*-^- '^V^S \\s. ^^ ^ 1 % *'^A^^1»*. ^0 ,-^ \,' ^^^^v^t\ #> Qlotnell Utttoeraitg Cibrarjj THE LIBRARY OF EMIL KUICHLING. C. E. ROCHESTER. NEW YORK THE GIFT OF SARAH L. KUICHLING 1919 Cornell University Library TD 430.F95 The purification of the water supply of 3 1924 004 984 708 The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924004984708 Vol. XXXV. AUGUST, 1909. M>. 6. AMERICAN SOCIETY OF CIVIL ENGINEERS INSTITUTED 185 8 PAPERS AND DISCUSSIONS This Society is not responsible, as a body, for the facts and opinions advanced in any of its publications. THE PUEIFIOATION OF THE WATER SUPPLY OF STEELTON, PENNSYLVANIA. By James H. Fueetes, M. Am. Soc. C E. To BE Presented October 6th, 1909. The plant for the purification of the Steelton water departs, in several particulars, from previously constructed types, and embraces features of some novelty. The interest expressed by engineers and municipal authorities vsrho have visited it, and seen it in operation, inclines the writer to believe that, although the plant is not a large one, a description of the conditions leading to its design, as well as an account of its construction and the results thus far obtained in its operation, would bring out an instructive and useful discussion of certain subjects on which more information is desirable. The plant is a slow sand filter having a capacity of 3 000 000 gal. of filtered water daily; and among its novel features are: 1. The use of coarse-grained filters instead of subsiding basins in the preparatory treatment of the water ; 2. The small area required — about one-fourth of that ordinarily required for a slow sand filter plant of equal capacity; 3. The system of handling and caring for the sand removed from the slow filters during scraping; Note. — These papers are issued before the date set for presentation and dis- cussion. Correspondence is Invited from those who cannot be present at the meeting, and may be sent by mail to the Secretary. Discussion, eltber oral or written, will be published in a subsequent number of Proceedings, and, when finally closed, the papers, with discussion in full, will be published In Transactions. 660 WATER PURIFICATION AT STEBLTON, PA. [Papers. 4. The system of determining, measuring, and applying the chemical solutions; 5. The construction of, and the filtering materials in, the rough- ing filters; 6. The method of operating the roughing filters, and the safe- guards to prevent the deterioration of the effluent; Y. The control of the rate of filtration of the slow filters. Among the novel features pertaining to the operation of the plant are the following: 1. The very high rate of operation of the slow filters; 2. The small quantity of coagulant necessary to secure satisfac- tory results — about one-fifth of that required in a mechan- ical filter plant of equal efficiency, and about one-eighth of that required to secure equally satisfactory results with coagulation and subsidence; 3. The ease of manipulation of the chemical dosing, and the certainty in its application; 4. The long runs of the roughing filters when no coagulant is necessary, and the consequent saving in wash-water, power, and attention; 5. The long runs of the slow filters, and the resulting large quantities of water filtered between scrapings; 6. The uniformity in the quality of the water applied to the slow filters, and consequently the satisfactory results ob- tained in the final purification; ^ 3 / '!'• The brilliancy of the effluent from the slow filters ; S-3 ~ 9. The moderate cost of construction and operation. The Borough of Steelton, having a population of approximately 15 000, is in Dauphin County, Pa., on the same side of the Susque- hanna River as Harrisburg, and about 3 miles farther dovsm. Topo- graphically, the territory within the borough limits consists of a flat portion, about 20 ft. above the level of the river and from 1000 to 1 500 ft. in width, which extends the whole length of the borough, and the residential district, which extends up the side hill to the east, and is much broken, and traversed by streets having for Han 8. The absence of trouble with rusty hot water and troubles from the destruction of the solder in copper closet tanks; Papers.] WATER PURIFICATION AT STEELTON, PA. 661 most part heavy grades. The principal street skirts the eastern edge of the flat area, the larger portion of which is occupied by the works of the Pennsylvania Steel Company. Original Waier-Worhs. — The original water-works plant was built by the Steelton Home Water Company, and was purchased by the borough in the latter part of 1899 for $150 069.48, since which time improvements to the intake, pumping station, and reservoir, and exten- sions of the street mains swelled the total cost to $223 000, on January 1st, 1906, of which $145 000 was the cost of the street mains, $38 000 the cost of the reservoir, and about $40 000 the cost of the intake and pumping station. The plant, at that time, obtained its water from the Susquehanna River through a 16-in. intake pipe reaching out approximately 1 400 ft. from the pumping station, to within about 150 ft. of the shore of Stuckers' Island. The original supply had been derived from wells dug on that island, but in a comparatively short time it became impossible to keep up the supply from the wells, and a pipe was extended about 50 ft. out into the river from the most southerly well, the water being then dravm from the river to the wells and back to the pumping station through the 16-in. pipe. Sub- sequently, owing to the silting up of the wells and their inaccessibility during high water in the river, a section of the 16-in pipe was taken out at a point about 150 ft. east of the island, allowing the water to enter the pipe from the river direct; it was in this condition on Jan- uary 1st, 1907. The water flowed through this broken-into intake pipe to a briek- and-concrete-lined well situated about 150 ft. from the river, just east of the Pennsylvania Railroad tracks, and was pumped from this well to the service reservoir, 2Y0 ft. above the river, by two 1 500 000- gal. Deane, compound, direct-acting, outside-packed, plunger pumps, through a 12-in. force main. An additional 16-in. main, connected up at the pumping station to permit either pump to pump direct or to the reservoir, parallels the 12-in. force main for about 1 200 ft., extending to Second Street on Connestoga, and is cross-connected with the distribution system and the 12-in. force main. The service reservoir, when full, holds aBout 7 500 000 gal. Character of Susquehanna Biver Water. — The Susquehanna River, above Steelton, has a water-shed of approximately 24 300 sq. miles. It draips practically one-half of the State of Pennsylvania, and about 662 WATER PDKIFICATION AT STEELTON, PA. [Papers. 6 080 sq. miles in the southern central portion of New York State. Upon its banks are many large cities, including Binghamton, Coopers- town, Corning, Elmira, Hornellsville, Horseheads, , and Owego, in New York State, and Athens, Altoona, Bellefonte, Clearfield, Emporium, Harrisburg, Hazleton, Huntingdon, Lewistown, Lock Haven, Mahon- ing, Mifflintown, Mount Carmel, Nanticoke, Pittston, Plymouth, Sayre, Scranton, Shamokin, Sunbury, Towanda, Tunkhannock, Tyrone, Wilkes-Barre, and Williamsport, in Pennsylvania. Approximately, 1 000 000 persons reside in the cities and towns along the river and its branches above Steelton, of which probably 750 000 are located along the North Branch, about 100 000 on the West Branch, and 150 000 on the Juniata, which streams unite a few miles above Steel- ton to form the main Susquehanna River. In addition to this urban population, there is a very large rural population, and, of the larger cities, there are upwards of 60, having a total population of about 700 000, which have sewers discharging directly into the river and its branches. The anthracite coal fields of Pennsylvania are on the water-shed of the North Branch, and along the whole length of the river and its tributaries there are many important manufacturing industries, the wastes from which are discharged into its waters. The West Branch lies largely in a sandstone country, and the principal indus- tries along it have been in connection with the manufacture of lum- ber. The Juniata lies very largely in a limestone district. In the days of canal navigation, several dams were built on each of the main streams, canals paralleling the streams on one side, or both, between the pools formed by the dams. The principal dams on the North Branch were at Clark's Ferry, Sunbury, Nanticoke, Wilkes-Barre, and Binghamton; on the West Branch, at Lewisburg, Muncy, Williamsport, Lock Haven, and Queen's Run; on the Juniata, at Millerstown, Lewistown, Newton, Hamilton, and Huntingdon; and on the branches at Pipers, Petersburg, Big Water Street, Little Water Street, Willow Street, Donnellys, Smokers, Mud, Williamsburg, Three Mile, Crooked, Prankstown, and Hollidaysburg. The existence of these dams is mentioned because they play an important part in the modi- fication of the character of the river water during the season of low water as well as during the early stages of the fall and spring floods. Essentially, it will be seen that these various streams consist of ,a serips Papers.] WATER PURIFICATION AT STEELTON, PA. 663 of comparatively steep natural slopes joining level pools, vphicli, dur- ing low and moderate stages of water, form subsiding basins in which suspended matters carried by the water may be deposited in greater or lesser amounts. During freshets, however, the deposits which may have accumulated in the pools during the preceding quiet periods are scoured loose from the bottom, picked up by the rushing waters in a more or less changed condition, and carried down the stream in large quantities. The larger part of the water-shed is semi-mountainous, the main streams flowing, for the greater part, through unerodable, and com- paratively steep valleys, and the hillsides being largely denuded of forest areas; for this reason the stream and its branches are more or less flashy, big floods rising to a height of many feet in a few hours and falling almost as rapidly. The amount and character of the sediment in a turbid water is of importance in determining the method of preliminary treatment best adapted to prepare the water for final purification. . At Steelton, the quantity which will subside to the bottom of a 1-gal. bottle will sometimes reach a depth of i in., and is usually a composite of sand, coal dust, leaves, stones, silt, nitrogenous matter, coagulated clays, as well as aluminum hydrate and the hydrated oxides of iron mixed with various forms of dirt, bacteria, and other matters. In the Susquehanna River water at Steelton a brown sediment, varying from red to yellow in color, either flocculent or otherwise (the flocculence depending largely on the stage of the river and the in-- fluence of the stage on the accumulation of bacterial masses and clay •particles into agglomerations), is generally characteristic of the water in the west channel, although a sediment of this color may sometimes occur over the whole width of the river, following heavy general rains. ' A microscopic examination of this brownish sediment shows that it consists of particles of coal and dirt agglomerated by growths of small water plants and filamentous bacteria, probably of the order Bacterium, or Leptothrix. A greenish-black sediment, resulting from the mixture of coal dust and clay, is generally flocculent, and is characteristic of the water in the east channel during heavy floods. A grayish sediment, due either to finely-divided suspended matter other than clay, or to the hydrates of aluminum and of iron, which 664 WATER PURIFICATION AT STEELTON, PA. [Papers. settle to the bottom when the river is low, is characteristic of the water in the east channel, and is locally called "sulphur water." A black sediment, generally more or less flocculent, consisting of finely-divided particles of coal from the mining districts, is character- istic of the east channel, and can be observed on the bottom of the river below islands and bridge piers, and in other protected positions when the river is at low stages. All possible gradations of color intermediate between these are observable at different times owing to the commingling of the dif- ferent waters. The presence of red or yellow clay in the Susquehanna River water at Steelton, in increasing quantities, is almost uniformly accom- panied by decreasing alkalinity; such water clears up slowly by sub- sidence. The water carrying flocculent sediment, which is generally black during floods but often the clean hydrate when the river is low, will clarify very quickly by subsidence. Such waters are characterized by low alkalinity and correspondingly high permanent hardness. The color of the turbidity of the Susquehanna River water, there- fore, is a fairly good index of its character, and, during the operation of the filters, plays an important role as a guide to the proper use of the coagulant. The changes which take place in the character of the sediment, as the result of bacterial activity, are intimately associated with the iron, sulphur, and alkaline constituents of the water. Shortly after the Harrisburg experimental filtration plant was put in operation, in the fall of 1902, one of the roughing filters, then operating without a coagulant, became suddenly clogged to a condition of water-tightness within a very few minutes after the turbid raw water was turned on, and an examination disclosed a heavy blanket of black, sticky mud on the surface of the sand. It was surmised that this mud might be similar to that which was reported to have accumulated in the sedi- mentation basins at Cincinnati during the experiments conducted by George W. Fuller, M. Am. Soc. C. E., but a chemical examination of the mud on the Harrisburg filter showed that while iron was present in considerable quantities it was not in the sulphide form. The mud was found to consist essentially of very fine particles of coal dust mixed with the hydrates of iron and aluminum. Although it had been understood for a loog time that the drainage Papers.] WATEK PURIFICATION AT STEELTON^ PA. 665 from coal mines contained a considerable quantity of free sulphuric acid, an explanation of the processes by which this was formed was not generally current. Recognizing that ferrous' sulphide, as pyrites, was present in the coal dust in the river, it was suspected that the "sulphur bacteria" and "iron bacteria," which work various trans- formations in iron compounds, might play some part in the formation of this sulphuric acid, and an experiment was made to test the hypothesis. A small quantity of coal dust or culm was pulverized in a mortar, and equal portions were placed in narrow-mouthed bottles. In each of these bottles was then placed a definite quantity of un- treated river water, and, the mouths of the bottles having been closed with cotton plugs, half the number were sterilized in an autoclave at a pressure of 15 lb. for 30 min. After standing in the laboratory for about a month the bottles which had not been sterilized exhibited very different characteristics from those which had been so treated. After both sets of bottles had been shaken violently the turbidity in the set which had not been sterilized was observed to decrease very much more rapidly than that in the set which had been sterilized, some- what after the manner that the untreated river water containing this hydrated oxide of iron acts in comparison with water which does not contain it. The water in these bottles represented, in some degree, the condition of the water in the river, except that the bottles which had been sterilized contained no living bacteria, while the others, supposedly, did, having been filled with river water which at the time of filling was known to contain a large number. The results of this preliminary experiment seemed to indicate that the change which had taken place in the unsterlized bottles was due to the action of the bacteria on the iron compounds contained in the culm. Following out this idea, a further elaborate set of tests finally led to the conviction that a hydrate of iron was formed from the pyrites, in the presence of bacteria, while it was not formed when bacteria were absent; further, there was no evidence of the formation of soluble iron except when bacteria were present. The reactions which took place in the bottles used in the experi- ments are not known, and are probably of a very complicated character, some resulting directly from, or being intimately associated with, the biological activities present, and others in all probability being spon- taneous chemical reactions. It would be foreign to the subject of 666 WATER PUKIFICATION AT STEELTON, PA. [Papers. this paper to enter into a discussion of these experiments in detail and of others which have since been made both on Susquehanna and on Schuylkill waters; it will be sufficient to state that at the conclu- sion of the Harrisburg experiments the data indicated that the bac- teria occurring normally in the Susquehanna Eiver, in the presence of organic matter on which they can live, and under proper conditions of storage, are able to transform the pyrites in the coal into compounds which furnish the material for the hydrated oxide of iron naturally occurring in the water. The indications seem to be that the sulphur in the pyrites, in contact with water, furnishes the material from which the bacteria make sulphuric acid, which, coming in contact wilii the iron, forms the soluble ferrous sulphate. The water containing this soluble ferrous sulphate commingles with the alkaline waters of the river farther down stream, the sulphate of iron being decomposed, the sulphur combining with the lime and magnesia to form the sulphates of these bases, and the iron being converted into the hydrate and sub- sequently into the hydrated oxide. This seems to be at least one method by which free sulphuric acid is formed in the drainage of the mining regions; the process also takes place to a very considerable extent in the level reaches of the river, where culm is deposited as floods recede. When the river rises, and the dirt on these flats is washed away, there is borne with it the hydrate and hydrated oxide of iron which had been formed during exposure. At the Clark's Ferry dam, some miles above Harrisburg, the water in the east channel some- times shows this coagulated condition so plainly that frequently within an area of a few square feet one may collect a sample of either per- fectly clear, or very black water. At times of very low water there occurs along the east shore a brownish flocculent sediment composed of the hydrate of aluminum as well as the hydrate of iron. The source of the aluminum entering into the composition of this hydrate is unknown ; possibly it is derived in some way from the clays. As a result of the natural formation of the hydrated oxide of iron, and of aluminum hydrate, in the water of the river, a considerable degree of apparent self-purification takes place during the summer months. This is brought about, in the pools above the dams, by the entangling of the bacteria into flocculent masses and their conse- quent settlement to the bottom of the river with the other suspended Papers.] WATER PURIFICATION AT STEELTONj PA. 667 matter. The action is, on a large scale, very similar to the results of the treatment of sewage by chemical precipitation, and the active precipitants are the same as those used in the artificial process. Ulti- mately, a large proportion of the organisms thus carried to the bottom cease to exist in an active condition, finding their environment not suited to sustain their vitality. The apparent purification is at times so great that, although from the many cities along its course enough sewage is flowing continuously into the river to render the water at its extreme low stage offensive, yet at such times the water, as it flows past Steelton, is actually at its best, both in appearance and as to its hygienic qualities. The analyses of different samples of river water collected at Harrisburg during low water have shown as few as 10 colonies of bacteria per cu. cm., with no intestinal bacteria present; and generally, during low-water stages, when the quantity of sewage discharged into the river is greatest relatively to the stream flow, the numbers of bacteria in the river water opposite Steelton (from points far enough away to be beyond the influence of the Harrisburg sewage) are lowest and the numbers of intestinal bacteria fewest. This condition obtains, however, only during dry weather, for the first floods which come down in the fall scour out a portion of the deposits from the pools formed by the dams up stream, and then the counts show sometimes as many as 300 000 colonies of bacteria per cu. cm. of water, as well as large numbers of bacteria characteristic of intestinal discharges. The total hardness of the river water varies with the stage of the river, yet not necessarily in proportion thereto, and is about the same, as a general thing, from bank to bank. The water is hardest at low stages when the flow is largely from springs, that on the west side being high in carbonates and low in sulphates while that on the east side is low in carbonates and high in sulphates, owing to the chemical changes taking place, as above described. These modifying conditions destroy any relationship between tur- bidities, river stages, and numbers of bacteria. High bacterial counts sometimes accompany very low, as well as very high, turbidities, and, similarly, relatively low turbidities are sometimes found at very high stages of the river; further, the color of the turbidity is no indication of its total amount, as a comparison of the records of ten samples of water, each with a turbidity of 1 000 parts per million, collected at 668 WATER PURIFICATION AT STEBLTON, PA. [Papers. different times, and at different parts of the river, shows all grada- tions of colors between bright yellow, red, dark brown, and black. As a result of these complex conditions, the water in the river may change quickly from a moderately hard, clear water to a soft water carrying several thousand parts per million of turbidity, which, on occasions, may intermittently come down coagulated to a more or less slimy or flocculent condition with hydrated oxide of iron and hydrate of alumina. On the other hand, the river from bank to bank may be carrying its maximum load of yellow clay from the Oonodo- guinet and Juniata, or reddish-brown mud from the West Branch. During a general flood from the entire water-shed, all these condi- tions may prevail at the same time, the yellow water from the Juniata keeping on the western side, the West Branch water keeping in the central part, and the black water from the North Branch along the eastern shore, each stream occupying practically one-third the width of the river, which, opposite Steelton, is approximately four-fifths of a mile. Although of great width, the river is comparatively shallow, the bed in that vicinity being of shale and limestone, and the depth of the water at its lowest stage being not more than 3 ft. at any point and not more than IJ ft. for the greater part of the width. Some trouble is caused during the winter by slush-ice and anchor-ice, the formation of which is favored by these conditions. The Steelton Filter Plant. Those