7 29 Total... 218 65 11 17 311 1.01 10 1918 Jan. .. . 5 5 10 27 4 35 Feb 10 7 3 4 24 2 22 March 16 o 4 26 1 l 13 April Mav 10 24 12 2 4 26 26 1 3 6 4 June 23 2 25 .12 8 Julv 28 3 26 5 \ug 25 25 8 Sept 19 5 1 25 3 6 Oct 20 7 27 3 4 Nov. . . Dec 22 19 2 4 '"2 1 25 25 .4 6 7 5 Total 216 57 19 15 307 86 10 1919. Jan Feb 19 22 7 2 26 24 .3 11' 7 '.i March. . . . \pril 24 26 1 25 26 . .16 11 5 May June July .. 19 17 20 6 3 6 ::::::: ""i" 25 22 26 .3 .7 .3 7 Aug 24 2 26 8 Sept 20 5 l 26 .5 7 Oct X en- 16 20 9 4 2 i 27 26 .7 .4 ',) 4 Dec 18 4 2 2 26 .9 7 Total ... 245 49 6 4 304 .37 7 62 BACILLUS COLI COMMUNIS Number of dayj on which there occurred various numbers per 100 c. c. TAP WATER Date 1-2 3-5 6-10 Over 10 Total Test Days B. Coli per 100 c. c. Bacteria per c. c. 37 1915. Jan 14 4 1 3 22 1 4 25 Feb 17 7 24 5 14 March 26 26 9 April May June July Aug Sept Oct Nov 24 23 14 3 1 12 3 2 2 1 6 7 6 8 18 8 '"3"' 4 6 3 1 3 8 8 2 2 7 '"4"' 4 '"2"' 4 26 25 26 26 25 25 26 26 .12 .32 1.7 7 8 1.9 3.8 6 5 5 7 10 8 6 8 9 Dec 6 4 3 11 2 26 7 14 Total. . . 145 71 26 45 16 303 3.1 10 1916. Jan Feb 6 20 9 1 2 2 8 1 1 26 24 4.5 7 11 (j March. . . . April May June July Aug Sept . 24 22 7 8 11 9 13 3 2 4 7 10 5 9 2 2 \ 1 6 7 3 4 y 2 '"2"' 27 25 26 26 25 24 24 0.2 0.5 11 5 2 4.8 1 i 8 6 10 29 72 22 7 Oct Nov 19 19 6 3 26 23 0.6 7 7 8 Dec 23 2 1 25 0.16 15 Total... 181 61 16 30 13 301 2.6 17 1917. Jan Feb 21 21 2 5 1 o 26 24 1.9 75 20 20 March April 21 18 2 4 2 o 25 22 .7 7 8 g May 25 o 26 15 6 June July 22 1!) 2 2 2 4 o 26 25 1.1 3 4 5 10 Aug Sept 12 12 11 7 4 5 o 27 24 3.1 3 2 9 7 Oct. .. Nov Dec 11 14 15 5 3 5 3 4 1 9 25 21 24 3 1.3 4 7 12 41 Total. . . 211 - 8 39 37 295 1.94 13 1918. Jan Feb 17 18 8 3 1 1 26 22 7 5 41 30 March.... April May June July Aug 22 10 12 5 11 13 1 13 11 18 10 9 "Y" 1 1 1 5 3 1 ""i" ....... 3 26 26 24 25 26 27 1.2 1.3 .7 2.4 5 1 11 12 5 11 8 14 Sept... Oct Nov Dec 4 17 19 21 6 9 4 4 9 1 i ....... 5 25 26 25 25 9 .4 .7 .2 10 5 8 5 Total. . . 169 85 20 18 11 303 2.9 13 BACILLI'S Col.l Co.M.Ml M- Number ill days on whir'i there oerurred various numbers per 101) c. c TAP WATER Cent. Date 1-2 3-5 6-10 Over 10 Total Test Days B.Coli tOOe.e. Bacteria per c. c. 37 1919. Jan 16 9 25 4 9 Feb 22 24 08 g March ... April... 20 4 4 1 25 26 .6 2 11 6 May.-- 2 j 25 24 5 June 16 5 1 22 5 10 July. . . 4 26 2 7 AUK.... - 21 4 1 26 3 g >ept 8 13 5 25 1 g 7 14 6 27 1 6 4 Nov Dec 17 18 3 g 2 1 23 26 .8 3 4 6 Total. . 211 71 16 2 300 .52 7 BACILLUS, COLI CU.MMUMS Five Year Totals Number of days on which there occurred various numbers per 100 C. C. RAW WATER Date 50 51-100 101-500 501- 1,000 1,001- 5,000 Over 5,000 Total Test Days Average B. Coli perlOOc.c. Bacteria ' Pe-- 1915 1916 1917 64 66 44 44 63 77 104 96 28 21 113 60 43 10 15 74 310 304 308 1,409 1,173 6 584 203 1 913 1918 1919 24 55 39 88 111 66 36 26 62 50 33 309 305 4,305 2,895 1,946 981 Total. 222 14 5 278 18 1 399 26 264 17 1 191 12 4 182 11 9 1,536 Average. 44 55 100 53 48 36 307 SETTLED WATEK Date 10 11-50 51-100 101- 500 501- 1,000 Over 1,000 Total Test Days Average B. Coli perlOOc.c. Bacteria per c. c. 37 1915 1916 1917 1918 1919 114 149 127 79 87 35 53 '"73"" 81 55 37 84 56 28 58 35 '"25" 58 16 13 56 40 19 14 4 18 22 32 292 291 285 295 305 355 118 1,091 647 565 83 86 443 . 456 289 Total 556 242 260 176 144 90 1 468 % 37 9 16 5 17 7 12 9 8 6 1 111 60 52 44 29 18 293 64 BACILLUS COLI COMMUNIS Five Year Totals Number of days on which there occurred various numbers per 100 c. c. FILTERED WAIER Date 1-2 3-5 6-10 11-25 26-50 Over 50 Total Test Days Average B. Coli perlOOc.c. Bacteria per c. c. 37 1915 1916 1917 72 109 137 75 69 25 29 68 57 104 21 15 19 15 42 27 12 22 307 306 305 27 12 186 19 18 57 1918 1919 90 74 54 68 42 42 41 34 12 21 20 24 45 42 304 305 38 35.5 57 35 Total. % 482 31 6 266 17 4 138 9 304 19.9 69 4.5 120 7.9 148 9.7 1,527 Average. 