who have had experience in the filtration of polluted waters will appreciate some of the difficulties encountered in the purification of the Susquehanna Eiver water, particularly as to bacterial purifica- tion, for the reason that the numbers of bacteria in the raw water may be increasing rapidly while the turbidity is decreasing; and hence, if the attempt is made to control the bacterial purification by operat- ing the plant on the indications of the removal of turbidity only, the result may be anything but satisfactory. Another complication arises from the exhaustion of the alkalinity of the raw water by the acid mine wastes, and also its reduction by dilution during heavy freshets. This, however, can readily be taken care of by adding enough lime or soda-ash to supply the deficiency in alkalinity, in cage a coagulant is used in connection with the process Papers.] WATER PURIFICATION AT STEELTON^ PA. 669 of purification. Among otlier complications is the difficulty of secur- ing the satisfactory coagulation of the raw water when the bacterial counts are high and the turbidity is low, and also the difficulties en- countered in securing a sufficiently heavy coagulum when the tem- perature of the water approaches the freezing point, at which times the turbidities are also usually low. While all these difficulties can be overcome and the water can be handled successfully with a mechanical filter plant such as that installed for the City of Harrisburg, the successful operation of such a plant depends on the maintenance of a fully equipped laboratory} and on a superior grade of supervision; and, to secure these, it is essential, of course, that the revenues derived from the sale of the filtered water be sufficient to cover the additional expense. Owing to the considerable general pollution of the river, the dis- charge of the sewage of Harrisburg into it only a couple of miles above Steelton, the excessive turbidity, and the enormous quantities of sand and coal delivered into the borough's distributing reservoir by the supply pumps (ultimately to enter the distribution mains, interfering with the action of meters, and clogging service pipes), the question of securing a purer and more satisfactory source of supply was brought to an issue in the winter of 1906, and the writer was com- missioned to investigate and report on what could be done toward securing a new supply of a satisfactory quality, or toward purifying the existing supply. The report recommending the purification of the present supply was submitted to Councils early in 190Y, and, following its adoption, the borough voted to issue $80 000 in bonds to cover the cost of the necessary works. The plant, as built, was designed to place the securing of satis- factory results, as to purification, within the reach of the Water De- partment without the necessity of maintaining a laboratory, and with- out incurring unduly heavy construction and operation costs. The general basis of the design is as follows : First. — The removal by subsidence of such of the suspended mat- ters as will settle out from the raw water in about 12 min. ; Second. — The removal, by passing the water rapidly through deep, coarse-grained filters, of at least 90% of the applied turbidity, using a coagulant to secure this result when necessary, and allowing no coagulated turbidity greater than 25 parts per million to issue from 670 WATER PDEinCATION AT STEBLTON, PA. [Papers. the coarse-grained filters, no matter what the turbidity of the applied water; Third. — The filtration of the effluent of the coarse-grained filters through slow sand filters at a relatively high rate. New Intake. The works include a new 30-in. cast-iron intake pipe extending out about 1 500 ft. into the river to a point where observations indi- cated that there would be a minimum of trouble with silt, sand, and coal dust. At the point where the intake was located prior to the commencement of these improvements large quantities of sand and fine particles of coal were drawn into the intake well during fioods in the river, as much as a car load a week requiring removal in order to keep the pumps in operation. The new intake pipe was laid to proper line and grade, in coffer- dams, in a trench excavated in the rock bottom of the river, the joints being poured with lead and caulked water-tight. The intake cage consists of a reinforced concrete structure inclos- ing the end of the intake pipe and spreading out like a fan to give an opening in front 1 ft. high and 20 ft. long, with vertical screen bars of 1-in. round steel, spaced 6 in. apart, in two rows, and staggered. The top of the opening was located at a depth of 1 ft. below extreme low water, and the rock bottom of the river was blasted away in front of the cage, a smooth concrete apron being laid thereon to assist the current in rolling sand and coal particles along past the intake. The line of the face of the intake is parallel with the trend of the current of the river at that point. A valve is placed in the intake pipe, at the shore end, and a con- nection is made with the existing 24-in. pipe leading to the suction well. The work on the intake, which was commenced on July 1st and completed early in September, 1908, was done by the Water Board by day labor, from the plans and under the general supervision of the Improvements at the Pumping Station. New Pumps.— The plans, as approved by Councils on July 10th, 1907, called for the construction of a 3 000 000-gal. slow sand filter plant with roughing filters, to be built on a piece of ground 100 ft. wide and 265 ft. long, situated about YOO ft. east and 900 ft. north of Papers.] WATER PURIFICATION AT STEELTON, PA. 671 the pumping station. Before the construction of the works had been authorized, appropriations had been made for the installation of a new 3 000 000-gal., Heissler, compound-condensing, crank-and-fly-wheel, pumping engine at the pumping station. In order to suit the new conditions, it was necessary to remodel the piping in the pumping station and install two 12-in. centrifugal pumps, direct-connected to simple, vertical, condensing engines of the marine type, the suction pipes of the new pumps being connected to the remodeled original suction to which the suctions of the old Deane pumps were attached. The discharge pipe from the two centrifugal pumps was piped to the filter plant, with provision for a cross-connec- tion to the suction pipe of the new Heissler pump, so that, in case of accident to the new filtered-water well, raw water could be pumped directly into the Heissler pump and thence to the city. The Deane pumps were located in a pit sufficiently low to enable them originally to take their own suction from the intake well, but the new Heissler pump, as well as the remodeled Deane pumps now take the filtered water from a new filtered-water well constructed just outside of the pumping station. Prom this brief description it will be seen that in the remodeled plant the raw water is pumped from the intake well by centrifugal pumps to the filter plant, and the filtered water is returned to the new pump well just outside of the pumping station, from which either the new Heissler or the old Deane pumps can draw the filtered water and force it through a 12-in. force main to the Jistribution reservoir on the hill above the town, or through the 16-in. return main directly into the distribution system, or into the reservoir and the distribution system at the same time by properly manipulating the control valves on the cross-connections. By this arrangement the construction of an additional force main to the reservoir, which would have been an immediate necessity owing to the increased consumption of water, was deferred for many years by reason of the provision of a pure water which could be pumped into the mains direct. Baw-Water Delivery and Filtered-Water Return Mains. — The raw- water pipe leading from the pumping station to the filter plant, and the return pipe leading from the filter plant to the new pump-well are of wood-stave construction, with cast-iron specials for bends and connections to the concrete. Papers.] WATER PDHIFICATION AT STEELTON, PA. 673 preparing the raw water for slow filtration by removing therefrom enough of the turbidity and bacteria to permit the slow filters to be operated at a relatively high rate. The object of the treatment is to prepare the water in such a way as to be able to discharge upon the slow filters, for final purification, a water of reasonably constant com- position, as to turbidity and bacterial contents, throughout the entire year, keeping the maximum limits of turbidity and numbers of bacteria in the applied water down below figures which experience indicates to be desirable. The treatment begins with the removal of the floating matters from the water by coarse screening and of the particles of sand, silt, and coal dust by subsidence. The water is then passed through roughing filters containing a deep bed of coarse-grained filtering material in which an additional portion of the remaining suspended matter (in- cluding a varying percentage of bacteria) is retained, the effluent water passing to the slow sand filters for final purification. As a general rule, when the numbers of bacteria in the raw water are relatively low and when the turbidity does not exceed 50 parts per million, the rough- ing filters produce an effluent of satisfactory quality without the use of artificial coagulation ; but when the turbidity of the raw water is higher than 50 parts per million, or when it is caused by particularly fine particles, or when the bacteria number more than about 5 000 per cu. cm. in the raw water, a sufficient quantity of coagulant is mixed with the incoming raw water, as it enters the deposit chamber, to produce a rough filtered effluent with a turbidity of not more than 10% of that of the applied water, and not more than 25 parts per million, in any case. Adherence to these limits, which are as yet only tentative, will apparently give sufficiently long runs of the roughing filters to permit of easily keeping the plant in operation, and sufficiently long periods of operation of the slow filters to enable them to pass a satisfactory quantity of filtered water between scrapings. The period of subsidence allowed in the deposit chamber, for the maximum capacity for which the plant is designed, is 12 min. The roughing filters are designed to operate at a rate of 1Y2 000 000 gal. per acre of filter surface per day when delivering 1 500 000 gal. per filter, three filters being provided so that two can yield the full 3 000 000 gal. while the third is being washed. The net area of the sand surface of each slow filter is 0.1446 acre, and when delivering 1500 000 gal. 674 WATER PURIFICATION AT STEELTON, PA. [Papers. per day the filter will operate at a rate of 10 373 000 gal. per acre per day, two filters yielding 3 000 000 gal. daily with the third in reserve for cleaning. Deposit Chamber. — The raw water is received at the filter plant at the bottom of a rectangular compartment in one end of the rein- forced concrete deposit chamber, Plate XL VII, a valve-controlled con- nection being provided between the bottom of this compartment and the sewer. One outlet from the upper part of this compartment leads to the deposit chamber and a second leads directly to the channel feeding the water to the roughing filters. By using one or the other of these outlets the water may be passed through the deposit chamber, or, if desired, the deposit chamber may be by-passed and the raw water delivered directly to the roughing filters.' After leaving the inlet cham^- ber, on its way to the deposit chamber, the water first passes vertically downward through a coarse screen composed of parallel lines of 2-in. planks standing on edge, horizontally, and spaced about 8 in. apart from center to center transversely to the direction of the entering water; between these planks, and in a vertical plane, one above the other, are the horizontal perforated pipes through which the lime-water and coagulant solutions are admitted to the raw water. The raw water is admitted to the deposit chamber in the manner described, in order to use up the entering velocity head and permit the water to pass through the deposit chamber quietly and thus drop out as much as possible of the suspended matter in the short time allowed therefor. Grit Elevator. — Near the end at which the raw water enters, a grit elevator is installed to remove continuously the sand and coal particles which may collect in the deposit chamber. The elevator. Plates XL VII and XL VIII, which has an estimated capacity of 20 tons per 24 hours, consists of an 8-in., 6-ply rubber belt carrying perforated, galvanized-iron buckets, 6 in. long, 4 in. wide and 4 in. deep, at inter- vals of 2 ft. The belt travels at the rate of 25 ft. per min., and the buckets discharge into a hopper arranged to fit into the end of a 6-in. cast-iron pipe leading to one of the sewer compartments of a roughing filter. The deposit chamber is 10 ft. wide and 42 ft. long, including the entrance chamber and screen box, and at full capacity the water flows through it with a depth of about 9.2 ft. and a velocity of 0.05 ft. Papers.] WATEE PURIFICATION AT STEELTON^ PA. 675 per sec, passing over a weir at the farther end into a channel leading to the inlet openings to the roughing filters. Chemical-Mixing Tanks. — ^Vertically above the deposit chamber, and carried on heavy reinforced concrete beams spanning the chamber, stand the mixing tanks. Plates XLIX and L, for the lime-water and coagulant solutions. These tanks are of monolithic reinforced concrete. In each tank there is a wooden dissolving rack with an open bottom carrying wooden frames provided with copper screens having 20 meshes per inch, reinforced by a coarse copper screen having 4 meshes per inch, with No. 11 bars. Each tank is provided vnih a system of g-in. perforated copper pipes at the bottom connecting with a line of galvanized-iron pipe leading from above the tops of the tanks to the air receiver, to provide for the agitation of the solution after it has been mixed. A galvanized-iron steam connection, with a copper branch reaching to the bottom of the coagulant tanks, is provided for warming the solution during cold weather. The solution is mixed in the usual manner. With a 2% solution, two tanksful will furnish a dose of 2 gr. per gal. for a consumption of 3 000 000 -gal. daily. The coagulant solution is drawn from the mixing tanks through 2-in. copper pipes arranged" with swivel- joints to permit of decanting the solution at any desired depth. Coagulant-Measuring Box. — Through the copper outlet pipes the solution runs to the coagulant-measuring box. Fig. 1, and Fig. 1, Plate LI. This is a reinforced concrete box with a plate-glass front in which are drilled three J-in. and three J-in. orifices, all at the same elevation and spaced 4 diameters apart and 4 diameters of the J-in. holes from the bottom and side edges of the plate. This glass plate is 11 in. wide, J in. thick, and 28 in. high, and the holes are counter- sunk one-haK the depth of the plate from the outside. On the back of the plate is etched a graduated scale reading to hundredths of a foot for a height of 2 ft. above the centers of the orifices. The plate is fastened to the front of the orifice box on soft rubber gaskets, made of tubing, by brass plates and expansion bolts secured in the concrete. The depth of the solution in the orifice box is regulated with a bronze lever valve and float arranged to admit the solution near the bottom and permit the regulation of the depth in the box by changing the height of the float. The float is of "copper, 10 in. in diameter, with a flat bottom and a coned top, sliding on a J-in. brass rod, the rod being 676 WATEE PDEII'ICATION AT STEELTON, PA. [Papers. hinged to the lever of the valve, and the height of the float being adjusted on the rod with a bronze set-screw having a milled head. Each orifice is provided with a soft rubber cork, or stopper, at- tached to a handle consisting of a piece of brass pipe with a cap on one end and an elbow on the other, the cork being fastened to a nipple entering the elbow; there is a rack, with thumb-screws, across the top of the orifice box, for holding the handles where desired. Solution Tank O .".".■. o.-o.a • COAGULANT ORIFICE BOX olant Solution Adjustable Float / -m ■:oi."b/ff Pipe to Raw Water Fig. 1. The coagulant solution flows through the orifices in the glass plate to a small concrete box, from which a 2-in. copper pipe connects with the perforated pipes between the screen bars at the entrance to the deposit chamber. Lime-Water Mixing Tanks. — The lime-water apparatus consists of two tanks for making the lime-water (Plates XLIX and L), a slacking box, and an orifice box for measuring the quantity of water to be supplied to the lime-water tanks. Lime being only slightly soluble in water, a relatively large quantity of solution is required FLOOR PLAN OF OPERATING OF ROUGHING AND SLOW-FILTER ROOMS, MACHINERY ROOM COAGULANT STORAGE R0( Papers.] WATEE PUEIFIOATION AT STEBLTOlf, PA. 677 unless the lime be added as cream of lime. In the Steelton plant the apparatus is designed to make use of a saturated solution of lime-water. The lime is slaked in a reinforced concrete box supported on top of the partition separating the two lime-mixing tanks, a sluice-gate being provided on each side for emptying the milk of lime through funnels into pipes leading down to the bottom of the lime-mixing tanks. The latter are rectangular in plan, the bottom of each being in the form of an inverted frustum of a pyramid. From the bottom of each a pipe leads to the sewer, with a connection immediately below the bottom of each tank leading to the lime-water measuring box which stands on brackets above the coagrulant-mixing tanks and receives its water through a float and lever valve (with spindle and float as described for the coagulant orifice box) either from the city mains or from a 2-in. centrifugal pump taking its suction from the channel containing the rough-filtered water. Valves in the various pipes permit of cleaning the tanks and pipes, and the admission of water to either of the tanks as desired. The lime-water orifice box is also of reinforced concrete, with a plate-glass front, J in. thick, 20 in. wide and 28 in. long, containing two J-in., two 1-in., and one IJ-in., orifices; there is also a graduated scale etched on the back of the glass, the details of the arrangements being similar to those for the coagulant orifice box. The effluent from the solution-measuring box issues vertically into the mixing tanks through the center of the bottom, rising as a saturated solution through the milk of lime, which has been mixed and deposited in the bottom of the tank. The quantity of water rising through the tank is varied from time to time in ac- cordance with the different conditions of the raw water and the vary- ing consumption of water by the borough, the maximum rate, for a dose of 2 gr. per gal., for a consumption of 3 000 000 gal. daily with both tanks in service, being such that the water would rise through the tanks at the rate of Y in. per hour if the outlet valves were closed. The water, having picked up its lime at the bottom of the tank, is skimmed off just below the surface by a system of perforated, hori- zontal, galvanized-iron pipes. Application of Chemicals. — ^For mixing the lime-water with the raw water, six lines of li-in. galvanized-iron pipes, perforated with double rows of fV-in. holes 3i in. apart, screwed into a 4-in. galvanized- iron manifold connected vri.th a loose-sleeve joint to the lime-water 678 WATER PURII'ICATION' AT STEELTON, PA. [Paper supply pipe, lie between the wooden bars of the screen at the entrance to tbe deposit cbamber. The copper pipes for distributing the coagulant solution consist similarly of six lines of 1-in. pipes perforated with double rows of A -in. holes spaced 3| in. apart and fastened to a 2J-in. copper manifold ; they lie vertically below the lime-water mixing pipes. EouGHiNQ Filters. The raw water, after passing through the deposit chamber, or in case it is by-passed around the deposit chamber, enters a channel, running lengthwise along one side of the deposit chamber from the SECTION THROUGH INLET TO ROUGHING FILTER Fia 2. weir at the east end to the entrance chamber at the west end, the side and bottom of this channel being supported by vertical steel bars built in the concrete and passing continuously through the side of the channel and the beams spanning the deposit chamber and carrying the mixing tanks. From this channel, Plate XL VII, the water is taken to each of the three roughing filters through 16-in. cast-iron pipes con- trolled by 16-in. sluice-gates operated with chain-hoists. Between the ends of the three filters and the east side of the deposit chamber are arranged a series of chambers, three for each filter, containing con- nections to the sewer, to the filtered-water effluent pipe, and to the raw- water supply, a connection being made at the bottom, between this SECTION THROUGH REGULATING WELLS OF ROUGHING FILTERS AND ELEVATION OF COAGULANT MIXING TANKS ^Si-^ffalv. iron steam pipe ZJ^'galv. iri ID pipe ^ ] WATER PURIFICATION AT STEELTON^ PA. 679 latter chamber and tlie chamber leading to the sewer, the purpose of which is to permit the dirty water resulting from the washing of the filters to flow backward through this inlet chamber to the sewer leading to the river. Between the row of regulating chambers for the filters and the deposit chamber is a narrow channel, 3 ft. wide, and about 10.5 ft. deep, into which the rough-filtered water is discharged from the roughing filters, and from which the supply pipe leads to the slow filters. The 16-in. supply pipes to the roughing filters cross the channel containing the rough-filtered water, and enter the supply chambers immediately in front of the center of each roughing filter. The wash- water troughs, which serve the purpose of distributing the incoming water over the filters during operation, as well as of removing the wash-water when the filters are being cleaned, are suspended from the roof beams of the roughing filters by 1-in. bolts engaging saddle pieces of flat steel, over the top and under the bottom of the troughs, with jamb-nuts at top and bottom to permit of adjustment. The roughing filters are each 12" ft. 2 in. wide and 29 ft. 6 in. long at the bottom, the sides and ends battering outward 6 in. in the height of the filter, the outside surfaces of the walls being vertical. The side walls and roofs of the filters are of reinforced concrete, and the bottoms of concrete without reinforcement. Eoughing-Fitter Under drains. — In each filter there is an under- drainage system, similar in principle to that designed by the writer in 1902 for the Harrisburg Filter Plant. It consists of 59 parallel 2-in. galvanized-iron pipes, extending across the width of the filter and drilled along the bottom with J-in. holes 3 in from center to center, the pipes entering a manifold made of 12-in. flanged pipe built into the concrete wall on one side of the filter and terminating in the effluent regulating chamber in front of the filter in a 12-in. gate-valve operated by hand from the floor above. The 2-in. galvanized-iron pipes are placed in the filter so that their bottoms stand 1 in. above the floor of the filter, and are surrounded with a 4-in. layer of gravel composed of stones which will pass through a screen having |-in. meshes in the clear between the wires and remain on a screen having J-in meshes in the clear between the wires, and containing no particles finer than J in. in largest dimensions. Upon this rests a second layer of fine gravel, 3 in. thick, all the particles of which will pass through a screen having J-in. meshes, but will remain on a standard brass sieve having 680 WATER PUKIFICATION AT STEBLTON, PA. [Papers. 12 meshes per lin. in., and containing not more than 2% of particles which will pass through a standard sieve having 15 meshes per Im. m. These specifications were written for local materials, and might require modification for other materials of different subsiding values. Filtering Materials in Roughing Filters. — The filtering material in the roughing filters consists of a layer of fine anthracite coal screenings, 5 ft. thick, prepared from particles of fine coal washed down the river from the culm piles of the mines on the North Branch water-shed during floods and recovered from the river in the neighbor- hood of Harrisburg and Steelton by centrifugal pumps. Considerable care was required in the selection of the raw material, as it was desired that the finished product should have an effective size of about 1 mm. and a uniformity coefScient as low as practicable without making the cost of preparation too great. The material as received at the plant contained considerable moisture in the pile and was prepared by casting over a screen having 4 meshes per in., the actual separation with the wet coal being practically the same as would be obtained by screen- ing the dry material through a sieve having 6 meshes per in. Each of the three filters contains 5 ft. in depth of this screened coal, which, as prepared, has an effective size of 1 mm., a uniformity coefficient of about 2.4, and does not contain more than 2% of particles which would pass through a screen having 30 meshes per in. In preparing the river coal, 144 tons of the finished product, weighing about 53 lb. per cu. ft., was required, to secure which it was necessary to screen 176 tons, rejecting 32 tons of coarse particles. Boughing-Filter Controller. — The rate of filtration of the roughing filters is controlled by regulating the depth at which the rough-filtered water flows over the edge of standard bronze weir-plates, 12 in. long, Plate L and Fig. 3, the weirs being placed in orifices in the front walls of the effluent regulating chambers of the filters, the rough- filtered water falling over the weirs and into its channel, from which it is conducted to the slow filters, the elevation of the weirs being fixed so as to allow losses of head up to about 2J ft. Loss-of-Head and Bate Gauges. — The rate of filtration and loss of head are indicated, for each filter, by gauges and indexes operated by spherical, seamless, spun-copper floats, resting on the water in three 8-in. spiral riveted tubes, Plate XL VII and Tig. 3, placed in the regu- lating chamber of each roughing filter. PLATE LI. PAPERS, AM. SOC. C. E. AUGUST, 1909. FUERTES ON PURIFICATION OF WATER, Fig. 1.— Roughing-Filter Operating Room. Fig. 2. — Forms for Roofs op Slow Filter. Papers.] WATER PURIFICATION AT STEELTON, PA. 681 The rate gauge was graduated at the plant from the actual meas- ured discharge over each weir in stated periods of time, and at dif- ferent depths of flow, as indicated by the float gauge. The loss-of-head gauge was divided into feet and hundredths. The gauge boards were made of strips of poplar, painted with three coats of white-zinc paint, graduated by hand with indelible India ink, the figures and letters being lead pattern-makers' letters attached to the gauge boards with shellac, on the first coat of paint, and painted white with the board before the latter was graduated. After graduation the tops of the letters were blackened, and the board was given two coats of spar varnish and stiffened by small brass cleats Fig. 3. screwed to the backs at the top and bottom, the cleats being provided with ears on each side drilled with holes, tjV in- in. diameter, parallel to the back of the board and lengthwise of the same. The gauges were strung on parallel guides formed of No. 20, B. & S. gauge, phosphor- bronze wire fastened to screws at the top and bottom in the head and foot blocks. Roughing-Filter Washing Devices. — Provision for washing each roughing filter was made through an 8-in. connection with a 12-in. pipe laid especially to bring filtered water to the filter plant from the force main leading from the city pumping station to the service reser- voir, with a cross-connection also to the 16-in. return main, so that 682 WATER PCTEIFICATION AT STEELTON, PA. [Papers. water may be drawn from either one or the other of these two pipes as desired. The 12-in. pipe divides, on reaching the filter plant, an 8-in. branch, Plate XLVII, entering the roughing-filter plant at the bottom of the channel which receives the rough-filtered water, with connections with the underdrain of each filter back of the effluent con- trol valve, each of these connections being controlled by its own valve operated by a hand-wheel from the operating floor above. In addition, one other 8-in. branch is taken off from the 8-in. wash-water main and terminates in the sewer chamber of the central filter for use in flushing the sewer, if necessary. In addition to the wash-water, provision is made for scouring the beds of the roughing filters with air admitted to the underdrains through 2-in. galvanized-iron pipes leading from the air receiver and provided with control valves and pressures gauges. The air pipes connect with the wash-water pipes in the bottom of the filtered-water effluent chamber of each filter between the wash-water control valve and the effluent underdrain. Seiuer Connection. — Under the regulating chambers of the rough- ing filters there is an 18-in. vitrified pipe sewer. Plates XLVII and L, laid in concrete and extending to the river, a distance of approxi- mately 1 000 ft. Into this sewer all parts of the plant can ultimately be drained. The floor of the operating room of the roughing filters forms the roof over the various regulating chambers and the channel for the rough-filtered water; and the manholes providing for entrance into the various chambers are covered with light, 20-in., circular, cast-iron covers with ring-frames built into the concrete, manhole steps being provided in the walls. General Water-Supply Pipes. — The water supply for mixing the coagulant solution, for supplying the steam-heating plant, the toilets, and the sand-washer plant of the slow filters, is taken from the 8-in. wash-water main in the effluent channel of the roughing filters by a vertical 3-in. galvanized-iron pipe rising above the ceiling of the operating room and extending lengthwise of the roughing filters to the sand-washer room in the second story of the building over the regulating chambers of the slow filters. The branches for the different fixtures are taken from this main line where required. Operating Platform for Chemical Tanhs. — From the floor of the operating room of the roughing filters, reinforced concrete stairways SLOW FILTERS 200— All aide walU reinforced ' horizontally thuB Sand Trough Overflow Weirs Sand return Pipe JBl. 301.83' SECTION ON A-B SECTION ON OD Papers.] WATEE PDEIFIOATION- AT STEBLTON, PA. 683 Plate L, lead up on either hand to the elevated platform in front of the coagulant-mixing tanks, and the lime-water mixing tanks. These are provided with iron-pipe railings, painted black and varnished. Centrifugal Pump.— The centrifugal pump. Fig. 1, Plate LI, which supplies the water through the measuring box to the lime-mixing tanks, has a 2-in. discharge pipe, and a 2-in. suction pipe, extending down into the channel containing the rough-filtered water, having on its lower end an elbow, short nipple, and gate-valve with extension stem reaching up above the floor of the operating room and there provided with a hand-wheel. Just below the floor level of the operating room a 2-in. horizontal branch extends through one of the roughing filters to the slow-filter operating room where a hose-valve for 2-in. suction hose is provided in order to permit of using the centrifugal pump to empty the regulating wells of the slow filters below the elevation to which they could be drained naturally. Electric Motors. — Power to operate the machinery in the plant is derived from the electric current supplied by the York Haven Power Company, and is of the alternating-current type delivered at a voltage of 220 at the switch-board. The motors, Plate SLlX, of which there are two, manufactured by the Westinghouse Electric and Manu- facturing Company, one of 5 h.p. and the other of 15 h.p, are of the alternating-current type, wound for 3-phase, 60-cycle, 200-volt terminal voltage, capable of standing an overload of 25% for 2 hours without injury, and provided with oil-immersion starting boxes. A marbleized- slate switch-board, mounted with volt meter, ampere meter, watt meters, circuit breakers, ground detectors, pilot lights, clock, and main and circuit switches for the different services required, stands in the machinery room. Lighting. — All the filters and different parts of the plant are pro- vided with electric light fixtures wherever light may be required, and all the rooms of the building containing the machinery and the regu- lating chambers of the different filters are also provided with gas jets. Air Compressor. — The air compressor for supplying compressed air for scouring the filtering materials in the roughing filters, and for agitating the coagulant solution, stands between the two motors, and can be operated from either one. It is an Ingersoll-Rand compressor, known as the Imperial Type 11, and consists of two cylinders stand- ing vertically, with single-acting plungers driven by crank shafts on opposite sides of a heavy fly-wheel which also serves as the belt pulley. 684 WATER PUEIFICATION AT STEELTON, PA. [Papers. There is provided on the outlet from the compressor an unloading device, so-called, which, when the pressure reaches the maximum for which the machine is intended, allows the excess pressure to be relieved automatically. The air compressor has a capacity of 39 eu. ft. of free air per min. at a speed of 200 rev. per min., and is capable of delivering the air at that rate under a pressure of 100 lb. per sq. in. The air cylinders are water- jacketed, with hooded ends, the cooling water being discharged into the trough of the filter below; the air valves are of the poppet type, and work vertically. Air Receiver. — A 2-in. galvanized-iron discharge pipe, with a brass check-valve near the compressor, leads from the compressor to the air receiver, which is a vertical cylindrical tank, with domed ends, 5 ft. in diameter, and 12 ft. high. The shell is of flange-steel, is furnished with a manhole and a cast-iron base, and is proportioned to stand a working air pressure of 110 lb. per sq. in. and remain air-tight under that pressure. The receiver is provided with connections for the 2-in. galvanized-iron discharge pipe from the compressor and the 2-in. galvanized-iron air pipe leading from the receiver to the air wash-pipes of the roughing filters, and has also a IJ-in drip at the bottom, pro- vided with a double valve. The receiver is provided with a safety valve set at 105 lb. per sq. in., a gauge to indicate the pressure in the receiver, and a IJ-in. Foster reducing valve capable of discharging the entire contents of the re- ceiver in 2 min. at a uniform pressure on the filter side of the valve of about 5 lb. per sq. in., the pressure of the receiver falling, at the same time, from 100 to about 10 lb. per sq. in. The two motors, the air compressor, and the air receiver stand on foundations built monolithic with the roof of the west roughing filter, and no vibration is apparent when these machines are operating. Countershaft. — The countershaft to which the motors and air com- pressor are belted is provided with a jaw coupler between the belt pulleys from the two motors, and is extended through the south wall of the machinery room into the regulating room of the roughing filters, Plates XLVni and L, where additional countershafts are installed in order to reduce the speed as required for the operation of the grit elevator, and to increase the speed for the operation of the centrifugal pump supplying the water for the lime-water solution. PLATE Llll. PAPERS, AM. SOC. C. E. AUGUST, 1909. FUERTES ON PURIFICATION OF WATER. Fig. 1. — Tops of Slow Filters During Placing op Earth and Top-soil Covering. Fig. 3.— Interior of Slow Filter, Showing Placing of Underdsains and Gravel Layers. Papers.] WATER PURIFICATION AT STEELTON, PA. 685 Superstructure.- — The building covering the operating room of the roughing filters, the coagulant storage room, the' machinery room, and the regulating room of the slow filters. Fig. 2, Plate LIV, is L-shaped, one leg of which is 44 and the other 65 ft. long, measuring on the long sides, the two wings being, respectively, 21 and 26 ft. wide. The outer faces of the side walls are of red brick, white brick being used on the inside faces. The building is surmounted by a slate roof on a timber framing, with copper ridge, finials, gutters, and down- spouts, the down-spouts discharging into the troughs of the roughing filters, or into the deposit chamber, as necessary. Sampling Devices. — ^In front of each roughing filter, and by the side of the entrance chamber for raw water, a small concrete bracket. Fig. 1, Plate. LI, and Fig. 1, Plate LV, containing a copper funnel in the center, is fastened to the wall, and carries a nickel-plated goose-neck on the discharge pipe of a small rotary hand-pump, the mechanism of which is submerged in the water immediately below the shelf. The pumps are operated by extension shafts carrying nickel- plated hand-wheels. These pumps are used for obtaining samples of the raw water as it enters the plant and of the effluent from each roughing filter. g^^^ -^^^^^^^_ The rough-filtered water is conducted to the slow sand filters through a line of 16-in. wood-stave pipe with 12-in. connections to each slow filter, controlled by sluice-gates with stems extending to the tops of the filters under a manhole. i.rea.— The slow filters, Plate LII, and Fig. 2, Plate LIV, of which there are three, with a total net area of filtering surface of 0.4340 acre, lie side by side immediately north of the roughing filters. They are each 97 ft. 6 in. long, the two end filters being 64 ft. 7J in., and the central filter, 64 ft. 9 in. wide, all dimensions being measured inside at the bottom, and all interior wall surfaces battering out toward the top 6 in. in their height. The thickness of the outside walls of the filters is 15 in. at the bottom, and 9 in. at the top; the thickness of the division walls is 21 in. at the bottom and 9 in. at the top. The walls are reinforced horizontally and vertically to take the necessary strains, and are built in sections ending against headers, with copper strips in all vertical joints to cut off any possible leakage at such points. 686 WATEE PURIFICATION AT STEELTON, PA. [Papers. The floors of the filters are laid in blocks, and form inverted groined arches, with the columns supporting the roof standing on the high points. Roof, Bide Walls, and Floors. — The roof of each filter is supported on forty reinforced concrete posts, 12 in. square from the roof to a point 4:i ft. from the bottom, battering out to a width of 18 in. on each side at the floor level. Longitudinally of each fllter, reinforced concrete beams, 19 in. deep, run across the tops of the columns in alternate rows, while transversely of the filters, beams, 12 in. deep, run across the tops of all the columns in each row; on the inter- mediate lengthwise rows of columns the functions of beams for carry- ing the roof are performed by the deep sides of the troughs intended for the storage of sand removed from the filters during the periodical scrapings, the roof of the filter and the intersecting roof beams being tied to, and suspended from, the side walls of these troughs by vertical steel rods. The reinforcement of the side walls of the filters consists of three vertical 1-in. bars, opposite each panel point of about 11 ft., running from top to bottom of the wall, the horizontal reinforcement con- sisting of f-in. bars, 12 in. from center to center, for the lower 8 ft. in depth, and |-in. bars, 12 in. from center to center, for the remainder of the height, bars being used in both faces of each wall. The side walls rest in grooves formed along the outer edge of the floor slabs when the latter were laid. In laying the floor slabs, no particular pre- caution was taken to prevent leakage, further than to form grooves in each side of each slab as it was laid, allowing plenty of time for the concrete to harden and contract, before laying the next slab against it, and using a grillage of J-in. bars, 4 ft. long, in the concrete of the floor under the bottom of each post supporting the roof. The grooves in the edges of the floor blocks were made V-shaped so that the settlement of any particular block would tend to wedge the joints tight. Sand Troughs. — The sand troughs. Fig. 2, Plate LI, and Fig. 1, Plate LIII, on top of the filters, of which there are three to each filter, are divided by cross partitions into boxes from 8 to 9 ft. long, separated at one end by weirs and at the other by manholes opening into the filter. The boxes are arranged so that the weirs of two adja- cent boxes stand 12 in. apart, forming a channel crosswise of the sand box, from one end of which leads a 6-in. drain-tile connecting with a PLATE LIV. PAPERS, AM. SOC. C. E. AUGUST, 1909. FUERTES ON PURIFICATION OF WATER. -i^^^^B H^g I E 9 B^Kf^p— ""M a 1 alH^Wi 3ga o^ f ^^K ¥illi!alii ^.; ■-"0^^ Fig. 1.— Interior, of Slow Filter, Showing Sand in Place. fTiG. '2.— Tops of Slow Filters. Showing Sand Trougbs Papers.] WATEH PUEIi'ICATION AT STEELTON, PA. 687 system of drains running over the tops of tlie filters and discharging to the sewer through the overflow of each slow filter. The sand boxes have covers of i-in. checkered steel plates stiffened by angle irons riveted to the underside thereof, the cover-plates being in two halves for convenience in handling. Manhole heads and covers, 18 in. in diameter, are provided for the compartments containing the weirs, and 24-in. heads and covers for the manhole opening into the filter, except for the main entrance, for which there is a rectangular checkered steel cover, 2 by 4 ft. The sand boxes have 2-in. drain-tile connections at the bottom of each, and lead into the 4-in. tile drains between two adjacent rows of sand boxes; these connections allow the water to drain out of the sand when the box is filled from the sand-washing plant. Frost Proofing. — Around the periphery of the plant a curb, 2.5 ft. high, of ornamental design was built on top of the roofs of the filters, and the entire roof, after the pipe drains were laid, was covered with earth. Fig. 1, Plate LIII, to a depth of 2.5 ft., the upper 9 in. being of top soil. The whole area was then raked over and sowed with grass seed. Exterior Finish. — The visible exterior faces of the slow filter walls were divided into panels by pilasters moulded monolithic with the walls, the effect being carried around the roughing filters as a water-table from which to start the brickwork of the building containing the regulating rooms, storage rooms, etc. Supply and Drain Valves. — In the southwest corner of each of the three slow filters is a compartment, Plate LII, the top edge of which is level with the surface of the sand. The rough-filtered water is delivered to the filters through these compartments. Drains for emptying the water from the surface of the filters, and overflow pipes connecting with a 12-in. drain, lead to the sewer. The drains from the sand troughs on the roof discharge vertically downward into the over- flow pipe. Slow-Filter Under drains. — The underdrainage system of the slow filters consists, Plate LII, and Pig. 2, Plate LTV, of lines of 10-in. half -tiles lying at right angles to the length of the filter in the center of each panel between the rows of posts, and connecting with a main underdrain running lengthwise of the filter in one of the panels ad.iacent to the center line; a 12-in. cast-iron pipe leads from 688 WATER PUEIFICATION AT STEELTON, PA. [Papers. this central drain to the regulating house. The effluent control valves for all three filters are placed in a special house forming a part of the structure standing on the roughing filters. The half-tile drains were extended to a point about 4.5 ft. from the walls, all around, the ends being closed, and where they cross the central underdrain they were covered with concrete slabs forming part of the tight cover of the main underdrain. The vitrified hub and spigot tile drains were laid with open joints, and were surrounded with layers of broken stone, the first or bottom layer, about 6 in. thick, being composed of frag- ments of crushed sandstone the largest pieces of which were not more than 3 in. in diameter and the smallest not less than | in. in diameter; the second layer, 3 in. thick, was formed of sandstone particles from i to I in. in diameter, from which the dust had been removed and in which there were few particles larger than i in. in diameter; and the top layer, also 3 in. thick, was prepared from sandstone screenings, from which the dust had been removed and in which there were no particles larger than J in. in diameter and very few larger than i in. in diameter. The material was secured from a quarry at Marysville, on the Susquehanna, some miles above Harrisburg, and was in three sizes, I in. to 3 in., J to | in., and screenings, the latter containing- a considerable quantity of fine dust as well as particles up to f in. in diameter. By proper sorting the only material requiring a second handling was the screenings, which, on sifting through J-in. and |-in. sieves, provided for all the fine material and for nearly one-half the i to |-in. gravel required; but the weight of the wasted dust from the screenings exceeded, by approximately 25%, the weight of the useful materials secured. As received at the works, the materials were quite dirty, and a simple plan for washing out the stone dust was devised by the contractor's foreman. This consisted of a couple of oil barrels mounted tandem on horizontal axes slightly above the centers of the barrels, on a framework high enough to allow a dump-cart to stand thereunder. The two barrels were provided with a simple lever arrangement for turning them upside down. After being filled with water from a hydrant, they were suddenly upset on a cart containing, perhaps, J cu. yd. of broken stone. This quick flush effectively washed the mud from the whole depth of stones in the cart, the surplus water draining out under the tailboard. The gravel underdrains were not carried out to the side walls of the PLATE LV. PAPERS, AM. SOC. C. E. AUGUST, 1909. FUERTES ON PURIFICATION OF WATER. Fig. 1.— Slow-Filter Operating Room. FiF. 2.