96 66 34 61 17 24 29 305 FILIER PLANT EFFLUENT Date 1-2 3-5 6-10 Over 10 Total Test Days Average B. Coli per 100 c.c. Bacteria per c. c. 37 1915 1916 1917 1918 1919 194 273 218 216 245 80 24 65 57 49 17 3 11 19 6 11 5 17 15 4 5 3 307 308 311 307 304 1.4 ' 0.64 1.01 0.86 0.37 8 5 29 10 7 Total. 1,146 74 6 275 17 9 56 3 6 52 3 4 8 5 1,537 Average. 229 55 11 10 1 307 TAP WATEH Bate 1-2 3-5 6-1 J Over 10 Total Test Days Average B. Coli perlOOc.c. Bacteria per c. c. 37 1915 1916 1917 191,8 1919 145 181 211 169 211 71 61 8 85 71 26 16 39 20 16 45 30 37 8 2 16 13 11 303 301 295 303 300 3.1 2.6 1.94 2.9 0.52 10 17 13 13 7 Total. 917 306 117 122 40 1 502 % Average. 61.0 183 20.4 63 7.8 23 8.1 24 2.7 8 300 65 \\ATKR WORKS STATISTICS FOR THE YEAR 1920. INDIANAPOLIS \\ATKK COMPANY, INDIANAPOLIS. INDIANA. Population of Indianapolis January 1, 1920 314,194 Total estimated population supplied 252,000 Date of original construction, 1870. By whom owned, Indianapolis Water Company. C. H. Geist, President. C. L. Kirk, Vice-President and General Manager. F. C. Jordan, Secretary. E. C. Leible, Treasurer. Source of supply, White River and deep wells. Emergency supply from Fall Creek. Water flows from White River near Broad Ripple, a town about eight miles north of the center of the city of Indianapo- lis, through a canal owned by the Water Company to the Filter Plant, where it flows through a Sedimentation Basin and thence through six slow sand filters. The slow sand filters have a daily capacity of 36 million gallons, the average daily yield for 1920 being 24.7 million gallons. The filtered water after being chlorinated flows to the pump well at the main station of the Indianapolis Water Company, known as the Riverside Station, and to a reservoir at this station having a capacity of 5V& million gallons. From the reservoir water flows by gravity to another station known as the Washing- ton Station, where all the pumping is done by hydraulic power furnished by the Canal. The flow from Broad Ripple to and through the Filter Plant, thence to the reservoir at the Riverside Station and to the Washington Station is entirely by gravity. The Filter Plant is located near Fall Creek and the pump- ing station at the Filter Plant is provided with low lift pumps in order to obtain an emergency supply from Fall Creek in the event of any interruption of the flow of the water through the Canal caused by a break or on account of im- provements which are made along the Canal necessitating such interruption. At the Riverside Station there are 43 deep wells with a capacity of from 18 to 22 million gallons daily; 18 of these wells are operated by air pressure from the central pumping station. The others are provided with electrically driven centrifugal pumps. 66 The eastern portion of the city, which is approximately 100 feet higher in elevation than the main part of the city, is supplied from the Fall Creek Station, which obtains its supply from 12 deep wells with a capacity of 6,200,000 gallons daily. In addition to the two main sources of supply the Water Company owns a small -station near Broad Ripple with a capac- ity of one million gallons. The supply is obtained from a deep well. All wells are from 330 to 350 feet deep and have been drilled through rock. The well supply is practically sterile. The Water Company maintains a well-equipped laboratory at the filter plant for the chemical and bacteriological exam- ination of water. Samples are collected daily from all parts of the purification, pumping and distribution systems, and deter- minations made for the total number of bacteria and B. Coli in each sample. About twenty thousand bacterial counts and fifty thousand B. Coli estimations are made yearly. (1) Pumping: Riverside Station Capacity One Hamilton Rarig Vertical Triple Expansion High Duty Pumping Engine 30 m. g. d. One Snow Vertical Triple Expansion High Duty Pumping Engine 20 m. g. d. One DeLaval Steam Turbine Driven Centrifugal Pump ... 