— Sand Washer. Papers.] WATER PUEIFIOATION AT STEELTONj PA. 689 filter, but were stopped about 2 ft. therefrom and banked up to the proper height, the finer layers of gravel covering the coarser layers on the slopes, and the top surface being brought to a uniform level. Filter Sand. — The filter sand vyas secured from the Susquehanna Eiver, and required no further preparation than screening, and, at times, the mixing of sands obtained from different localities. The specifications for the sand v^ere as foUov^s : Not more than 10% should be smaller than 0.29 mm. in diameter. Not more than 5% should be smaller than 0.24 mm. in diameter. At least 90% should be finer than 0.80 mm. in diameter. Practically all the particles should be finer than 1 mm. in diameter, and the uniformity coefficient should be not greater than 1.6 to 1.8. These specifications were written for the Susquehanna River sand with which the writer had had much experience; a modification of these requirements might be desirable for sands from other localities. As to chemical composition, the Susquehanna River sands were satisfactory, containing less than 2% of lime and magnesia, and less than 5% of an aggregate amount of iron and aluminum in forma likely to disintegrate under the action of water or chemical changes produced by matters the water might contain when passing through the sand. The analyses of the samples of sand taken after the filters were ready to be put in operation are given in Table 1. Filter No. 3 and practically all of Filter No. 2 were filled with sand obtained at Northumberland, from the West Branch of the Sus- quehanna, and required no further preparation than the careful selec- tion of the barge loads as the sand was dredged, accepting those which would fill the requirements satisfactorily, as indicated by field tests conducted as the dredging proceeded, and rejecting those in which the sand was too fine or too dirty. The only difficulty experienced with this sand was the likelihood of its containing noticeable quantities of saw-dust, shreds of bark, and friable particles, but casting it by hand over screens having J-in. meshes removed the objectionable matters to a degree which rendered the sand satisfactory for filtration purposes. Owing to the frequent changes in the conditions of flow in the Susquehanna, the character of the sand deposited is subject to varia- tion, and, before a sufficient quantity had been secured, the North- umberland deposits had become too fine for use and it was necessary 690 WATER PURIFICATION AT 8TEELT0N, PA. [Papers. to seek a coarser sand to mix with it in order to increase its effective size. An inspection of several deposits farther up the river finally disclosed one at Nesbit, from which much of the finer material had been washed, and by combining the mechanical analyses of this and the Northumberland sands, it was found that a satisfactory composite sand could be made by mixing the two in proper proportions. TABLE 1. — Mechanical Analyses of Sand in Slow Filters of Steelton Filter Plant. Note. — In this table the figures in Column 1 give the depth from which the sample was taken, those in Column 2 refer to samples collected on the northeast corner of the filter, those in Column 3 to samples of the southeast corner, those in Column 4 to samples col- lected in the center of the filter, those in Column 5 to samples collected in the northwest corner of the filter, and those in Column 6 to samples collected in the southwest corner of the filter. (I) (2) (3) (4) (s) (6) Northeast SODTHBAST Center. Northwest Southwest Corner. Corner. Corner. Corner. Depth from which sample was taken. ■§.s la 1 ■ ii •L 11 > II 1. II H W H ^8 m &8 H ^8 Filter No. 1. Top 0.5 ft... 0.5 to 1.5 ft. 1.5 to 3.0 ft. 3.0 to 4.0 ft. 0.38 1.7 0.32 1.5 0.33 17 0.31 1.5 0.31 0.30 1.8 0.30 1.6 0.31 1.7 0.31 1.6 0.31 0.39 1.9 0.32 1.5 0.32 1.5 0.30 1.6 0.32 0.31 1.5 0.33 1.3 0.29 1.7 0.31 1.6 0.82 a.o 1.9 2.0 1.8 Filter No. 2. Top 0.5 ft... 0.5 to 1.5 ft. 1.5 to 3.0 ft. 3.0 to 4.0 ft. 0.32 1.5 0.30 1.5 0.30 1.4 0.31 1.4 0.30 0.32 1.5 0.30 1.5 0.30 1.4 0.30 1.5 0.30 0.30 1.4 0.31 1.4 0.30 1.5 0.81 1.4 0.30 0.31 1.4 0.40 1.3 0.81 1.4 0.30 1.4 0.29 1.4 1.4 1.4 1.5 Filter No. 3. Top 0.5 ft.., 0.5 to 1.5 ft 1.5 to 3.0 ft 8.0 to 4.0 ft. 0.31 1.6 0.30 1.8 0.81 1.8 0.31 1.7 0.32 1.6 0.30 1.8 0.33 1.7 0.29 1.8 0.32 1.6 0.29 1.7 0.86 1.5 0.30 1.8 0.31 1.7 0.32 1.7 0.36 1.5 0.28 1.8 not taken. Papers.] WATER PURIFICATION AT STEELTON^ PA. 691 A3 an example. Table 2 exhibits the analyses of the Northumberland and Nesbit sands on July 2d, 1908, at which time the mixing of the two sands iu equal proportions produced the desired composite. TABLE 2. Sieve: Meshes, per inch. Size of separa- tion of sieves, in millimeters. Nesbit sand. Percentage passing sieve. Northumberland sand. Percentage passing sieve. Composite sand. Percentage passing sieve. 80 i.oa 0.57 0.38 0.29 0.21 100 64 8 2 0.3 100 65 23 14 0.6 100 30 65 50 13 60 7 80 0.5 Effective size 0.4 mm. 1.4 0.27 mm. 2.00 0.32 mm. Uniformity coeiii Bient 1.6 When Filter No. 1 was being filled, the Northumberland sand had an effective size of only about 0.27 or 0.28 mm. The two sands were mixed by dumping cart loads of each kind upon separate platforms on each side of a manhole, and casting simultaneously the correct relative numbers of shovelsful from each pile upon a screen standing over the manhole. In passing through the screen and falling to the filter below, the two sands became thoroughly mixed, as will be seen from an inspection of the analyses in Table 1. The sand was delivered to the filters in one-horse, two-wheeled carts. It was dumped through the manholes, and spread in three layers, care being taken not to compact it unevenly. The top sur- face. Fig. 1, Plate LIV, was leveled off with straight-edges to a practically uniform elevation, the final operation being to rsske the entire area over lightly with an iron rake. A typical mechanical analysis of the Northumberland sand, when running at its best, is given in Table 3. The Northumberland sand, during the process of dredging from the river bottom, was quite thoroughly washed, and required no further treatment except screening through J-in. mesh screens to remove mussel shells, shreds of bark, etc., before being placed in the filters. The total quantity of sand shipped to Steelton for the slow filters was 3 989 tons, in' lOY car loads, and made, in the filters, 2 953 cu. yd. In order to help out the contractor in the matter of securing this 692 WATEH PDEIFICATION AT STBBLTON, PA. [Papers. sand, the writer kept an inspector at the sand dredges, whenever sand was being secured, during the entire period covering November and December, 1907, and January, April, May, June, and July, 1908; actual dredging operations, however, were limited to 92 days during this time. TABLE 3. — Mechanical Analysis of Samples op Northumberland Sand, October 31st, 1907. Size of sieve. Meshes, per linear inch. EfEeotlve size of separation, in millimeters. Weight of sand passing sieve, in pounds. Percentage passing. 4 8.4 2.15 1.02 0.57 0.45 0.38 0.29 0.21 0.17 0.12 2.295 2.289 2.274 2.183 1.774 1.067 0.170 0.025 o.oia 0.005 100 10 99.7 20 99 30 95 40 76 50 46 60 7 4 80 1.1 100 0.6 150 0.0 Note.— EfEeclive size, 0.31 mm.; uniformity ooeflHcient, 1.37: finer than 0.24 mm., 2.5%; finer than 1 mm., 9iX; finer than 1.5 mm., 99.5 per cent. The total cost of inspecting the sand, including the salary of the inspector, his board, and traveling expenses, was 9.5 cents per ton, or 13.0 cents per eu. yd. As the sand was running, a satisfactory indi- cation of its character was obtained by a mechanical analysis of one sample to about each 30 cu. yd. dredged. The sand was sampled on the barge before unloading, each sample being a composite of samples collected at about four places from freshly excavated vertical faces in the sand as piled on the barge, and collected so as to be representative of the material for the full depth of the pile. These samples were then thoroughly mixed, the sample for analysis being taken therefrom, dried, and sifted through carefully rated screens. Sand obtained in the river deposits in the neighborhood of Harris- burg could not have been used without expensive preparatory treat- ment to remove the fine material, as its effective size usually runs from about 0.23 to 0.24 mm. To remove some of the fine material, washing is necessary, the total waste ranging from about 10% to more than 50%, depending largely on the quantity of coal dust in the sand. The sand from Northumberland, being obtained from the West Branch above the junction of the North Branch, is free from coal, and is also coarser than that obtained farther down the river. Papers.] WATEE PURIFICATION AT STBBLTON, PA. 693 Effluent Control of Slow Filters. — The effluent pipes pass beneath the floors of the slow filters, and each terminates in its own regulating chamber with a sluice-gate controlled from a valve-stand and hand- wheel on the floor of the operating room. The rate of filtration of each filter is controlled by causing the filtered water to flow through a submerged orifice. Fig. 4, 8 in. in diameter in the end of an outlet pipe having a swivel-joint at the lower end. Plates XLVII and XL VIII, and discharging through a gate-valve into a filtered-water well common to the three filters. The free or orifice end of the swivel- REGULATING ORIFICE AND FLOAT FOR SLOW-FILTER EFFLUENT PIPES Fis. 4. jointed pipe is suspended by brass links from a float resting on the surface of the filtered water in the effluent chamber of the filter; the center of the 8-in. orifice is adjusted to stand 1 ft. 9 in. below the surface of the water in the effluent well, at which depth, with a free discharge through the orifice into the atmosphere, the filter will deliver 1 500 000 gal. per day to the outer well. If the pumps in the Borough pumping station do not take the water away as fast as the three filters can furnish it, the water rises in the outer well, submerges the orifices, and cuts down the head by an amount just sufficient to keep up the supply, the plant being capable, therefore, of yielding the water to the 694 WATER PURIFICATION AT STEELTON, PA. [Papers. pumping station automatically at any rate neqessary, up to the maxi- mum for which the plant is designed, namely, 1 500 000 gal. per day per filter. In order to prevent any draft-tube effect in the effluent pipe, between the effluent well and the outer filtered-water well, a 2-in. galvanized-iron pipe is connected with the swivel-jointed pipe at a point below the orifice, the free end of the 2-in. pipe being adjusted so as to be always above the water level in the effluent well, whatever position the float may take; this assures a discharge into the atmos- phere whenever the difference in level of the water between the effluent well and the filtered-water well of any filter exceeds 1 ft. 9 in. Gauge Boards. — The rate of filtration and loss of head are indi- cated on scale boards, Fig. 1, Plate LV, carried by bronze wires attached to floats in the different compartments of the regulating house, as described for the roughing filters. A sampling pump, similar to those for the roughing filters and the raw water, is provided for the effluent well of each slow filter. The filtered water leaves the central filtered-water well through a 24-in. wood-stave pipe leading to the new pump-well at the pumping station, a sluice-gate being provided at each end of this pipe line. Sand Washer. — The system of handling the sand when the slow filters require cleaning is not new in principle, but the details differ materially from those of other plants. Portable ejectors, supplied with water at about 100-lb. pressure from a line of 4-in. wrought-iron pipe, Plate LII and Fig. 2, Plate LIII, suspended from the ceiling of each slow filter, through 3-in. bronze hose-gates and 3-in. hose, lift the dirty sand and transport it to the sand-washer plant. The dirty sand, having been shoveled into piles in the filter, is cast into the hopper of the portable ejector, is picked up by the jet of water passing through the ejector, and delivered through a line of 4-in. wrought-iron pipe to the sand-washing hoppers, Plate XL VIII, and Fig. 2, Plate LV, in the second story of the slow-filter regulating house. After passing through two hoppers the washed sand is forced by a jet in the bot- tom of the second hopper through a line of 4-in. wrought-iron pipe, rest- ing on the roof of the three filters, to hose-valves from which lines of hose may be laid to discharge the returning clean sand into the compartments of the sand troughs running lengthwise of the tops of the filters. The operation is simple, and requires no special description or comment. Heating. — The building covering the operating rooms of the rough- Papers.] WATEE PURIFICATION AT STEBLTOK^ PA. 695 ing and slow filters, as wall as tlie machinery room and the coagulant- storage and sand-washer rooms, is heated by steam generated in a vertical tubular boiler standing in the second story of the slow-filter regulating house. The returns from the radiators are piped back to a trap in one corner of the slow-filter operating room, and discharge into the sewer. The provisions for heating required boiler capacity with sufficient heating surface and grate area to furnish steam to raise the temperature of 20 000 lb. of coagulant solution from 32° to 70° Fahr., in 1 hour. The boiler is designed to carry 100 lb. of steam, and is provided with a reducing valve, on the steam-heating connec- tion, set at 5 lb. The coils and radiators are proportioned to maintain the machinery and operating rooms at a temperature of Y0° and the coagulant-storage room and the sand-washer room at a temperature of 58° when the outdoor temperature is at zero. The toilet-room is pro- vided with a wash-basin with cold-water connection and a steam con- nection for securing hot water, and a low-dovyn tank closet. OONSTEDCTION. No special difficulties were encountered during construction, the bearing value of the foundations proving satisfactory for the relatively light loads they are required to carry. There was comparatively little ground-water in any of the work, although the permanent ground-water level was barely below the floor of the excavation for the slow filters; in fact, a modification of the design of the main underdrains of two of the filters was made necessary in order to keep the work out of the water. A considerable portion of the site of the slow filters was, in early days, a low swale which had been used subsequently as a dump- ing ground for slag, cinders, and ashes from the steel works, and it was owing to the presence of these materials, which had had such a long time to become thoroughly settled, that satisfactory foundations were secured without extra work. Leakage. — Before the sand was placed in the slow filters they were filled with water from the street mains up to the level required for opreration, the water from the street being then shut off and the drop in the surface level being observed. It was not expected that the filters would be perfectly tight, owing to the construction of the bottoms in blocks about 11 ft. square, with no special provision for stopping leakage between the blocks. The leakage from each of the 696 WATER PURIFICATION AT STEELTON, PA. [Papers. three filters was about the same, and aggregated about 0.66 of 1% of the nominal daily capacity of the plant. No tests to determine the amount of leakage have been made since the filters have been in operation, but it is believed that there has not been an increase. . Cost of the Plant. — The plans were completed in July, 1907, and, in response to advertisements, three bids were opened on August 7th, 1907, aggregating $71522.85, $80 482.48, and- $89 592.59, respectively. The contract, which was awarded August 14th, 1907, to the Bunting Construction Company, of Tlushing, N. T., stipulated that the work should be completed within 6 months from the date of the contract, or by February 14th, 1908. Various delays resulted, however, and it was not until the middle of September that the plant could be put in operation, the final estimate to the contractor being dated September 24th, 1908, and amounting to $70 730.87. This, however, did not repre- sent the full cost of the work, as the installation of the centrifugal pumps, the changing of the suction pipes at the pumping station, and some repair work on the checkered steel covers on the slow filters, were done by the Water Board outside of the contract. In addition to the above items, there were the cost of the site and of engineering and incidental expenses, which were high, by reason of the fact that the time occupied by the contractor in the construction of the plant was more than twice that in which the work should have been completed under favorable conditions. The following are the principal items miaking up the cost of the plant : Excavation, grading, roadways, sodding, etc.. $3 936.88 Sewers and drains 2 314.67 Concrete, all classes 18 708.32 Eeinforcing steel 3 475.80 Manhole heads and covers, indoors and out- doors 1 590.50 Cast-iron pipes and specials 3 956.27 Wood-stave pipes between filter plant and pumping station 9 034.21 Filter regulating devices, gauges, coagulant and lime-water apparatus 2 650.00 Superstructure over regulating houses, etc., including steam heating, lighting, and plumbing 7 900.00 Papers.] WATER PURIFICATION AT STEELTON, PA. 697 Machinery, plant, motors, compressor, shafting, belting, air receiver, sand- washing plant, etc 3 750.00 Gate and sluice valves 1 485.00 Filtering materials, including underdrains . . . 10 811.34 Extra work 1 117.88 $70 730.87 Of the above amount, $57 000.00 is the cost of the purification works, the remainder, $13 700.00, being the cost of the pipes to the pumping station, the sewer, to the river and other details growing out of local conditions. Operation. The plant was put in service on September 12th, 1908, but was shut down later in the day, and was started again on September 16th, since which date it has furnished filtered water continuously. The regular force required for operation consists of a general super- intendent and two filter attendants, the latter working on 12-hour shifts. When the slow filters require scraping an additional force is taken on temporarily, the men working under the direction of the general superintendent. The permit for the construction of the plant, issued by the Pennsylvania State Department of Health, provided that the filters should be operated for a considerable period under the direction of its designer, and in compliance with this condition the writer instructed the superintendent and attendants in their duties, provided specific instructions regarding the manipulation of the roughing and slow filters, and the handling of the coagulant, and has advised with refer- ence to special features of the operation whenever necessary. From September 16th, 1908, until January 9th, 1909, no coagulant was used in connection with the operation of the roughing filters, the raw water being sufficiently free from turbidity to yield a satisfactory water without coagulant. Quantity of Coagulant Bequired. — The general instructions given to the superintendent called for the application of a coagulant to the raw water whenever its turbidity exceeded 50 parts per million, and the continuance of its use for much lower turbidities whenever the 698 WATER PURIFICATION AT STEBLTON, PA. [Papers. last bacterial analyses showed 5 000 or more bacteria per cu. cm. in the raw water, or in case the effluent from the slow filters showed objectionable color. It is the intention, in dosing with coagulant, to use a sufficient quantity to produce an effluent from the roughing filters having a turbidity of not more than 10% of that of the applied water, and not, in any case, more than 25 parts per million, regardless of the turbidity of the raw water. It has also been found desirable, particularly during the winter and the cold spring weather, to manipu- late the dose of coagulant so as to produce an effluent from the rough- 3.0 2.5 2.0 2 1-5 1.0 a ■a 1 mill 1 1 1 1 1 1 1 1 1 II 1 1 1 1 1 1 1 1 PROVISIONAL CURVES FOR COAGULANT DOSING / ^ / ^ y ^ f X / y ^ ^ y y y ^ ^ / y ^"^ y £0' ^ .'"►' > r ■''^. y' ^% \«" r" y^ ^ i " ,. -■0 'r^ r -! r-. ^^ ,1^^-^ •'^ y •> ' ^ p y^ ^^^ ' ^ / »"> ' r / ^ / ^ ^ ' ^ f^ ^ /■ ■' ,' ,' 1 "S ss Sgg sss S \ 2 ; \ \ 1 ' 1 V' \ i: 1 • 1 1 1 1 3 e. H 1 1 i M SO 0.5 Turbidities of Raw Water Fio. 5. ing filters having a turbidity of practically zero, as by this means the runs of the slow filters are very greatly lengthened and the cost of operation is not increased by the use of the comparatively small additional amount of coagulant necessary to secure this result. The quantity of coagulant to be used is determined by the super- intendent from a diagram, Fig. 5, prepared by the writer from data secured origijially during the operation of the Harrisburg Testing Station, in 1902 and 1903, but modified somewhat to take into account the difference in the character of the filtering materials, and the slightly different character of the raw water at the Steelton intake; the Papers.] WATER PURIFICATION AT STEELTON, PA. 699 diagram may require further modification as additional knowledge is gained during the operation of the plant. The dose of coagulant indi- cated by the lowest curve is for application during the summer when the turbidities do not rise very high, or very suddenly. The dosing indicated by the middle curve is required when the turbidity is rising rapidly, during the first parts of the floods in the fall, winter and spring; the dosing indicated by the upper curve is required when the turbidity is falling, after the passage of a flood. It will be noticed that for equal turbidities more coagulant is required when the turbidi- ties are falling than when they are rising, on account of the finer character of the turbidity and the persistence of high numbers of bacteria in the Susquehanna after the turbidities have begun to fall. The superintendent must use judgment in determining the proper dose of coagulant. The character of the turbidity of the river water varies so greatly and changes so quickly that a dose satisfactory with a given clay turbidity, say, of 500 parts per million, might prove alto- gether too much for an equal turbidity from the North Branch. At times, also, particularly after a protracted season of low water, a turbidity caused largely by the fine shreds of vegetation torn and scoured loose from the rocky river bottom is particularly difScult to handle by reason of its high clogging value when it becomes matted upon the filter surface. In order to keep the plant in efficient opera- tion, therefore, the superintendent must watch the effect of his dosing and reduce the quantities if he finds the loss of head increasing on the, roughing filters too rapidly, or increase the quantity if the condi- tions require it. When to Wash the Roughing Filters. — If, when coagulant is being applied to the raw water in the proper amount, the effluent of a roughing filter shows objectionable turbidity, indicating the passage of coagulated material through the filter bed, the filter is immediately washed; otherwise, it is left in operation until the maximum allowed loss of head, about 2.5 ft., is reached. This method of operation is based on the theory, which seems to be supported by considerable evidence gained both at this plant and at the Harrisburg Testing Station, that these coarse-grained filters act very much as settling basins with bottom areas of great extent. The mud carried into the filter by the entering water is deposited on the granules of the filter bed through practically its entire depth, and one 700 WATER PURIFICATION AT STEELTON, PA. [Papers. of the requirements in controlling their operation is to prevent the washing of this sediment through the filters by pushing their operation too rapidly, or by using too great losses of head, or by suddenly increasing the rates of filtration. To, prevent most of these evils the mechanisms at the plant limit the loss of head to about 2.5 ft. Regulation of Coagulant Dosing. — The quantity of solution re- quired to supply a given dose of coagulant depends on the quantity of water being filtered and the percentage strength of the solution. COAGULANT ORIFICE DIAGRAM Grains per gallon of Aluminmn Sulphate Directions: Find intersection of water consumption and required dose of coag. of given ^ strength of solution, then follow vertically down to intersection wifli one of inclined lines and read head required on scale Co left. Fio. 6. Knowing these, the requisite quantity of solution is measured out by continuous discharge through orifices in the glass plate forming the front of the measuring box, the size and number of orifices, as well as the head required to give the necessary dose, in grains per gallon, being ascertained by the superintendent from a diagram. Fig. 6, prepared for his use. The lines indicating the orifices to be used were placed on the diagram in accordance with the measured discharge of each orifice. Use of Lime-Water, — During floods the alkalinity of the river water Papers.] WATER PUEIFICATION AT STEELTON, PA. 701 falls sometimes too low to permit of the decomposition. of the requisite dose of aluminum sulphate, and provisions are made to add, at such times, sufficient milk of lime to supply the deficiency in alkalinity. When lime must be used, enough is added to neutralize the free CO^ in the river water, combine with the aluminum sulphate, and leave a residual alkalinity in the water of about 6 parts per million. The free COj and alkalinity are determined by the methods recommended by the Committee on Standard Methods of Water Analysis, of the Ameri- can Public Health Association. Knowing the dose of aluminum sul- phate being used, and having ascertained the alkalinity of the raw water, and the amount of free CO^ carried, reference to a diagram including all these data will give the quantity of lime required, if any. JRegulation of Lime Dosing. — The quantity of water required to add this dose of lime to the raw water in a saturated solution is regulated at the lime-water measuring box in the manner in which the coagulant dose is regulated, a diagram similar to the coagulant- dosing diagram being supplied for the lime-water. The amount of lime to be slaked daily to produce the required quantity of lime-water, for different dosings and different daily consumptions of water, is determined by the superintendent from another diagram, small quanti- ties being slaked from time to time during the day and discharged as milk of lime through the funnel and pipe leading to the bottom of the lime-mixing tanks. The strength of the lime-water solution is to be tested from time to time by the usual simple chemical test, and it should be kept up to full strength by the attendant. Another diagram gives the number of pounds of aluminum sulphate required for making solutions of different percentage strength, the capacity of the tanks having been calibrated for the purpose. Simplicity of the Dosing System. — ^No calculations being required on the part of the superintendent during the operation of the plant, the rapid changing of the dosing, when required, can be effected with accuracy and without the likelihood of errors from faulty calculations. The lime-water and the coagulant solution are admitted to the water practically together. The conditions requiring the use of lime being comparatively rare no great pains have been taken to go into refinements in the use of lime that would be desirable under conditions requiring its more continuous use. 703 WATER PURIFICATION AT STEELTON, PA. [Papers. Reduction of Alkalinity hy Coagulant. — As to the proportion of the coagulant rendered inactive through its absorption by suspended mat- ters in the water, sufficient data have not yet been obtained at the Steelton plant to speak authoritatively, but it is believed, from present indications, that the reduction of alkalinity of the rav? v^ater, instead of th6 theoretical 8.2, will probably be about 6 parts per million, per grain of coagulant used, as observed at the Harrisburg Filter Plant during the past three years and at the Harrisburg Testing Station in 1902 and 1903. Operation of Eoughing Filters. Operation with Respect to Lengths of Runs. — ^While the roughing filters must be operated in a manner to secure an effluent sufficiently free from turbidity and bacteria to be handled satisfactorily by the slow filters, their operation must be controlled in a manner to secure sufficiently long periods of operation between washings, otherwise the plant could not be kept going during periods of very turbid water. A little experience is required to enable this to be done, but it presents no practical difficulty with any range of conditions thus far encountered . on the Susquehanna at Harrisburg and Steelton. The highest tur- bidity experienced since the plant has been in operation occurred in the forenoon of January 26th, reaching 1 600 parts per million, remain- ing at that figure for about 5 hours, and dropping by steps to 500 parts during the next 12 hours. The shortest run during this period was about 2i hours, which occurred after the turbidity had dropped to about 700 parts. The first turbid water the plant was required to handle occurred about January 8th, at a time when the writer was not able to be present, and, owing to the deterioration of certain of the silica-turbidity standards, the turbidities of the raw water and of the effluents from the roughing filters were incorrectly read, the actual turbidities being much higher than the figures recorded. As a result, the quantity of coagulant used was entirely too small, and a con- siderable number of bacteria passed through both the roughing filters and the final filters. Records of Operation.— Since January 23d, 1909, with but very few exceptions, the coagulant has been properly applied, and the satis- factory results obtained in the removal of turbidity are exhibited in Table 4, which gives the hourly records of the operation of Eoughing Filter No. 1, through runs Nos. 58 to 81, January 26th to 31st, 1909, inclusive. Papers.] WATEE PUEIFIOATION AT STEBLTONj PA. 703 TABLE 4. — Hourly Eecords or Operation of Eoughing Filter No. 1 Through Euns Nos. 58 to 81, January 26th to 31st, 1909. Len(}ths of Date. Turbidities, in parts RDN. PER MILLION. Number Coagulant used, in of run. Hours. Minutes. Day. Hour. Eaw water. Eough- flltered Percentage of grains per gallon. water. remoTal. 58 9 58 Jan. 26 6 A.M. 7 " 8 " 80 800 800 98.8 99.9 99.9 0.42 0.42 0.92 B9 4 16. 9 " 10 " 11 " 12 M. 1P.M. 800 1600 1600 1600 1600 99.9 99.9 99.9 99.9 99.9 1.81 8.22 2.22 2.22 2.22 60 3 02 2 " 3 " 4 " 160O 900 900 99.9 99.9 99.9 2.22 2.00 2.00 61 a 31 5 " 6 " 7 " 900 900 800 99.9 99.9 99.9 2.00 2.00 2.00 62 2 26 8 " 9 " 700 700 99.8 99.8 1.50 1.50 63 4 35 10 " 11 '■ 12 " 700 700 500 99.8 99.8 99.8 1.50 1.00 0.75 Jan. 27 1A.M. 500 99.8 0.75 2 " 500 2 99.6 0.75 64 2 58 3 " 500 2 99.6 0.75 4 " 500 2 99.6 0.75 5 " BOO 1 99.8 0.75 65 2 11 6 " 500 1 99.8 0.75 7 " 600 2 99.7 0.63 8 " 750 2 99.7 0.63 66 3 18 9 " 800 3 99.6 0.63 10 " 800 2 99.8 0.63 11 " 750 2 99.7 0.63 12 M. 750 2 99.7 0.63 67 3 17 1P.M. 700 2 99.7 0.63 a •' 600 3 99.5 0.68 3 " 550 2 99.6 0.63 68 4 15 4 " 550 2 99.6 0.63 5 " 550 2 99.6 0.63 6 " 550 3 99.4 0.63 , 7 " 550 3 99.4 0.63 69 3 53 8 " 550 3 99.4 0.63 9 " 550 3 99.4 0.63 10 " 550 3 99.4 0.80 11 " 500 10 98.0 1.95 12 " 500 2 99.6 1.95 70 3 00 Jan. 28 1A.M. .100 1 99.8 1.50 2 " 500 ] 99.8 1.50 3 " 50O 1 99.8 1.50 71 4 03 4 " •■iOO 1 99.8 1.50 5 " 500 1 . 99.8 1.50 6 " 500 1 99.8 1.50 7 " 500 1 99.8 1.30 704 WATER PURIFICATIOHr AT STEELTON, PA. TABLE i.— (Continued.) [Papers. Lengths op RUN. Number of run. Hours. Minutes. 7S 75 77 78 10 17 04 56 13 19 50 05 59 Date. Day. Jan. 26. Jan. 00 Jan. Hour. 8 A.M. 9 " 10 " 11 " 12 M. ) P.M. a " 3 " 4 " 5 " 6 " 7 " 11 " 18 " 1A.M. 2 " 3 " 4 " 5 '■ 6 " 7 " 10 " 11 " 12 M. 1P.M. TUKBIDITIES, IN PARTS PER MILLION. Baw water. 9 " 10 " 11 " 12 " 1A.M. 7 " 8 " 9 " 10 " 11 " 12 M. 1P.M. 2 " 3 " 500 500 500 500 450 500 500 500 500 sop 500 500 450 450 450 450 450 450 450 450 450 450 450 400 400 300 300 300 300 300 275 275 275 275 250 250 250 250 250 250 850 250 250 250 250 250 200 200 Kough- flltered water. Percentage of removal. 99.8 99.8 99.8 99.8 99.7 9.8 9.8 9.8 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 99.7 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 lOo 100 Coagulant used, in grains per gallon. 1.30 1.30 1.30 1.22 1.22 1.22 1.82 1.22 1.22 1.22 1.22 1.30 1.30 1.30 1.30 1.30 1.30 1.30 1.80 1.40 1.40 1.40 1.40 1.40 1.40 1.40 1.30 1.30 1.30 1.30 1.30 1.20 1.20 1.80 1.12 1.12 1.12 1.12 1.12 1.18 1.18 1.12 1.18 1.18 1.12 1.12 1.12 1.12 1.12 1.12 1.18 1.12 1.12 1.12 1.12 0.99 Papers.] WATER PURIFICATION AT STBELTON, PA. 705 TABLE i.— (Continued.) Lengths of Turbidities, in pabts BUW. UAIIS, PER MILLION. Number Coagulant used, in of run. Hours. Minutes. Day. Hour. Raw •water. Eough- flltered water. Percentage of removal. grains per gallon. 80 17 00 Jan. 80 4 P. M. 200 100 0.99 5 " 200 100 0.99 6 " 200 100 0.99 7 " 200 100 0.99 8 " 200 100 0.99 9 " 200 100 0.99 10 " 200 100 0.99 11 " 200 100 0.99 13 " 200 100 0.99 1A.M. 200 100 0.99 a " 200 100 0.99 S " 200 100 0.99 4. " 150 100 0.99 5 " 150 100 0.99 81 13 40 6 " 150 100 0.99 7 " 150 100 0.99 8 " 100 100 0.63 9 " 100 100 0.63 10 " 100 100 0.63 11 " 100 100 0.63 12 M. 75 100 0.55 1 P.M. 75 100 0.55. 2 " 75 100 0.55 Four of the runs included in Table 4 were too short, and, with the experience since gained in handling the plant, particularly the coagulant, could have been extended by an hour or two without difficulty. The complete records of the operation of Roughing Filter No. 1 by runs, from September 12th, 1908, to February 3d, 1909, are given in Table 5. Analyses. — There is no laboratory for chemical and bacteriological work at the plant, but an arrangement has been made by which samples of the raw water, the rough-filtered water, the final-filtered water, and water secured from a faucet in the borough, are collected and examined by the cSemist and bacteriologist of the Harrisburg Filter Plant once each week, and at such other times as desired. Two plates, for total numbers of bacteria, are prepared from each sample, and presumptive tests are made for the presence of B. Ooli in the raw- and filtered-water samples, five 1-cu. cm. sowings usually being made of each sample; the tests for turbidity, free COj, and alkalinity are made at the plant by the superintendent, who is a skilled chemist. 706 WATER P DEIFICATION AT STEBLTON, PA. [Papers. SSgggSSgSSSS3§S!Sg82ggS:SfeS!!&fe!fel&SgSS5g Run No. 1 .............. ..................^ ^ 1 gs;:5SS§8gSSgSS£SKgS&£8SSgfeg£SggCSSSS i CO O 3 M Average rate of operation, in gallons daily; - SD Vj CO W '<( 50 1^ '#. O 10 -5 Vl Ol O V ■-^-JCOOMOtOOO-^eDWaDeDCDOM Millions of gallons per acre per day. 1& )^ »^ t^ tCb tC*C0K>M»MI-'i-i|-i|-i.Mi-'K)i-'Ci5tfr.ie.^(f>.U>J oooooooooooooooooooooooooo'oo'o'ooooo ■«1 (6 2 g: 1^ ^: :: g: ::::::::::: : B»^ s <" s iOOOOOi-»_MN)tOtOOOOO' ooooooooooooooooo- ias'oscsas-5e2WQOQ*.iOMK-*. io'co03t^u'«o'eocD'co' h- 1 I— ' hJi 1^ hJ- to to to to I-* >-^ OOOOOOcDtO«OCOQOaoao-Q-^0309»-iCC)»^tOOOOlt^aDi^ 303»ODCocococn t-l CO o a o o IT) w OTC0 4^O.WM*.C=4^KlcO,»5SSS?8g2ggSisSSg?Si8&li CD a rn r" o 3 00000000000000000000000000000000000 w m eg w CT OS m w 52 m *; jfe; OS ►&. 52 01 s w m tP-owtfr-M-^o-^wtoaDODotooDwcnoiuiooooooaooait— wi^EncacnwcoOT QpO0Q0QOa0QOO0QOQOGO-3Q00POOO0-CROO^COeOCOQOOOQD*QOO-i^ -I o 'm^ m *»^ -J *co o rf*. *i-i h-. ba *-j w 03 OS o o ►-i o bo o CO ^ 10 is CO V^ Millions c gallons p acre per a. »^l^»(^j£k«.c;TCn»C>.>f:>.»C>.COeOtO^ ^ w ^^ QoceDSOTaaceo9i-^eot^a3aoooi(^i^>-i)-^i-^coceco)-AMOoooooooooo< O 00 00 000 "00 o'o "00 coo 00000 00 00 00 000 03 01 • ^cc^«0cn-^a:-^oaoooaooQooooc 5os-^o:asowoswtooo5-^ooooc a»2 J^' fflMjfe- IO>-^ 1^ s ■<1 CD 2 CD CJ p O p _0 p p p ^t- ; oscececoososAco- Coagulant, f rains per g on (averag 708 WATER PUEIFICATION AT STEBLTON, PA. [FaperB. CO CO CO CO lA 10 %a COMlc»»-i»-'i-kOOcOtt5eo 3 i OC&COOCO'^OODOC1Ci5C7li4^i;k.ri^Ci9 '^•4o5aoooeocnoeDeocera»4^coc3 oooooooooooooooo '^ £0 1-1 M to to so SOOOQ'OQOOOOSOO Sooooooooooooo -ao-JcjicD»oia^oo(OOiM- cncn m ►-j- ocngs-2offiw^®QofWooQo ^oocnoooooooooooo tOOOOOOOOOOi-n-'MMi-'-H- o o o b b b b b b b b b b b b b CDOOOOOOOOOCOCOOCOeOeO • O O O O »-*»-' M. M (-1. 1- t-l l-l l-l l-l M 1^ Millions of gallons per acre per day. Si C3 fla ^ 5»S Ma a M ■ a D a H i| Coagulant, in grains per gal- lon (average). o I Papers.] WATEK PDHIFIOATION AT STEELTON, PA. 709 Air Wash. — The roughing filters, when clogged, or when not yield- ing a satisfactory effluent, are cleaned by washing with air and water, first shutting off the raw-water supply and then, if pushed for time, opening the sewer gate and wasting the water upon the surface of the filter down to the level of the wash-water troughs. The filter, in the meantime, being in operation, is allowed to run until the surface of the water on it has dropped about 6 in. below the edges of the wash- water troughs; the effluent gate is then closed, and compressed air is discharged into the underdrains from the air receiver. The compressed air is stored in the receiver at a pressure of approximately 100 lb. per sq. in. at the beginning of the wash, and is discharged through a reducing valve set at about 10 lb. on the filter side of the valve ; this, however, does not represent the pressure at which the air is applied during the wash, as the 10 lb. is largely used up in friction through the small air pipes leading to the filters. The air issues into the filter through the orifices in the bottom of the 2-in. galvanized-iron underdrain pipes 1 in. above the floor of the filter. The bottoms of the pipes lying all in the same plane, the air issues from the orifices only as it is forced out by the pressure in the delivery pipe, its tendency to issue by its own buoyancy being overcomes by placing the orifices in the bottoms of the pipes; within limits, therefore, it is possible to deliver to the filter any quantity of air desired, and to have it issue through all the orifices in the underdrain pipes with practical uni- formity. The application of the air is continued until the contents of the receiver, approximating 1 500 cu. ft. of free air, is discharged, which takes from 4 to 5 min., as the reducing valve is now set, and this corresponds to approximately 1 cu. ft. of free air per square foot of filter surface per minute. Apparently, this is sufficient, for, as far as can be ascertained, there has been no tendency for suspended mat- ters to accumulate in the filter permanently. The objection to the use of a large quantity of air, or to very heavy air washing, is the likeli- hood of the disturbance of the gravel underdrains to be followed by the loss of some of the filtering material ; and the tendency to aid in produc- ing vertical stratification in the filter, in connection with the water-wash, allowing some portions to clog up completely, and increasing the velocity of filtration through the more porous parts to such an extent as to render the filter ineffective. In order to guard against this tendency, the superintendent, from time to time, stirs up by hand the 710 WATER PUEIFICATION AT STEELTON, PA. [Papers. filtering materials in the roughing filters, using a rake while the wash- water is being applied; this "dissipates effectively any tendency of the materials to collect in vertical strata, and maintains the porosity and eflSciency of the filter. Water-Wash. — After the air is turned off the filters are washed with filtered water from the street mains. The water is applied through the underdrains, and the dirty water overflows into the wash-water troughs and passes thence to the river through the sewer. The wash-water is ordinarily applied at a rate equivalent to about three and one-half times the rate of filtration, or between 8 and 9 vertical inches in the filter per minute, the duration of the water-wash being from 5 to 6 min., and the quantity of water used per wash about 10 000 gal. In addition to that used for washing, about 5 000 gal. are wasted froin the surface of the filter at each wash when the turbidity is high and the filters have to be washed frequently. It cannot, as yet, be told with accuracy just what percentage of wash-water will be required during the year, but a prediction of the quantity can be made from data secured at the Harrisburg Testing Station, and at the Harrisburg Filter Plant. The number of washes required will depend on the turbidity of the raw water, as this causes the clogging of the fi[lter and necessitates the use of a coagulant. TAELE 6.^ — Effect of Turbidity op Eaw Water on Lengths op Runs OF EouGHiNQ Filters at Steelton. Turbidity of applied water, in parts Number of days PER MILLION. per year when such turbidities may be expected to Average lengths Range. Average. prevail. (I) (2) (3) (4) 0- 50 25 160 6 days. 51- 75 62 65 2* hours. 76- IOC 88 85 17 " 101- 150 125 29 12 " 151- 200 175 as 10 " 201- 250 225 18 9 " 251- 500 875 25 7 " 501- 800 650 7 5 " 801-1200 loon 5 4 " 1200+ 1400 3 8 " Lengths of Runs. — The data thus far secured from the Steelton plant, in addition to information from the Harrisburg Testing Station, indicate that the lengths of runs from the roughing filters, when operat- Papers.] WATER PURIPIOATION AT STEELTON. PA. 711 ing at a rate of from Y5 000 000 to 125 000 000 gal. per acre per day, would be about as shown in Table 6. Column 1 indicates the range of turbidity of the raw water. Column 2 the average turbidity. Column 3 the approximate number of days per year when such turbidities may be expected to prevail, and Column 4 the lengths of runs of the .rough- ing filters. Using the data in Table 6, 428 washes per year per filter will be required, calling for a total of about 6 YOO 000 gal. of wash- and waste- water, assuming that the water on the surface of the roughing filters is wasted at each wash. As it takes about 15 min. to wash a filter, the necessary lost time for the 428 washes would be 4J days per year, and the output of each filter for the 360.5 days of operation, at its full rated capacity, 1500 000 gal. per day, would be 540 500 000 gal.; the wash- and waste- water, on this basis, would represent 1.2% of the n'et output of the filter for the year. The longest run yet experienced was about 24 days, and the shortest 2 hours 11 min. It is believed that with more experience the shortest runs need not be less than about 4 hours. TABLE 7. — Average Dose of Ooaqulant. Tdrbidity of applied water, in parts per million. Numtier of days per year when sucn tur- bidities may be .expected to prevail. (3) Average dose of aluminum sul- phate, in grains per gallon. (4) Average dose of lime, in Range. (■) Average. (2) gallon. (S) 0- 50 51- 75 76- 100 101- 150 151- 300 301- 250 251- 500 501- 800 801-1 300 1300+ 35 63 88 135 175 325 375 650 1000 1400 160 65 35 39 38 18 35 7 5 3 0.0 0.4 0.5 0.6 0.7 0.8 1.0 1.4 1.7 3.1 0.25 0.40 0.60 0.75 Average Dose of Coagulant. — From the data thus far secured, it would appear that the average quantities of coagulant necessary in the operation of the plant would be about as shown in Table 7, from which a calculation indicates that the average dose required through a series of years would be equivalent to 0.37 gr. per gal., or 53 lb. per million gallons of water treated, which, at $20 per ton, gives the average cost for coagulant as 53 cents per million gallons. For the 712 WATER PUHIFICATION AT STEELTON, PA. [Papers. first year the quantity required will fall far short of this figure, the average dose from September 16th, 1908, to April 18th, 1909, having been but 0.22 gr. per gal. Effect of Turbidity on Initial Loss of Head.— The initial loss of head, at the commencement of operation of a roughing filter, is not affected either by the turbidity of the raw water or the quantity of aluminum sulphate used, the effects of these two factors being evident only in the rate at which the loss of head increases, and consequently in the lengths of the runs. Effect of Bate on Quantity Filtered. — The rate of filtration affects only slightly the total quantity of water that may be filtered between washings of the roughing filters; whatever slight effect there may be apparently indicates that somewhat larger quantities can be filtered between washings at high rates than at low rates, but, of course, at the expense of somewhat more frequent washings. Operation of Slow Filters. The slow filters are operated on the continuous plan, the rough- filtered water being controlled so as to stand about 2 ft. deep on the surface of the slow filters. The method of regulating the rate of filtration need not be again described, it being sufficient to state that the yield of the slow filters is responsive to the draft of the municipal high-pressure pumps, up to the limit of the discharging capacity of the regulating devices on the filters. As filtration proceeds, the slow filters gradually clog up, and the loss of head increases; provision is made to allow for a total loss of head of 7 ft. in the slow filters, which is the full depth of the filtering materials, and the superincumbent water. Daily Records. — The slow filters were put in operation on Septem- ber 16th, 1908. The daily records of the operation of Filter No. 1 are given in Table 8, from which it will be seen that the accumulation of the loss of head proceeded with comparative slowness during Septem- ber, October, and November, while during December the occasional increase in turbidity of the raw water began to have its effect on clogging, bringing the total loss of head up to about 2 ft. on the last day of December. In January, 1909, during which there were two periods of bad water, the loss of' head increased to about 6 ft., a gain of 4 ft., during the month, and it became necessary to start scraping. Papers.] WATER PUKIFIOATION AT STEELTON, PA. 713 On the morning of January 30tli, therefore, the supply to Filter No. 1 was shut off, the water was drained down to the surface of the filter, and filtration was continued for a short time to allow the water to fall a sufficient distance below the sand surface. TABLE 8.— E ECORD OF OpEI 16th, 1908, lATioN OF Slow Filter No. TO February 1st, 1909. 1, September Time in service: Days. Hours. (2) Loss of head, in feet. (3) Rate of Operation: Date. Gallons daily. (4) MiUions of gallons per acre per day. (S) Sept. 16 'i :: 15 18 23 24 28 31 37 42 45 49 59 70 75 76 80 83 86 92 98 101 102 106 107 109 110 111 115 118 121 124 128 131 133 1.34 135 135 7 0.38 0.36 0.33 0.35 0.45 0.30 0.49 0.35 0.40 0.35 0.30 0.40 0.63 0.61 0.75 0.75 0.85 1.70 1.50 1.20 1.60 1.33 1.95 2.07 2.09 2.20 2.35 2.42 2.92 3.70 3.74 4.23 4.12 5.01 5.35 5.99 5.75 6.01 680 ono 650 000 610 000 600 000 600 000 520 000 510 000 500 000 530 000 630 000 520 000 580 000 500 000 500 000 500 000 500 000 500 000 500 000 680 000 500 000 500 000 500 000 600 000 500 000 500 000 500 000 .500 000 500 000 500 000 500 000 500 000 500 000 500 000 500 000 500 000 500 000 500 000 500 000 4.7 " 17.:::.::.... 