30 m. g. d. One DeLaval Steam Turbine Driven Centrifugal Pump. . . T 1 /^ m. g. d. One DeLaval Steam Turbine Driven Centriugal Pump... 6 m. g. d. Total capacity at the Riverside Station for fire service 93 MJ m. g. d. For domestic service the capacity is over 100 m. g. d. .Fall Creek Station One Allis-Chalmers Horizontal Cross Compound Pumping Engine 6 m. g. d. One DeLaval Steam Turbine Driven Centrifugal Pump ... 6 m. g. d. Total 12 m. g. d. Washington Street Station Three Water Turbines operating DeLaval Centrifugal Multi-Stage Pumps, with a capacity of 14% m. g. d. Broad Ripple Station One DeLaval Centrifugal Pump 1 m. g. d. Supplementary pumping is done for the Fall Creek Station by Booster Pumps at two stations, with a capacity of 6 m. g. at one sta- 67 tion and 12 m. g. at the other. These pumps are DeLaval Centrifugal Pumps, electrically driven. (2) 1> u of Coal Used: Indiana Coal Mine Run and Screenings. The Riverside Station is provided with coal crushing machinery. Percentage of Ash varies from 10 to 20 Coal consumed during year 1920, approximately 15,000 tons. Total Pumpage for the year 1920, 11,037,902,000 gallons. Total fixed capital, or value of property and plants, as shown by books of the Indianapolis Water Company as of December, 1920 $11,694,476.56 Capitalization Funded debt: General Mortgage 5% Bonds $2,359,000 First and Refunding 4Ms% Bonds 3,711,000 Total funded debt $6,070,000.00 Capital Stock common $5,000,000 Capital Stock preferred l f . .'< 295,000 Total Capital Stock 5,295,000.00 Total Capitalization $11,365,000.00 Consumption 1920 1. Estimated total population of district at date, 314,914. 2. Estimated total population supplied by the Indianapolis Water Company, 252,000. 3. Total number of gallons consumed for year, 11,037,902,000. 4. Percentage of consumption metered (estimated), 54' 0. Average daily consumption in gallons, 30,159,000. 6. Gallons per day to each inhabitant, 87. 7. Gallons per day to each consumer, 120. 8. Gallons per day to each tap, 653. Distribution 1. Kind of pipe used, cast iron. 2. Sizes, 4-inch to 40-inch. (All new mains arc 6-inch in diameter or larger.) 3. Extensions made in 1920, 60,000 feet. 4. Discontinued, none. 5. Total now in use, 450 miles. 9. Fire hydrants added in 1920, 87. 10. Number of public hydrants now in use, 3,763. 11. Gate valves added in 1920, 92. 12. Number now in use, 3,689. 15. Range of pressure on mains at center of city Domestic service, 45 to 50 Ibs. Fire pressure service, 85 to 95 Ibs. 68 Services 16. Kind of pipe Lead, %-inch to 1%-inch. Byers W. I. pipe, 1^-inch to 3-inch. Cast iron, 3-inch to 8-inch. 17. Sizes, %-inch to 8-inch. 21. Services added in 1920, 1,933. 22. Number of services now in use, 46,165. 25. Meters added in 1920, 466. 26. Meters now in use, 6,325. THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW AN INITIAL FINE OF 25 CENTS WFLL BE ASSESSED FOR FAILURE TO RETURN THIS BOOK ON THE DATE DUE. THE PENALTY WILL INCREASE TO SO CENTS ON THE FOURTH DAY AND TO $1.OO ON THE SEVENTH DAY OVERDUE. IN STACKS LD 21-100m-7,'33 Gaylord Bros. Makers Syracuse. N. Y. PAT. JAN. 21, 1908 UNIVERSITY OF CAL.FORNIA L I BRARY P":^S