4.5 Oct. 1 4.2 " 4 4.1 " 9 4.1 " 10 3.6 " 14 8.5 " 17 3.4 " 23 3.7 " 88 4.3 " 31 3.6 Nov. 4 4.0 " 14 3.4 " 25 3.4 " 30 3.4 Deo. 1 3.4 " 5 3.4 8 3.4 " 11 4.7 " 17 " 23 3.4 3.4 " 26 3.4 " 27 " 81 3.4 3.4 Jan. 1 3.4 " 3 3.4 " 4 3.4 " 5 3.4 " 9 3.4 " la 8.4 " 1.5 •' 18 3.4 3.4 " 22 3.4 " 25 3.4 '■27 3.4 " 28 3.4 " 29 " 30 3.4 Filter put in service September loth, 1908. Filter scraped January 30tli, 1909. Sand removed, washed, and stored, 18 cu. yd, — 124 cu. yd. per acre. Quantity filtered during run = 71 000 000 gal. Millions of gallons per acre filtered between sorapuigs = 491. Average rate of filtration = 3 623 000 gal. per acre daily. Net area of sand surface =- 0.1446 acre. Wastedfromsurfaceof filter for scraping 104 500 ga). Water drained out, and water used for refillmg from below. . .156 800 261 300 — 0.36% of filtrate. Scraping. — Filter No. 1 was scraped hy the superintendent of the plant and a force of five men, the actual time occupied in piling up i'14 WATER PURIFICATION AT STEELTON, PA. [Papers. the scrapings ready for transportation to the washer being 3 hours, ar at the rate of about 7 sq. ft. of filter scraped per man per minute. The scraping on this filter was rather deep, averaging nearly an inch, and being less than i in. in very few places. A portion of the extra scraping was made necessary by the accidental dropping of one of the manhole covers into the filter early in January. This caused a great disturbance of the sand, over an area of perhaps 100 sq. ft., as the cover turned when it struck the water and penetrated the sand bed on its edge, the lower corner having reached almost to the gravel under- drains. The cover was taken out, and all the discolored sand which had been driven down into the bed from the schmutzdeche was dug out and sent to the washer. Sand Transportation and Washing. — ^The ejector and washing plant worked satisfactorily as to the handling of the sand, but, as it had not been in use before, the men did not attain a high degree of efficiency in its operation. It is believed that greater familiarity with its use, and the proper regulation of the relative force of the different jets, will permit of handling the sand much more rapidly and efficiently than was done. The total quantity of sand removed from Filter No.l was about 18 cu. yd., or 124 cu. yd. per acre of filter surface, to trans- port which to the sand washer, wash it, and return it to the sand troughs on top of the filters occupied about 8 hours, one man attending to the sand washer, one to the hose returning the sand to the sand troughs, one raking the scraped surface of the filter, and an average of three shoveling the dirty sand to the portable ejector. After the sand for one filter had all been washed, the portable ejector was taken apart and an irregularly-shaped stone, which had carelessly been left in one of the water mains, was found wedged in the nozzle. This had caused the deflection of the jet of water so that the throat on the discharge side had been cut deeply. After the throat had been replaced with a new one, the original capacity of the ejector was restored, and the scraping and transporting of sand to the washer from the other filters proceeded with more expedition. Ejector Nozzles. — The nozzle in the portable ejector throws a f-in. jet of water into a throat with a IJ-in opening reduced to 1 in. in diameter IJ in. from the front of the throat and increasing to If in. in diameter at the back end of the throat 7 in. from the face; the nozzle and throat stand IJ in. apart. The nozzle reduces from 2 in. in Papers.] WATBK PURIFICATION AT STEBLTON, PA. 715 diameter to the f-in. jet in a length of about 6 in. The nozzles and throats are of chilled cast iron. The nozzles for the sand washer have a if-in. jet, and the throats reduce from IJ in. to IJ in. in diameter in IJ in., increasing to If in. in 5J in. The portable ejector and the washing hoppers were supplied by the Norwood Engineering Company, from designs by Allen Hazen, M. Am. Soc. C. E., originally worked out for the filter plant at Wash- ington, D. C, and containing recent improvements suggested by Mr. Hazen. The washing plant consists of only two hoppers, and practi- cally all the washing is done in the first one, the sand as it is thrown to the second hopper from the first being sufficiently clean to be used in the filter. The wash-water overflows the second hopper with very little turbidity. The lift from the portable ejector to the top of the pipe discharging into the washing hopper is about 15 ft.; from the second washing hopper to the sand troughs there is a drop of about 6 ft., so that com- paratively little pressure is required to transport the sand from the sand washer back to the sand bins. The washing hoppers are provided with auxiliary jets at the bottom to compensate for the quantity of dirty water actually picked up by the jet and forced through to the succeeding hopper, the auxiliary jet also serving the purpose of stirring up the sand at the bottom of the hopper so as to facilitate its transportation by the water jet. - The dirty water overflowing the washing hoppers passes through a reinforced con- crete box in order to catch such sand as may escape with the wash- water; the dirty water escapes from this box to the sewer by an overflow. The water which transports the washed sand to the sand troughs overflows a weir at one end of each sand box and escapes to the sewer. Refilling after Scraping. — After the dirty sand is all thrown out of the filter and the surface of the clean sand has been raked over to remove footprints and give a smooth even surface, filtered water is admitted through the underdrains until the surface of the filter sand is covered about 2 in. in depth; the raw water is then admitted, gently at first, and the filter is refilled to its proper operating depth with the rough-filtered water. Resumption of Filtration. — In starting this filter in operation after scraping, the effluent control valve was regulated so that the filter would 716 WATEH PUEIFICATIOIT AT STEBLTON, PA. [Papers. deliver, according to its rate-gauge, 800 000 gal. a day, it being necessary to choke down the discharge of the clean filter to a proper rate because the two other filters were very much clogged up and could not yield their proportional part of the water without a loss of head of 5 or 6 ft.; with such a loss of head the clean filter would at once start off with its maximum allowable rate, which was deemed inadvisable, the filter being so new, and having been started in opera- tion so late in the season. Lengths of Buns. — Filter No. 2 was scraped on February 2d and was put back in service on February 3d, and Filter No. 3 on February 4th, 1909. The lengths of runs of these three filters, therefore, were 135.5, 137.5 and 140.5 days, respectively. Filter No. 1 delivered during its run 11 000 000 gal. at an average rate of 3 623 000 gal. per acre daily, the total yield corresponding to 491 000 000 gal. per acre between scrapings. The figures for the other two filters were a trifle larger. The total quantity of water wasted from the surface of the filter prior to scraping, and the quantity used for refilling the filter from below, before starting in operation again, corresponded to 0.14 of 1% of the quantity of water filtered. The three slow filters have now (May 10th, 1909) been in service for 14 weeks since the last (and only) scraping, and the loss of head is but O.Y ft. Efficiency of the Plant. The efficiency of the entire plant, in the removal of turbidity and bacteria, is exhibited in the daily records given in Table 9, which date from November 1st, 1908, for the reason that no bacterial analyses were made prior to November 5th. The efficiency in the removal of turbidity has been 100 per cent. During November and December, when the turbidity of the raw water was low and the bacteria were not numerous, the effluent water was satisfactory. The same is true during January, with the exception of the second week when, as has already been explained, the superin- tendent was deceived by the deterioration of his silica-turbidity stand- ards, and the quantity of coagulant was inadequate to produce a satis- factory effluent. The results subsequent to that date are good. The samples collected on the afternoon of February 5th, when Filter No. 3 was being scraped, represent the bacterial content's of the effluents of Filters Nos. 1 and 2, which had been scraped, No. 2 within two days Papers.] TABLE WATBE PUKIFICATION AT STBELTON, PA. 717 9. — Daily Efficiency of Plant in Eemoval of Turbidity AND Bacteria. Average turbidities, ih Bacteria, per cubic i 1^ PARTS PER MlLLIOf centimeter: II ill 5" ^a Date. 1 il f MS i i k it 1 1 p 2 < 1908. Nov. 1... 5 100 8... 4 100 3... 3 100 4... 2 100 5... 3 100 80 3 96 6... 3 100 7... 3 100 8... 3 100 9... 2 100 10... 2 100 11... 2 100 12... 2 100 13... 8 100 14... 2 100 15... 1 100 16... 2 100 17... 3 1 100 18... 3 1 100 19... 2 100 20... 2 100 119 76 6 95 21... 2 100 22... 2 100 23... 2 100 24... 2 100 25... 2 100 26... 100 27... 100 214 12 3 90 28... 100- 29... 100 30... 100 Deo. 1... 100 8... 100 3 .. 100 4... 100 5 .. 100 6... 100 408 310 ,52 87.8 7... 100 8... 100 9... 100 10... 100 11 100 12... 100 220 185 8 96.4 13.. 100 14.. 100 15.. 100 16 100 17 . 100 1 680 680 71 95.9 18.. 100 19.. 8 100 20 i 100 . > • 21 5 100 4 435 800 58 98.7 82.. 15 100 83.. 18 2 100 718 WATER PURIFICATION AT STBELTON, PA. TABLE 9.—(GonUnued.) [Papers. Average turbidities, in Bacteria, PER cnBic i in PABTS PER million: Is ss S "• II Date. 1 ft 1 1 l| 1 II 1 K o 1 p " 1 1 p |l M fK Ph « b, &( o 5 1908. Dec. Z4. . . 10 3 100 25... 10 3 100 26... 6 2 100 2T... 5 100 38... 4 100 89... 3 100 30... 3 100 31... 8 100 1909. Jan. 1... 4 100 5 800 1 125 100 98.3 2... 2 100 S... 3 100 4... 3 100 5... 3 8 100 6... 7 100 7... 180 100 8... 850 100 9... 450 100 0.5 10... 400 100 0.15 11... 425 100 18 200 6 000 1105 93.9 0.30 18... 450 100 0.90 13... 250 100 0.10 14... 180 100 0.20 15... 125 100 32 000 6 200 320 99.0 0.35 16... 75 100 0.30 17... 60 100 o.to 18... 30 100 19... 25 100 SO... 20 100 81... 35 100 5 975 1 445 115 98.1 82... 15 100 0.25 23... 10 100 0.25 24... 20 100 0.25 18 25... 60 100 0.25 83 26... 770 100 1.38 86 27... 600 100 13 160 1 085 11 99.9 0.78 83 28... 490. 100 1.32 16 89... 825 100 1.35 •m 80... 285 100 1.00 20 31... 100 100 0.65 28 Feb. 1... 75 100 0.55 16 8... 60 100 0.50 18 3... 80 3 1 95 14 4... 20 2 100 0.25 S*.. 15 100 3 025 295 188 94 0.25 6... 70 KlO 0.25 14 7... 45 100 0.40 17 8... 190 2 100 0.60 14 9... 170 100 0.60 12 10... 215 100 0.60 13 11... 170 100 0.60 12 * Filter No. 3 out of service for scraping; Filter No. 1 was scraped on January 30tli, and No. 2 on February 3d. Papers.] WATER PURIFICATION AT STEELTON, PA. 719 TABLE 9.—(Oontinued.) Date. 1909. Feb. 12. IS. 14. 15. 16. 17. 18. 19. 20. 21. 27. 28. Mch. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. IS. 14. 15. 16. 17. 18. 19. 20. 2S. 24. 25. 26. 27. 28. 29. 30. 31. Apr. 1 . 2. 3. 4., 5., 6., 7., 8., 9.. 10., 11.. 12., IS.. 14.. 16.. 16.. 17.. 18.. AVERAQE TURBIDITIES, IN PARTS PER million: 270 160 80 30 120 200 310 190 125 300 390 465 380 275 310 310 260 150 120 100 65 52 47 45 60 60 50 50 90 120 115 100 75 47 42 45 40 SO 25 22 20 20 25 110 140 120 150 135 100 85 72 60 50 SO 50 55 100 95 80 95 140 345 350 340 TV® n 9 >a:s &■= » f:a ■Sfei. ^ 3 18 .ago, ^ fe ^^ 0.65 13 0,65 15 0.35 16 0.83 14 0.50 13 0.70 10 0.80 6 1.00 6.5 1.20 7.5 0.90 9 0.80 9 1.10 11 1.00 11.5 0.85 11.3 0.80 10 0.80 11 0.80 11 0.65 11 0.65 8 0.50 7 0.40 6 0.40 6 0.00 8 0.00 10 0.35 11 0.35 12 0.35 12 0.35 U 0.33 11 0.40 14 0.40 16 0.40 14 O.SO 14 0.30 12 0.30 14 0.30 14 O.SO 14 0.25 14 0.25 16 0.23 16 0.00 15 0.25 15 0.25 14 0.40 14 0.55 14 0.57 12 0.62 14 0.57 14 0.47 12 0.45 14 0.40 16 0.32 16 0.35 17 0.40 19 0.40 19 0.45 18 0.52 15 0.48 14 0.45 16 0.50 15 0.45 14 0.S5 14 0.49 13 0.80 12 1.25 12 1.00 11 ■5 f^S 3 3.5 3 a> g S ic ft SB a bo 0.42 0.42 0.30 0.30 0.25 0.25 * Raw water sample not taken. 720 WATEE PURIFICATION AT STEELTON, PA. [Papers. and No. 1 within 5 days of the date of taking the sample. The indi- vidual samples of the effluents of Nos. 1 and 2 showed the following counts : Filter No. 1, put in service, January 31st. Sample No. 1 185 (collected February 5th) Sample No. 2 lYO ( " " " ) Average 176 Filter No. 2, put in service February 3d. Sample No. 1 200 (collected February 5th) Sample No. 2 185 ( " " " ) Average 198 A slight improvement shows in the effluent of the older filter. The efficiency in the removal of B. Ooli is shown in Table 10. TABLE 10.— Presumptive Tksts foe Presence op B. Ooli in the Eaw' Water, Filtered Water at the Filter Plant, and Tap Water in Steelton. Pa. Raw Water. Filtered Water. Tap Water. Date. Number of Positive Number of Positive Number of Positive 1-cu. cm. indica- 1-cu. cm. indica- 1-ou. cm. indica- ' sowings. tions. sowings. tions. sowings. tions. Nov. 20, 1908. 5 3 5 5 " 27- 5 3 5 5 Dec. 6 5 3 5 5 " 12 5 1 5 " 17 5 1 5 5 " 21 4 1 4 1 4 Jan. 1, 1909. 5 1 5 5 " 11 5 5 5 4 5 5 " 15 5 3 5 1 5 •' 21 S 4 5 5 1 " 27 5 5 5 5 Feb. 5 4 3 4 4 " 12 4 4 4 4 " 19 3 1 3 3 " 27 5 S 5 1 5 Moh. 6 5 2 5 5 " 13 5 3 5 5 " 28 5 4 6 5 " 27 S 5 Apr. 3 5 5 5 S " 18 4 3 4 4 ' 17 4 4 4 4 The presence of B. Goli in the filtered water on December 21st, January 11th, and January 15th was in part due to lack of, or insuffi- cient use of, coagulant in the preliminary process. The dropping of Papers.] WATER PUEIFIOATION AT STEELTON, PA. 721 a manliole cover into Filter No. 3 on January 1st, however, may have contributed to the deterioration of the effluent during the early part of that month. Thus far, no chemical analyses have been made of the applied water and final filtrate at the Steelton plant, and no data, therefore, have been secured on the degree of nitrification accomplished, or on the other usual changes expected. It is evident that between periods of turbid water, when no coagulant is being used, the slow sand filters will be penetrated deeply by such suspended matters as may pass through the roughing filters, among which would be bacteria, and in this respect the work required of these filters will differ from that performed by the ordinary slow sand filter. During the past few months no visible suspended matters have been left in the rough- filtered water. In nearly all plants using aluminum sulphate as a coagulant, trouble, either continuously or intermittently, has been experienced with discoloration of the water, generally only the hot water, by iron rust. This trouble has been more or less serious at Harrisburg, Pa., Watertown, N. T., Charleston, S. C, Hackensack, N. J., and various other places. Thus far, there has been no trouble of this sort at Steelton, although at times as much as 2 gr. of aluminum sulphate has been used per gallon of water filtered. Cost of Operation. — The plant has not yet been in operation suffi- ciently long to obtain accurate data, but an approximation for an average year would indicate an expense, per milUon gallons, when filtering 3 000 000 gal. of water daily, about as follows : Cost of Purification op Steelton Water per Million Gallons, on a Basis of Filtering 3 000 000 Gal. per Day. Operation of Roughing Filters. Labor $1.55 Superintendence 0.42 Supplies and analyses of water 0.30 Coagulant 0.53 Power 0.08 Wash-water • 0.13 Light '. 0.05 Coal 0.10 $3.16 722 WATER PDEIFIOATION AT STBELTON, PA. [Papers. Operation of Slow Filters. Scraping, 5 times per year $0.60 Transporting and washing sand 0.06 Ee-sanding filters 0.25 Superintendence 0.42 $1.33 Total $4.49 Interest and sinking fund charges are not included. Achnowledgments. — The principal credit for the existence of the Steelton filter plant is due to J. V. W. Reynders, M. Am. Soc. 0. E., President of Councils of Steelton, Chairman of the Filtration Com- mittee of Councils and Vice-President of the Pennsylvania Steel Com- pany, who, by arranging a well-conducted campaign of education and devoting much thought and time to the cause, prepared the way to enable Councils to see the necessity of making provisions for a purer water supply. On its completion, the plant was turned over for operation to the Board of Water Commissioners, Mr. George H. Roberts, President, under the general direction of the Superintendent of the Water-Works, Mr. O. P. Baskin. The operation of the filters is in charge of Mr. M. B. Litch, as Filter Superintendent, the writer, acting in an advisory capacity, as required. The works were built by contract under the supervision of the writer's engineering force, Mr. Paul Hooker being Principal Assistant during the designing of the works, and Resident Engineer during their construction; to his faithfulness the plant bears permanent witness. Papers.] MODERN HTDBATJLIC TURBINES 735 TABLE 2. — Tests of a 28-inoh K. H. Wellman-Seaver-Morqan Company Turbine "Wheel, No. 1Y96. Date, February 25tli, 1909. Wheel supported by ball-bearing step. Swing-gate.' Conical draft-tube. Propoetional I ■3 a « ■3 .1 'S'S Part op: o-g 1 III ■s II ID a la SI 1 It "i'C llll 8, 1 If 95 1.077 0.971 17.11 4 153.00 97.00 126.20 66.58 94 1.077 1.017 16.97 3 199.67 101.16 147.66 75.84 93 1.077 1.036 16.94 3 284.33 102.98 156.38 79.04 92 1.077 1.053 16.89 3 239.33 104.50 159.58 79.72 91 1.077 1.061 16.87 3 247. .S3 105.22 161.17 80.06 89 1.077 1.068 16.81 3 253.67 105.70 161.45 80.12 90 1 077 1.072 16.80 3 259.00 106.08 161.71 80.01 88 1.077 1.079 16.82 4 267.50 106.82 162.15 79.58 87 1.077 1.026 17.05 3 294.67 102.27 125.03 63.23 86 1.000 0.913 17.28 3 147.00 91.65 120.29 66.98 85 1.000 0.957 17.16 2 190.50 95.70 144.34 77.50 84 •1.000 0.972 17.14 3 211.67 97.10 152.69 80.89 83 1.000 0.981 17.13 3 285.00 98.03 156.85 82.36 W 1 000 0.990 17.11 3 232.67 98.90 157.96 82.31 80 1.000 0.996 17.09 4 240.25 99.43 160.19 83.13 81 1 000 1.003 17.07 3 247.33 100.07 161.17 83.19 79 1.000 1.004 17.07 4 252.25 100.14 160.55 82.82 78 1 000 1.001 17.13 4 259.00 100.00 167.00 80.82 77 1 000 0.983 17.22 3 268.33 98.43 146.39 76.15 76 1 000 0.911 17.47 4 893.60 91.96 106.75 58.59 106 923 0.8ti8 17.32 4 143.85 87.24 115.49 67.39 105 923 0.899 17.24 5 176.00 90.12 135.49 76.90 104 0'923 0.920 17.15 4 201 .00 91.96 146.21 81.74 101 0.923 0.931 16.93 5 213.20 92.52 148.62 83.66 100 923 0.936 16.93 6 220.67 92.96 150.48 84.31 99 0.923 0.942 16.91 4 227.35 93.51 151.52 34.50 102 923 0.945 16.93 4 282.00 93.82 151.88 94.81 lOS 923 0.945 17.04 4 235.50 94.14 152.74 83.96 98 0!923 0.945 16.93 4 23T.75 93.82 151.32 84.00 97 923 0.924 17.02 4 254.50 92.04 188.84 78.15 96 923 0.823 17.25 4 288.25 82.47 87.36 54.15 42 923 0.870 17.17 3 146.33 87.02 116.20 68.57 41 923 0.895 17.10 4 170.00 89.37 130.87 75.51 40 0'92S 0.921 17.04 4 202.75 91.80 146.35 82.44 39 923 0.928 16.99 3 208.67 92.35 147.99 83.17 35 0'923 0.932 17.03 4 216.25 92.81 150.75 84.10 88 923 0.937 16.97 4 220.00 93.20 151.36 84.38 36 o!923 0.939 17.01 4 223.75 93.44 152.58 84.65 34 923 0.940 17.02 4 226.25 93.66 150.86 83.45 37 0'923 0.944 16.97 3 238.33 93.90 151.69 83.94 33 933 0.921 17.13 4 256.25 98.05 139.80 78.18 32 923 0.823 17.27 3 288.00 82.54 87.29 53.99 31 0i923 0.730 17.50 4 334.75 73.75 74 846 0.824 17.46 3 158.67 83.15 120! 23 '7s!62 75 72 70 71 73 69 68 67 66 65 o!84B 0.846' 0.836 17.46 3 175.67 84.35 129.91 77.78 0861 17.3* 5 202.20 86.50 143.40 84.30 o!846 865 17.33 4 209.00 86.95 145.69 85.25 0*846 868 17.33 3 215.00 87.24 147.27 85.89 0'.846 0.846 0.846 0.846 ' 846' 0'870 17.34 4 219.25 87.47 148.19 86.15 0.869 17.32 4 221.25 87.32 147.53 86.01 866 17.38 4 227.75 87.02 144.96 84.76 0.858 17.36 4 231.75 86.25 140.48 82.73 845 17.39 4 243.75 85.11 132.98 79.22 0.846 0.828 17.44 4 256.50 83.44 124.39 75.37 736 MODEEN HYDRAULIC TURBINES [Papers. TABLE 2.— (Continued.) Proportional ■3 0) S « ■3 . o a Part of: If ill fell P i 1 t g-ssl it 64 0.846 0.754 17.59 3 282.00 76.31 85.47 56.15 80 0.769 O.7S0 17.38 3 141.00 75.52 102.56 68.90 26 0.769 0.766 17.35 3 166.00 77.08 115.72 76.36 27 0.769 0.779 17.38 3 183.00 78.24 124.84 80.84 25 0.769 0.789 17.81 3 194.00 79.85 129.36 83.15 89 0.769 0.793 17.25 4 200.75 79.48 131 AZ 84.52 as 0.769 ■ 0.792 17.86 4 206.00 79.40 131.11 84.36 24 0.769 0.773 17.38 4 226.25 77.82 123.43 80.47 23 0.769 0.735 17.49 3 251.33 74.17 106.64 72.49 22 0.769 0.690 17.56 3 269,67 69.82 81.73 58.t8 21 0.769 0.623 17.68 4 323.75 63.27 20 0.615 0.617 17.91 4 139.50 63.07 'ssira 'm.a , 16 0.615 0.627 17.79 3 158.33 63.81 95.97 74.55 17 0.615 0.634 17.80 3 171.33 64.55 101.26 77.71 15 0.615 0.638 17.73 4 179.50 64.82 103.37 79.31 18 0.615 0.636 17.77 4 183.00 64.74 103.16 79.07 19 0.615 0.634 17.80 2 188.00 64.61 108.56 78.64 14 0.615 0.627 17.72 4 194.50 63.74 100.21 78.24 13 0.615 0.596 17.77 3 218.67 60.60 92.78 • 75.97 18 0.615 0.563 17.79 4 243.00 57.29 73.65 63.72 11 0.615 0.519 17.98 4 312.00 53.00 10 0.468 0.452 17.34 3 117.83 45.40 'hh'.bb 'esira 6 0.462 0.453 17.08 3 136. OC 45.05 61.83 70.17 7 0.468 0.461 17.08 4 146.75 46.04 64.94 72.82 6 0.462 0.462 17.05 4 152.00 46.04 66.34 74.53 4 0.462 0.462 16.92 4 155.25 45.90 65.88 74.79 9 0.462 0.459 17.16 8 162.00 45.94 66.78 74.69 8 0.462 0.457 17.15 3 166.67 45.69 65.67 73.90 3 0.462 0.451 16.98 4 17i.25 44.83 62.65 72.57 8 0.462 0.432 17.05 4 217.50 43.04 52.74 63.37 1 0.462 0.404 17.18 5 282.40 40.40 Note.— During the above experiments, ihe weight of the dynamometer, and of that portion of the shaft which was above the lowest coupling, b as 2 600.1b. With the flume empty, a strain of 0.5 lb., applied at a distance of 3.2 ft. from the center of the shaft, sufficed to start the wheel. by three-way cocks. By connecting a set of tubes with the vacuum tank, the columns were drawn above the floor level, after which, by closing the cock, they were held in a stationary position. If drawn too high they could be dropped back by opening the cock to the atmos- phere. After the columns were once adjusted and the cocks closed, they would stand indefinitely without attention. This apparatus for the measurement of velocities has been described at some length for the benefit of those who may be interested in the same line of experiments. Lack of space precludes the publication of any of the data obtained, although they are of great interest to the designer, and throw much light on some very obscure points in the theory of turbine design. REFOfiT Of. THE Board of Water CoMnissioNERG CITY OF OQPENSBURG; N. Y., TO THE HONORflBLO MAYORftND COMMON COBNGIL ^- ' RELATIHQ to' ' ■ _ . '■' -^/ATEK SliFFLT, FROFOSED FILTRATION ELANT AND ELAN ■r-\\ ■ poR- FINANCING SAtlE. TOQEThER WITH ■'REFORT ,0P HESSRS. liAZEN, VHIFFLE -AND FULLER, Sanitary^, Enqiweers, : AND. ,- ^ • COPT OF LETTER. FROM TH£ ENQIMfiERlNG DET^flRinENT OF THE STATE ■board OFjHfiftL^fH/ ENDORSMQ 'amp AFFROVlNG the said ^ '^ ; SANITARY ENGINEERS' REF'oRT.- SlJBniTTED, 'FEBRUART 2nD, 1909. REPORT OF THE Board of Water CoMnissiONERS CITT OF OQDE/NSBURG, N. T., TO THE HOfiORflBLE MAYOR and COMMON COUNCIL KELATINQ TO WATER 5UFFLT, PROPOSED FILTRATION FLANT AND FLAN FOR FINANCING SAME. TOGETHER WITH Report of Hessrs, Hazen, Whipple and Fuller, Sanitary Engineers, AMD COPY OP LETTER FROIi THE ENGINEERING DEFARTHENT OF THE STATE BOARD OF HEALTH, ENDORSING AND AFFROVING THE SAID SANITARY ENGINEERS' REPORT. SUBniTTED, FEBRUARY 2«d, 1909. REPORT RELATIVE TO NEW WATER SUPPLY. At a joint meeting of tlie Common Council, the Health Board, and the Board of Water Commissioners held more than one year since, the atten- tion of the City authorities was called, by Dr. Eugene H. Porter, Commis- sioner of Health of the State of New York, to the dangerous and hazardous conditions surrounding the Water Supply of the City. While for many years those directly chargable with the management of the Water Works, and a considerable number of our citizens, have realized that radical improvement in the conditions around and about the present intake was necessary, in order' to remove therefrom the saw-dust and filth nuisance, and while from time to time our medical fraternity has sounded notes of warning against the danger of sewage contamination, but a very small proportion of our Citizens appreciated the real seriousness of the situation as it was outlined and' explained by Or. Porter. His remarks and Tiis warnings were based upon the reports of Engineers and Chemists con- nected with the State Board of Health, after careful study and investiga- tion of the source of our Water Supply, and frequent and regular analysis of the water of the Oswegatchie River, samples taken from various points between the Railroad Bridge and the Dam. After listening to and carefully considering Dr. Porter's statements, the Water Board in discharge of what it believed to be its manifest duty, -engaged the services of Messrs. Hazen, Whipple and Fuller, Sanitary Engi- neers, recognized as standing at the head of their profession, to make all necessary surveys, examinations and estimates, and to report their con- clusions, findings and- recommendations. Their very full and able report was duly received by this Board, and by them published in the City news- papers, to the end that all of our citizens might have opportunity to see a,nd know the views and opinions regarding our present Water Supply, held by Expert Sanitary Engineers and their recommendations for improving the same. A copy of said .report is herewith submitted. It may appear that the Board of Water Commissioners have moved very slowly in these premises, and that the "Engineers' Report, accompanied with the opinion and recommendation of the Commissioners is very late in com- ing before the Mayor and Council! We can only say, appreciating the gravity of the situation, they have thought it wise to avoid hasty conclusions. The Board of Water Commis- sioners regard this, one of the greatest problems the citizens of Ogdensburg have ever been called upon to solve, a problem that in the near future they, and they alone, must solve, or the inevitable consequences, the loss of our largest and steadiest consumer, (the State Hospital) with its annual rate payment of $4,000.00, and a possible dangerous and destructive epidemic of typhoid will surely have to be faced. The Commissioners hiave studied this problem in all its phases; they know the regard held by many of our citizens for the soft waters of the Oswegatchie, they know only too well that our taxes are high, and they fully realize that there is a limit of taxation beyond which we cannot go and ought not to go. At the same time they also know that common sense, and due regard for the health, prosperity and happiness of our City and its citizens, demands, that we do not disregard the warnings of those compe- tent to judge of the dangers surrounding ' us, and that we do not delay action until we are face to face with death and destruction in our homes. In arriving at their conclusion that, should a filtration plant be installed, the Oswegatchie River should be abandoned as the future source of water supply, and the St. Lawrence River be substituted therefor, the well-estab- lished sentiments and prejudices of the majority of the Commissioners had to undergo a complete change before our minds met and the members of the Board as a unit stood ready to advocate the change. When due thought and consideration was given, to the variable and lessening fiow of water in the Oswegatchie, the danger of a future shortage in supply, the 4 REPORT .RELATIVE TO NEW WATER SUPPLY. fact that a, filtration plant for the treatment of its waters, is estimated tOK cost, when construction and operation are both considered as much' or more than the construction and operation of a filtration plant suitable for treat- ment of the St. Lawrence River water, the fact that the Oswegatchie water would require chemical treatment, whether mechanical or Sand Filtration was adopted, the fact that mechanical filtration has proved far from satis- factory at Watertown where a plant of this type has been installed re- cently, that its operation involves the steady employment of an experienced ; Chemist, that the proper treatment of its water supply requires a variable and changing dose of chemicals, according to the changing character of the water, that the life of mechanical filters is estimated to be 20 to 25 years, that the saw-dust, silt and debris with which the Oswegatchie River is: at times surcharged would without doubt cause great trouble and expense and might completely clog the filters, finally convinced the Commissioners that a change in our water supply is necessary. All these unquestionable facts have led the Commissioners to agree that wisdom and past experience dictate that when the citizens of Ogdens- burg decide to install a filtration plant, such plant should be located where the supply of water will be for all time unfailing, easily and economically flrtered and distributed, and that the filter should be of the practically in- destructible sand filtration type. The saving to all our citizens, that will result from the adoption of filtration, by eliminating dirt and filth from our water supply, and the con- sequent reduction in plumbers' bills, will, in the opinion of the Commis- sioners, be an item of no mean proportion. The impression has prevailed that Bonds necessary to meet the cost of extending and improving our Water Works, can not legally be issued to run for a longer period than 20 years, and must be paid at the rate of l-20th of the amount of any issue, each year. The Commissioners are satisfied this is an erroneous impression, a written opinion of the City Attorney, copy whereof is hereto annexed, and the pertinent fact that municipalities in New York .State of like classes with Ogdensburg have, within the past two or three years Issued bonds of this character, running for 40 years, convinces the members of this Board that they are right in their interpretation of the law, and that 30-year con- struction bonds can be legally issued as outlined in the plan herewith sub- mitted. It may appear that the Commissioners have far exceeded the bounds of economy aiid prudence in adding 10% to the Engineers' estimate of cost of construction, and $1,000.00 to their estimated cost of filter operation. The members of this Board maintain it is far better to overestimate than under- estimate on a proposition of this magnitude. Contingencies of a character Impossible to foresee may arise. Engineers' estimates may be at fault, and the work once commenced must be carried to completion. While provision should be made authorizing the issue of Bonds In amounts as recommended in this report, the Bonds should be issued only when and as funds are required. A marked saving in interest will thus result, and if the whole amount authorized is not needed they would not be issued. The members of this Board believe this is a matter that should be re- ferred to the taxpayers of this city, in such a way as to get an expression of opinion from the largest number, the final decision must rest with them; all that their chosen representatives, the Council, the Water Board and the Board of Health can do is to lay the plain and true facts before them, with such suggestions and recommendations as to them may seem good. With that view and purpose the WATER. COMMISSIONERS RECOMMEND, in event that a filtration plant be installed, that such plant be of the type known as slow sand filtration, located above the Ship-Yard, and that' the River St. Lawrence be the source of supply, as recommended in the Report of the Engineers, which report has been referred to, and approved by the State Board of Health. REPORT REILATIVE TO NEW WATER SUPPLY. 5' THEY FURTHER RECOMMEND.that the cost of the plant with its ap- purtenances and accessories be financed by the issue of- 30-year construction bonds, and 20-year refunding and meter bonds, so arranged as to maturity, that the entire issue of both classes of said bonds, and the present out- standing old Water Bonds, can be paid, principal and interest from the earnings of the Water Works, without recourse to direct taxation or an increase in Water Rates. This they feel confident can be done, as clearly outlined in the plan and. tables herewith submitted. Respectfully, JAMES M. WELDS, FRANK CHAPMAN, GEORGE F. DARROW, WILLARD N. BELL, Water Commissioners. The table hereto attached shows plan suggested by the Board of Water Commissioners for financing the cost of a Filtration Plant using the St. Lawrence River as the source of supply, in conformity with the suggestions and recommendations contained in the Report of Messrs. Hazen, Whipple- and Puller, Sanitary Engineers. The estimates and calculations are made on the following assumptions and basis, namely: FIRST. That a Sand Filtration Plant, with Pumping Station, Stand. Pipe, Connections and appurtenances, be constructed in the years 1910 and 1911, and that Bondsi to pay for same be issued in 1910. SBCONEi. That the cost of the aforesaid plant and appurtenances may exceed $160,000.00, the Engineers' estimate probably 10%, and therefore pro- vision should be made to authorize the issue of $175,000.00 "CONSTRUC- TION BONDS" to bear not to exceedi 4% interest and to be payable within thirty (30) years. THIRD. As outlined in the Engineers' report, it will be absolutely necessary, to prevent unreasonable waste of water, and overtaxing the Fil- tration Plant beyond its capacity that meters be installed as rapidly as is. practicable. It is estimated that $20,000.00 will be needed for this purpose. FOURTH. The amount of old water bonds outstanding in 1911 will be $74,550.00 These bonds mature at the rate of $9,300.00 each year from 1911 to 1917 inclusive, $5,550.00 of said Bonds mature in 1918, $1,550.00 of said, bonds mature each year 1919 and 1920, and the remainder thereof, $800.00 mature in 1921. If the old bonds and the $175,000.00 new- 30-year Filtration Construction Bonds, and the $20,000.00 needed for meter installation are to be paid from, and out of the earnings of the Water Works, without recourse to di- rect taxation or an increase of Water Rates, it will be necessary to refund $40 000 00 of the old bonds, and therefore provision must be made to author- ize the issue of $60,000.00 "REFUNDING AND METER BONDS" to bear not to exceed 4% interest and to be payable within twenty (20) years. FIFTH. The Board of Water Commissioners believe that bonds caa be issued as above outlined, in conformity with the Constitution and Laws of the State of New York, and they further believe that the earnings of the Water Works, based on the returns of the past 20 years, with due allow- ance made for the extra cost of operating the Filtration Plant, will be suf- ficient to pay the entire cost of a Filtration Plant, (such as is now under consideration,) the cost of meter installation, which must go hand in hand with filtration, and that said earnings can be safely relied upon to pay all of the bonds issued for water purposes, together with the interest thereon, as is clearly shown by the table hereto annexed. SIXTH. The earning power of the "JVater Works after the construction of the Filtration Plant, and installation of meters, is estimated on the fol- lowing basis: , ^,. .,. . ^ That in 1911 the yearly gross earnings will be $32,000.00 and that based on the average yearly gain for the past 20 years, these earnings may he- safely counted on to increase at the rate of $500.00 per year. « REPORT RELATIVE TO NEW WATER SUPPLY. That the cost of operation of th6 present plant,, now averages $10,500.00 per year; and That the additional cost of operation of the Filtration Plant will be $5,- 500.00 per year, this being $1,000.00 greater than is estimated by the Engi- neers. Under the plan as formulated there will, in the judgment of the Com- missioners, remain of earnings each year after payment of Bond and In- terest charges, a sufficient sum to provide for the necessary and profitable extension of water mains, thereby increasing the earnings of the Water Works. ESTIMATED CONDITION OF WATER SUPPLY FINANCES FOR EACH YEAR FROM 1911 TO 1940. Prin. Rate. Interest. Bonds Paid. I 32,000.00 3%% $1,120.00 $4,000.00 Old Bonds. 7,500.00 3%% 2G2.00 750.00 Old Bonds. 8,800.00 3%% 308.00 800.00 Old Bonds. 26,250.00 4 % 1,050.00 3,750.00 Old Bonds. 175,000.00 4 % 7,000.00 30 year Filter Bonds. $249,550.00 $9,740.00 $9,300.00 DISBURSEMENTS. Bonds ' $9,300.00 Interest on Bonds 9,740.00 $19,040.00 Current expenses ,. . 10,500.00 Filter expenses 5,500.00 Total disbursements $35,040.00 RECEIPTS. Water Rates $32,000.00 Proceeds Refunding and Meter Bonds 7,000.00 Total receipts 39,000.00 Total disbursements 35,040.00 $ 3,960.00 Say $ 4,000.00 1912. Prin. Rate. Interest. Bonds Paid. $ 28,000.00 31^% $ 980.00 $4,000.00 Old Bonds. 6,750.00 3%% 236.00 750.00 Old Bonds. 8,000.00 3%% 280.00 800.00 Old Bonds. 22,500.00 4 % 900'.00 3,750.00 Old Bonds. 175,000.00 4 % 7,000 00 Filtration Bonds. 7,000.00 4 % 280.00 Refunding & Met'r B'nds $247,250.00 $9,076.00 $9,300.00 DISBURSEMENTS. Bonds $9,300.00 Interest on Bonds 9,676.00 $18,976.00 Current expenses 10,500.00 Filter expenses 5,500.00 Meters 5,000.0(J $39,976.00 REPORT , RJ^LATIVE TO NEW WATEiR SUPPLY. T RECEIPTS. Water Rates $32,500.00 Proceeds Refunding and Meter Bonds 11,000.00 Total reqelpts $43,500.00 Total disbursements 39,976.00 Available . . ; $ 3,524.00 1913 Prin. Rate. Interest. Bonds Paid. $ 24,000.00 31/2% $ 840.00 $4,000.00 Old Bonds. 6,000.00 3y2% 210.00 750.00 Old Bonds. 7,200.00 31/^% 252.00 800.00 Old Bonds. 18,750.00 4' % 750.00 3,750.00 Old Bonds. 175,000.00 4 % 7,000.00 Filtration Bonds. 18,000.00 4 % 720.00 Refunding & Met'r $248,950.00 $9,772.00 $9,300.00 B'nds DISBURSEMENTS. Bonds $9,300.00 Interest on Bonds 9,772.00 $19,072,00 Current expenses 10,500.00 Filter expenses . . . .- 5,500.00 Refunding . and Meter Bo-^ds 5,000.00 $40,072.00 RECEIPTS. Water Rates $33,000.00 Proceeds Refunding and Meter Bonds 11,000.00 Total receipts $44,000.00 Total disbursements 40,072.00 Say $4,000 available ? 3,928.00 1914 Prin. Rate. Interest. Bonds Paid. ? 20,000.00 31^% $ 700.00 |4,ODO.O0 Old Bonds. 5,250.00 3%% 183.00 750.00 Old Bonds. 6,400.00 ZVz% 224.00 800.00 Old Bonds. 15,000.00 4 % GOO.OO 3,750.00 Old Bonds. 175,000.00 4 % 7,000 00 Filtration Bonds. 29,000.00 4 % 1,160.00 Refunding & Met'r B'nds $250,650.00 $9,867.00 $9,300.00 DISBURSEMENTS. Bonds . •■• ■• $9,300.00 Interest on ' bonds' '.'.'.■.■.■ . ' .' 9.807.00 $19,167.00 Current expenses "'^'^'cnnnn Filter expenses ^nnnna Refunding and Meter Bonds "'"""■"" $40,107.00 ■£ REPORT RELATIVE TO NEW WATER SUPPLY. RECEIPTS. Water Rates ^S3,5QafiO ,-, Proceeds Refunding and Meter Bonds ll.OOO'.OO Total receipts |44,500.00 Total disbursements 40,167.00 Available I 4,333.00 1915. Prin. Rate. Interesit. Bonds Paid. ,$ 16,000.00 3%% $ 560.00 $4,000.00 Old Bonds. 4,500.00 3%% 157.00 750.00 Old Bonds. 5,600.00 3%% 196.00 800.00 Old Bonds. 11,250.00 4 % 450.00 3,750.00 Old Bonds. 175,000.00 4 % 7,000.00 Filtration Bonds. 40,000.00 4 % 1,600.00 Refunding & Met'r B'nds .$252,350.00 $9,963.00. $9,300.00 DISBURSEMENTS. Bonds >$9,300.00 Interest on Bonds 9,963.00 $19,263.00 Current expenses 10,500.00 Filter expenses 5,500.00 Refunding and Meter Bonds 5,000.00 $40,263.00 RECEIPTS. Water Rates $34,000.00 Proceeds Refunding and Meter Bonds 10,500.00 Total receipts $44,500.00 Total disbursements 40,263.00 Available $4,237.00 1916. Prin. Rate. Interesit. Bonds Paid. $ 12,000.00 3%% $ 420.00 $4,000.00 Old Bonds. 3,750.00 3%% 131.00 750.00 Old Bonds. 4,800.00 31/2% 168.00 800.00 Old Bonds. 7,500.00 4 % 300.00 3,750.00 Old Bonds. 175,000.00 4 % 7,000.00 Filtration Bonds. 50,500.00 4 % 2,020.00 Refunding & Met'r B'nds $253,550.00 $10,039.00 $9,300.00 DISBURSEMENTS. Bonds •. $9,300.00 Interest on Bonds 10,039.00 $19,339.00 Current expenses 10,500.00 Filter expenses 5,500.00 $35,339.00 RECEIPTS. Water Rates $34,500.00 •Proceeds Refunding and Meter Bonds • 5,000.00 Total receipts $39,500.00 Total disbursements 35,339.00 Available $4,161.00 REPORT RELATIVE TO NEW WATER SUPPLY. 9 1917. P>rin. Rate. Interest. Bonds Paid. I 8,000.00 3%% I 280.00 $4,000.00 Old Bonds. 3,000.00 3%% 105.00 750.00 Old Bonds. iMMO.- 3%% 140.00 800.00 Old Bonds. 3,750.00 4 % 150.00 3,7S0,«Old Bonds. 175,000.00 4 % • 7,000.00 Filtration Bonds. 55,500.00 4 % 2,220.00 Refunding & Met'r B'nds $243,250.00 $9,895.00 $9,300.00 DISBURSEMENTS. Bonds $9,300.00 Interest on Bonds -9,895.00 $19,W5.00 Current expenses 10,500.00 Filter expenses 5,500.00 $35,195.00 RECEIPTS. Water Rates $35,000.00 proceeds Refunding and Meter Bonds 4,500.00 Total receipts $39,500.00 Total disbursements 35,195.00 Available $ 4,305.00 1918. Prin. Rate. Interest. Bonds Paid. $ 4,000.00 3%% $ 140.00 $4,000.00 Old Bonds. 2,250.00 3%% 78.00 750.00 Old Bonds. 3,200.00 3%% 112.00 800.00 Old Bonds. 175,000.00 4 % 7,000.00 Filtration Bonds. 60,000.00 4 % 2,400.00 Refunding & Met'r B'nds $244,450.00 $9,730.00 $5,550.00 DISBURSEMENTS. Bonds $5,550.00 Interest on Bonds 9,730.00 $15,280.00 Current expenses 10,500.00 Filter expenses 5,500.00 $31,280.00 RECEIPTS. Water Rates $35,500.00 Disbursements 31,280.00 Available $ 4,220.00 1919. Prin. Rate. Interest. Bonds Paid. § 1,500.00 3%% $ 52.00 $ 750.00 Old Bonds. 2,400.00 3%% 84.00 800.00 Old Bonds. 175,000.00 4 % 7,000.00 Filtration Bonds. 60,000.00 4 % " 2,400.00 3,000.00 Refunding & Met'r B'nds ?238,900.00. $9,536.00 $4,550.00 10 REPORT RELATIVE TO NEW WATER SUPPLY. DISBURSEMBiNTS. Bonds : $4,550.00 Interest on Bonds 9,536.00 ?14,086.0(> Current expen&es 10,500.00 Finer expenses 5,500.00 $30,086.00 RECEIPTS. Water Rates $36,000.00' Total receipts $36,000.00 Distursements 30,086.00 Available '. $ 5,914.00 1920. Prin. Rate. Interest. Bonds Paid. $ 750.00 3%% $ 26.00 $ 750.00 Old Bonds. , 1,600.00 3%% 56.00 800.00 Old Bonds. 175,000.00 4 % 7,000.00 Filtration Bonds. 57,000.00 4 % 2,280.00 4,000.00 Refunding &■ Met'r B'nd& $234,350.00 $9,362.00 $5,550.00 DISBURSEMENTS. Bonds $5,550.00 Interest on Bonds 9,362.00 $14,912.00 Current expenses 10,500.00 Filter expenses 5,500.0T) $30,912.00 RECEIPTS. Water Rates $36,500.00 Total receipts $36,500.00 Disbursements 30,912.00 Available $ 5,588.00 1921. Prin. Rate. Interest. Bonds Paid. $ 800.00 3%% $ 28.00 $ 800.00 Old Bonds.. 175,000.00 4 % 7,000.00 Filtraition Bonds. 53,000.00 4 % 2,120.00 6,000.00 Refunding & Met'r B'nds $228,800.00 $9,148.00 $6,800.00 DISBURSEMENTS. Bonds $6,800.00 Interest on Bonds 9,148.00 $15,948. Off Current expenses 10,500.00 Filter expenses 5,500.00 $31,948.00 RECEIPTS. Water Rates $37,000.00 Total receipts $37,000.00 Disbursements 31,948.00 Available $ 5,052.00 REIPORT RBiLATIVB TO NEW WATER SUPPLY. 11 1922. Prin. Rate. Interest. Bonds Paid. $175,000.00 4 % $7,000.00 Filtration Bonds. 47,000.00 4 % 1,880.00 7,000.00 Refunding & Met'r B'nds $222,000.00 $8,880.00 $7,000.00 DISBURSEMENTS. Bonds $7,000.00 Interest on Bonds 8,880.00 $15,880.00 Current expenses 10,500.00 Filter expenses 5,500.00 $31,880.00 REiCBIPTS. Water Rates $37,500.09 Total receipts $37,500.00 Disbursements 31,880.00 Available $ 5,620.00 1923. Prin. Rate. Interest. Bonds Paid. $175,000.00 4 % $7,000.00 $ Filtration Bonds. 40,000.00 4 % 1,600.00 8,000.00 Refunding & Met'r B'nds $215,000.00 $8,600.00 $8,000.00 DISBURSEMENTS. Bonds $8,000.00 Interest on Bonds 8,600.00 $16,600.00 Current expenses 10,500.00 Filter expenses 5,500.00 $32,600.00 RECEIPTS. Water Rates • $38,000.00 Disbursements! 32,600.00 • Available $ 5,400.00 1924. Prin. Rate. Interesit. Bonds Paid. $175,000.00 4 % $7,000.00 $ Filtration Bonds. 32,000.00 4 % 1,280.00 9,000.00 Refunding & Met'r B'nds $207,000.00 $8,280.00 $9,000.00 DISBURSEMENTS. Bonds $9,000.00 . Interest on Bonds 8,280.00 $17,280.00 Current expenses 10,500.00 Filter expenses 5,500.00 $33,280.00 RECEIPTS. Water Rates $38,500.00 Disbursements 33,280.00 Available ? 5,220.00 12 REPORT RELATIVE TO NEW WATER SUPPLY. 1925. Prin. Rate. lateresit. Bonds Paid. $175,000.00 4 % $7,000.00 $5,000.00 Filtration Bonds. 23,000.00 4 % 920.00 5,000.00 Refunding & Met'r B'nds $198,000.00 $7,920.00 $10,000.00 DISBURSEMENTS. Bonds $10,000.00 Interest on Bonds 7,920.00 $17,920.00' Current expenses 10,500.00 Filrer expenses 5,500.00' $33,920.00 RECEIPTS. Water Rates $39,000.00 Disbursements 33,920.00 Available $ 5,080.00- 1926. Prin. Rate. Interest. Bonds Paid. 517C,000.00 4 % $6,800.00 $6,000.00 Filtration Bonds. 18,000.00 4 % 720.00 4,000.00 Refunding & Met'r B'nds- $188,000.00 $7,520.00 $10,000.00 DISBURSEMENTS. Bonds $10,000.00 Interest on Bonds 7,520.00 $17,520.00 . Current expenses 10,500.00 Filter expenses ; 5,500.00' $33,520.00 I^ECEIPTS. Water Rates $39,500.00 Disbursements 33,520.00 Available $ 5,980.00 1927. Prin. Rate. Interest. Bonds Paid. $164,000.00 4 % $6,560.00 $7,000.00 Filtration Bonds. 14,000.00 4 % 560.00 4,000.00 Refunding & Met'r B'nds $178,000.00 $7,120.00 $11,000.00 DISBURSEMENTS. Bonds , $11,000.00 Interest on Bonds 7,120.00 $18,120.00- Current expenses 10,500.00 Filter expenses 5,500.00 $34,120.00 RECEIPTS. Water Rates $40,000.00 Disbursements 34,120.00 Available $ 5,880.00- REPORT RELATIVE TO NEW WATER SUPPLY. 13 1928. Prin. Rate. Interest. Bonds Paid. $157,000.00 4 % $6,280.00 $7,000.00 Filtration Bonds. 10,000.00 4 % 400.00 4,000.00 Refunding & Met'r B'nds $167,000.00 $6,680.00 11,000.00 DISBURSEMENTS. Bonds $11,000.00 Interest on Bonds 6,680.00 $17,680.00 RECEIPTS. Water Rates $40,500.00 Disbursements 33,680.00 Available $ 6,820.00 1929. Prin. Rate. Interestt. Bonds Paid. $150,000.00 4 % $6,000.00 $9,000.00 Filtration Bonds. 6,000.00 4 % 240.00 3,000.00 Refunding & Met'r B'nds. $156,000.00 $6,240.00 $12,000.00 DISBURSEMENTS. Bonds ■. $12,000.00 Interest on Bonds .- 6,240.00 $18,240.00 Current expenses 10,500.00 Filter expenses 5,500.00 $34,240.00 RECEIPTS. Water Rates $41,000.00 Disbursements 34,240.00 Available $ 6,760.00 1930. Prin. Rate. Interest. Bonds Paid. $141,000.00 4 % $5,640.00 $9,000.00 Filtration Bonds. 3,000.00 4 % 120.00 3,000.00 Refunding & JMet'r B'nds $144,000.00 $5,760.00 $12,000.00 DISBURSEMENTS. Bonds $12,000.00 Interest on Bonds 5,760.00 $17,760.00 Current expenses 10,500.00 Filter expenses 5,500.00 $33,760.00 RECEIPTS. Water Rates $41,500.00 Disbursements ■•' 33,760.00 $ 7,740.00 14 REPORT RELATIVE TO NEW WATER SUPPLY. 1931. Prin. Rate. Interest. Bonds Paid. $132,000.00 4 % $5,280.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 5,280.00 $18,280.00 Current expenses 10,500.00 Filter expenses 5,500.00 $34,280.00 RECEIPTS. Water Rates $42,000.00 Disbursements 34,280.00 Available $ 7,720.00 1932. Prin. Rate. Interest. Bonds Paid. $119,000.00 4 % $4,760.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 4,760.00 $17,760.00 Current expenses 10,500.00 Filter expenses 5,500.00 $33,760.00 RECEIPTS. Water Rates $42,500.00 Disbursements 33,760.00 Available $ 8,740.00 1933. Prin. Rate. Interest. Bonds Paid. $106,000.00 4 % $4,240.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 4,240.00 $17,240,00 Current expenses > 10,500.00 Filter expenses 5,500.00 $33,240.00 RECEIPTS. Water Rates $43,000.00 Disbursements 33,240.00 Available $ 9,760.00 1934. Prin. Rate. Interest. Bonds Paid. $ 93,000.00 4 % $3,720.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 3,720.00 $16,720.00 Current expenses 10,500.00 . Filter expenses 5,500.00 , $32,720.00 REiPORT RELATIVE TO NEW WATER SUPPLY. 15 RECEIPTS. Water Rates $43,500.00 Disbursements 32,720.00 Available $10,780.00 1935. Prin. liate. Interest. Bonds Paid. $ 80,000.00 4 % $3,200.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 3,200.00 $10,200.00 Current ex]jenses 10,500.00 Filtei; expenses 5,500.00 $32,200.00 RECEIPTS. Water Rates $44,000.00 Disbursements 32,200.00 Available $11,800.00 193G. Prin. Rcite. Interest. Bonds Paid. $ 07,000.00 4 % $2,680.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 2,680.00 $15,080.00 Filter expenses 5,500.00 Current expenses 10,500.00 $31,680.00 RECEIPTS. Water Rates $44,500.00 Disbursements 31,680.00 Available $12,820.00 1937. Prin. Rate. Interest. Bonds Paid. $ 54,000.00 4 9"c $2,160.00 $13,C00.00 Filtration Bonds. DISBl'RSBMENTS. Bonds : $13,000.00 Interest on Bonds 2,160.00 $15,100.00 Current expenses 10,500.00 Filter expenses 5,500.00 $31,100.00 RECEIPTS. Water Rates $45,000.00 Disbursements 31,160.00 Available $13,840.00 16 REPORT RE3LATIVE TO NEW WATER SUPPLY. 1938. Prin. ' Rate. Interest. Bonds Paid. $41,000.00 4 % $1,640.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 1,640.00 $14,640.00 Current expenses 10,500.00 Filter expenses 5,500.00 $30,640.00 RBCEflPTS. Water rates $45,500.00 Disbursements 30,640.00 Available $14,860.00 1939. Prin, Rate. Interest. Bonds Paid. ? 28,000.00 4 % $1,120.00 $13,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $13,000.00 Interest on Bonds 1,120.00 $14,120.00 Current expenses 10,500.00 Filter expenses 5,500.00 $30,120.00 RECEIPTS. V/ater Rates $46,000.00 Disbursements 30,120.00 Available $15,880.00 194t). Prin. Rate. Interest. Bonds Paid. ? 15,000.00 4 % $600.00 $15,000.00 Filtration Bonds. DISBURSEMENTS. Bonds $15,000.00 Interest on Bonds .- 600.00 $15,000.00 Current expenses 10,500.00 Filter expenses 5,500.00 $31,600.00 RECEIPTS. Water Rates $46,500.00 Disbursements . - 31,600.00 Available $14,900.00 REPORT RELATIVE TO NEW WATER SUPPLY. it: m tu o o $ a III 1- < a -I o a z < 111 z o z o z < z IL IT O U. < _l a. o >• < to 'd'i'ea penssi nsqAv « Tpws MOH 'spuofl j3;aiM: s ^ Snipunjaa '^ 3 O CD CO US »0 ■^'' OCOcqt-COOSlOOCOWCSOOOOOOOOoOOOOOOOOOOO Tt^CDi^MOOOOt^li3^COrHa>CDC^cnWTHCO(r<|C-eqir-(?qi>cqcC)THCDrHCO asosoro5050os«osarosoooooo"i>^t-t-'ocDU3i^'^^*cocoo4'c^rT-i"r:? oooooooooooooooooooooooooooodo OOOOOOOU3U3U300000000000000000000 cococopsc'swffoi^uamoooooooooocjoooooooooo aso^o505050^o^u^•^llCcDI^-cx^ai"oOTHTH(^fec^oo'ofeo'col^^c'5cococolO THTHiHT-HrHT-irHrHrHT-tT-lrHT-tTHT-tTH OOOOOOO^OOOOOOOOOOoOOOOOOOOt^OO loiiStoiAwiA^^oinoooooooooooooooooooo WMOTCOCOU3_c^'«:t^OTMOOOOOOOOO^OOOOOOOOOOOO ■^-^■^USlOtCTj^-^COCOCsItMT-IOOSOOt-Om'^COTHOOSQOOUSTtiCqTH 6e- ooooooooooooooooooo OOOCJOOOOi^OOOOOOOOOO ooooiouaoooooooooo^ooo I> 00 OT o' o u^" o o t^ co" t>^ O e^" fo" 00 ■*'" O CD CO T-lt^t>t-t-t^t^I>b-t-I>t-t-CD10lO-<*30t--^<35COOO ■^ la" »0 CD t- C» OD Oi" CO Cvl t- CD lO "* CO W rH •QiraiT ^-^C^CO^t^WCDt>OOaiOr^CCICO'!*^mCDt^OOO^Or^CC^CO^lOCDt^OOOSO ^+^U. tHtHiHt— (T-HTHTHTHTHCSo°°° Beach Thoroughfare crossings, including material and labor, I3.S00-00 Venturi Meter, - 2,000.00 Connecting to pumping station and force mains, •- 1.160.00 Connecting to distribution system, material and labor, 12,500.00 4 REPORT OF SUPERINTENDENT. Gate Houses, 3,000.00 Unloading and hauling, 3,000.00 Laying 26,700 feet 30-inch cast iron pipe, @ ^.80, 21,360.00 Extra for creek crossings, 2,500.00 1185,877.93 Engineering, inspecting and contingencies, ^}i per cent., i3;940.84 ^199,818.77 The plan upon which this estimate is based, pro- vides for a 30-inch cast iron main with two lines, each 20 inches in diameter, under Beach Thorough- fare, the main entering the city along North Missis- sippi Avenue, connecting with the stand pipe at Ohio and Baltic Avenues by a 20-inch main on Baltic Ave- nue, and continuing on Mississippi Avenue 24-inch to Arctic Avenue, 20-inch to Atlantic Avenue, and 16- inch to the Beach main at Columbia Avenue. The Beach Thoroughfare crossing will be made by two. lines of 20-incli ball joint pipe laid on saddles sup- ported on creosoted piles. The existing 12-inch and 20-inch force mains are cast iron with the usual lead joint every twelve feet. As these mains are laid parallel with and close to a railroad track for about four miles on the meadows, the jar of the trains is constantly straining and damaging the non-elastic lead joints, necessitating daily inspection and very frequent repairs. With a 30-inch pipe the weight, 4,000 to 4,200 pounds per length of twelve feet, would be so great as to cause more or less settlement and the consequent damage to the main. Another matter to be considered in connection with a cast iron pipe, is the deterioration of the metal. Where the present mains are bedded in the salt mud of the meadows, it is found in many places that a chemical change has taken place in the metal to a depth of one-eighth of an inch, and that that part REPORT OF SUPERINTENDENT. 5 of it can be cut like chalk. This action is probably caused by the galvanic action of the iron, the free carbon of the metal and the salt water, in which we have all the elements of an electric battery. We do not know the rate of this deterioration, nor whether it is increasing or diminishing. The ordinary coal tar coating which is put on cast iron pipe affords little protection, and hence the estimate includes cost of first-class asphalt coating. In making this estimate I have planned on lay- ing the pipe so as to barely cover it with a layer of sod. It will be connected at various places with the existing force mains and provided with a sufficient number of valves so that in case of break or serious leak, it will only be necessary to close out one section of the main for repairs. RIVETED STEEL PIPE. This has been used for mains in several places, and is cheaper in sizes over thirty inches in diameter, than cast iron. There is probably very little differ- ence in cost, however, between a 30-inch main of cast iron and a 30-inch steel main when the difference in carrying capacity is considered. Owing to the cer- tainty of damage to the coating of a steel pipe in lay- ing, the absolute impossibility of efficient repairing and inspection in a pipe only thirty inches in diam- eter, the rapid corrosion which will follow any exposure of the metal to the action of the salt mud of the meadows, and also to the fact that the carrying capacity of a 30-inch steel pipe after a few years service is about fifteen per cent, less than that of a cast iron pipe of the same size, I do not think it worth while to consider it for the proposed force main. WOODEN STAVE PIPE. In the West within the past twenty-five years, a pipe, built in place, of wooden staves banded with 6 REPORT OF SUPERINTENDENT. round or oval steel or wrought irou bands, has been extensively used, and were intelligently constructed, with excellent results. Denver, Col.; Butte, Mont.; Astoria, Or.; San Francisco, Cal.; and many other cities and villages are supplied with water through stave pipes, and many power plants and mines also have it in use. A similar pipe has been used in the East for mill flumes for from 25 to 50 years. A pipe is now in use in Manchester, N. H., which was built in 1872. It is six feet in diameter, about 600 feet long, and is under from 12 to 40 feet head. It has never leaked and has needed no repairs. At Wicopee, Dutchess Co., N. Y., on Fishkill Creek, there was, in 1893, a 30-inch stave pipe which had been in use for power purposes for 45 years. It was reported at that time as in good condition and good for many more years of service. At Portsmouth, N. H., in 1797, bored pine logs were laid for water pipe, having five inches internal diameter. The last of these was taken out after having been in service for about 75 years, and the wood was still sound. There are numerous other cases which could be cited. With Commissioners Reiley and Booye, and Councilman Long, Chairman of the Committee of Protection of Property and Improvement of City Council, I recently made a tour of inspection in the West, paying particular attention to the materials and methods of construction of wooden stave pipe, and to the condition of pipes which had been in service for some years. At Butte, Mont., there were ten miles of 24-inch stave pipe which had been in service nine years, and over twenty miles of 24-inch and 26-inch stave pipe are now being constructed. We saw the 26-inch stave pipe in process of con- struction Redwood staves were being used, banded with is-inch steel bands, which were coated with Min- eral Rubber. We noted particularly the speed with which the pipe was built, the strength and excellent 3 •a o 3 REPORT OF SUPERINTENDENT. 7 quality of the finished pipe, and we were greatly im- pressed *with the ease and rapidity with which the pipe can be repaired. We were extended every facil- ity for examining the work by Mr. Payne, the Bn- gineer in charge of the work, and we also visited the old line of 24-inch stave pipe. We found the staves of this pipe as good as new, and the bands, saddles, and nuts were also in excellent condition, after nine years of continuous use. At Seattle, Wash., the city is now constructing about twenty-two miles of 42-inch to 50-inch pipe. We saw the pipe being built. Washington Fir is being used for staves, and J^-inch steel bands are being used, the latter coated by being dipped in hot asphalt. Photographs were taken of both the Butte and the Seattle pipe. At Seattle we were accorded every courtesy by Messrs. C. P. Allen & Son of the contracting firm, and City Engineer R. H. Thomson. Among other things we were shown a 48-inch wooden stave pipe used for a sewer under most adverse con- ditions, being laid so that it is nearly covered at high tide, and thus exposed to the action of drift and ice, and entirely exposed to the air at low tide. To make these conditions worse, the sewer ordinarily flows from one-half to two-thirds full. This sewer was built seven years ago, and the fir staves are in ex- cellent condition. The bands are only very slightly rusted, not enough to be damaged at all, though only dipped in hot asphalt. At San Francisco we met Mr. D. C Henny. C. E., Manager and Engineer, of the Excelsior Wooden Pipe Co., from whom we obtained a number of photographs and much information on the sub- ject of stave pipe, especially where laid in salt marsh, as Mr. ' Henny 's Company has laid several such pipes. He stated that when those pipes had been examined, after being in service some years,they were in good condition. Mr. Henny mentioned 8 REPORT OF SUPERINTENDENT. one case where a steel riveted pipe laid in salt marsh had given out in a short time, and been re- placed by wooden stave pipe with success. Near Floriston, Cal., we saw from the train a stave pipe, nine feet in diameter, about one-half a mile in length. At Denver, while we were unable to see any of the pipe, the city has over 50 miles of stave pipe which is stated to be in good condition after eleven years of constant and satisfactory service. We visited Mr. C. P. Allen's office in Denver, and saw many samples of staves, bands, saddles, etc., and photographs of pipe lines. While on the trip we secured such information as. to cost of materials and construction as we were able, and the following is an estimate of the cost of a 30-inch wooden stave force main calculated on the same basis as the estimate of the 30-inch ca;st iron main. Under Beach Thoroughfare I have figured on two lines of 20-inch cast iron ball joint pipe laid on saddles supported on creosoted piles : 26,700 feet 30-inch Wooden Stave Pipe, at $3.55, $ 94,785.00 Valves and bolts for connections, iS>3S7-93 Special castings, - 3>550-73 Beach Thoroughfare crossing, including materials and labor, 13,500.00 Venturi Meter, 2,000.00 Cast iron pipe at pumping station, and laying, 550.00 Connecting to force mains, 1,000.00 Gate Houses, 3,000.00 Materials and connections to Distributing System, 12,500.00 Extra for crossing Adam's Ditch and Jonathan's Thoro- ughfare, 1,700.00 Unloading and hauling Gate Valves, etc., 500,00 Engineering inspection and contingencies, 10 per cent. r , 14,844.37 ;?i48,443-66 14,844-37 $163,288.03 Sketch showing Tongue and Slot in End of Stave for making Brett Joint. Section of Wooden Stave Pipe, showing Construction. REPORT OF SUPERINTENDENT. 9 You will note that compared with the cast iron pipe, the wooden stave pipe will cost $36,530.74 less. A comparison of the qualities of the wooden stave pipe with those of cast iron for our conditions, are almost wholly in favor of the former : viz : (a) The wooden stave pipe as built is contin- uous, while the cast iron has a non-elastic lead joint at least every 12 feet. Laid as our mains are, close to railroad tracks, this is an important matter as already pointed out. This has been demonstrated by our experience with the existing force mains. {d) The durability is the only point where cast iron might have the advantage over wooden stave pipe, and as already shown, even this is questionable under the conditions on the meadows. The staves are made of a thickness that insures their always being kept saturated by the w^ater under pressure within, and it is a well known fact that wood which is kept wet and free from contact with air, will never rot. Hence it is the life of the bands which will determine the life of the pipe. To coat and protect these bands with some suitable preservative, is but a matter of careful, painstaking, honest inspection, and will be far more easily accomplished than the coat- ing of heavy cast iron pipe. The stave pipe is de- signed with a safety factor of 4-5^ or 5 and should a few of the bands be destroyed, it would be an easy and inexpensive matter to replace them. (c) The wooden stave pipe is much less liable to breaks or leakage than a cast iron pipe. It is so de- signed that it cannot burst. Before a pressure suffi- cient to burst the bands could be applied, the staves would be forced back against the bands, forcing the bands into the staves and opening the joints between the edges of the staves, thus causing leakage, which would relieve the extra pressure. (d) The wooden stave pipe can be repaired with far greater ease than cast iron. If a band has lO REPORT OF SUPERINTENDENT. given out, it is but a few minutes work to replace it. If it is a damaged stave or staves, it is only neccessary to loosen a few bands, slip them back on tbe pipe, pry out and cut off the stave or staves to be replaced, cut new stave or staves to proper length, slot the new ends and put in the tongues, buckle staves into place and replace bands. This ease of repair is a very important matter in our case. Anyone who is familiar with the mud and other conditions on the meadows will realize at once the difficulty of repairing a 30-inch cast iron pipe which involves handling lengths of pipe weighing 4,000 to 4,200 pounds, as well as the further difficulty of using melted lead for making joints. Even with the present 20-inch main, in making repairs it is frequently impossible to run a joint with melted lead, and the only thing that can be done is to pound in cold lead which does not make a good joint. (e) The difference in weight is, under the con- ditions on the meadows, very much in favor of the wooden stave pipe. As before stated, there would be difficulty and expense in preventing serious settle- ment of a 30-inch cast iron pipe in the meadow mud. (/) The carrying capacity of wooden stave pipe is considerably greater than that of cast iron pipe, on account of the smoothness of the inner sur- face of the pipe, and it does not decrease as does that of cast iron pipe. Experiments show that for 30-inch pipes when new, the wooden stave pipe will carry 10 per cent, more than cast iron ; after a few years when the cast iron pipe has become tuberculated, the difference increases to about 40 per cent. For the reasons above enumerated, I am lead to recommend the use of wooden stave pipe for the pro- posed force main, as the most suitable and reliable for the purpose, as well as the cheapest in first cost, and also in cost of maintenance. This recommen- dation is endorsed by your Consulting Engineer, REPORT OF SUPBRINTENDKNT. II Mr. Emil Kuichling, member of the American Society of Civil Engineers. In connection with this matter I wish to call your attention to the following plan which was sug- gested to me by Mr. D. C. Henny, while in conver- sation with him in his office in San Francisco. When the present plant at the Absecon pump- ing station was constructed, Atlantic City had a population of .about 13,000 people. When the city purchased the plant in August 1895, the population was about 18,000. To-day it is about 28,000, and the summer crowd, which is what tests the capacity of the works, has increased in even a greater ratio. With the prospects of a continued growth, and the inability to furnish an adequate supply which we are now experiencing, it becomes evident that con- siderable improvements and additions must be made in the near future at the pumping stations. The first thing to be done is to develop and in- crease the underground supply. The existing suction well is not of sufficient depth and is too far from the basins. The old plank flume has about rotted out and must soon be replaced or abandoned. There is no room in the present pumping station for any additional pumps or boilers. The suction lift of the pumps, as now set, is greater than it should be. It will, hence, be but a short time before it will be necessary to make radical changes to the plant. A new suction well will have to be sunk nearer to the basins and the timber flume replaced by a pipe of proper size. When this is done, the present machinery should be moved into a new and larger pumping station, and two new high duty pumps, one 5,000,000, and one 3,000,000 gallons capacity in 24 hours, with the necessary boilers, be added.. The present 5,000,000 high duty pump should have new and larger plunger rings and plun- gers. 12 REPORT OF SUPERINTENDENT. Ill developing the underground supply, it will probably be necessary to use centrifugal or low lift reciprocating pumps to lift the water from a greater depth than the pressure pumps will reach. It will make very little difference in cost of oper- ation whether the water is merely lifted to the pumps as they now stand, or whether it is lifted to a suffi- cient height to cause it to flow by gravity across the meadows to a pumping station in the city. A wooden stave pipe to carry the water thus under very low pressure would however cost materially less than a main such as I have figured upon to carry the water under 75 to 50 pounds pressure, and there would be this very g-reat advantage, that with the plant con- structed and operated in this way, nearly all dan- ger of leaks or failure of the force mains on the meadows would be avoided. As there are frequently times during storms or high tides when the meadows are entirely inaccessible in case repairs should be needed, this is a considera- tion to which I would urgently call your attention, especially in view of the danger from fire in a city containing so many frame buildings, in case of any failure of the water supply In preparing the estimate which follows, I have assumed that it will cost no more to develop and in- crease the underground supply at the Absecon pump- ing station, whether the pumps are tbere or in the city. The city owns enough ground on North Kentuck}' A ve- nue in this city, where the present Consumers Pump ing Station is located, for the construction of a new pumping station, and both plants could be installed in one station and operated by one set of men. If, therefore, we do not now consider the cost of the new machinery, which will be the same in either locatiop of the pumping station, the only increased cost due to change of pumping station to Atlantic City will be the increased length of force main, the tanks to REPORT OF SUPERINTENDENT. 1 3 needed for storage, and the change in distribution system necessary for making connection to it. Part of the latter will be necessary in two or three years, whether the change in location of pumping station is made or not, and hence I have shown the cost of that separate. The estimate contemplates a force main of a maximum capacity of 10,000 000 gallons in 24 hours, under a pressure of 27 pounds at the main land, delivering into a tank at elevation 12 at North Kentucky Avenue. From the point on this side of Beach Thoroughfare, where Hummock Avenue pro- duced, intersects the present force mains, I have figured on a 36-inch wooden stave pipe along Hum- mock Avenue to Kentucky Avenue, thence on Ken- tucky Avenue to the pumping station. The neces- sary new mains to connect with the distribution mains would be a 20-inch line on Mediterranean Ave- nue from Kentucky to Ohio Avenue, on Ohio Ave- nue from Mediterranean Avenue to Baltic Avenue, connecting there with the stand pipe thence on Bal- tic Avenue to Missouri Avenue connecting with the present 20-inch force main there There will also be needed a main east from Kentucky Avenue on Mediterranean Avenue with cross mains to Pacific Avenue on. Virginia Avenue and Connecticut Ave- nue. The cost of this has been added to the esti- mate under the item " Extension of Distribution System." The following is the estimate of the cost of a wooden stave pipe built as proposed in this plan : 25,575 feet of 30-inch Wooden Stave Pipe, at $2.40 per foot, - ^61,380.00 3,300 feet of 36-inch Wooden Stave Pipe at $z-5° P^^ foot, - 11,550.00 Valves and bolts for connections, 14.857.93 Special castings, 3.55°- 73 Beach Thoroughfare Crossings, including material and labor, 13,500.00 Venturi Meter, - - - - 2,000.00 14 REPORT OF SUPERINTENDENT. Cast iron pipe at pumping station and laying, SS°-oo Connecting to force mains, 1,000.00 Gate Houses, - 3,000.00 Extra for crossing Adams Ditch and Johnathan's Thoroughfare, 1,700.00 Unloading and Hauling Gate Valves etc., 500.00 Connecting the Distribution System, 7,450.00 1121,038.66 Engineering, Inspection and Contingencies, 10 per cent. 12,103.87 $133,142.53 Extension of Distribution System, 9,400.75 $142,543-28 You will note that this shows a saving of $20, 744 75 compared with the high pressure wooden stave pipe, and $57,275.49 compared with the cast iron main. Owing to the limited time which I have had in which to consider this matter, I cannot give the exact cost of the proposed changes, but the following is an approximate estimate : Probable cost of developing underground supply at Absecon Pumping Station, including new machinery and alterations to pumping station, from $15,000 to $35,000.00 New pumping station on North Kentucky Avenue, including moving and resetting pumps and boilers, in- cluding new boiler, 27,000.00 Two steel tanks, zoo feet in diameter and 15 feet high, with foundations and pipes, or masonry basins of equal capacity, 26,000.00 Changes to present Consumers Pumping Station, 2,000.00 30,000.00 If these changes are made, the total expendi- ture will be about $230,000 to $235,000. The re- duced liability of breaks and leakage in the force mains on the meadows, and the resulting security against a water famine, is worth many times, to the REPORT OF SUPERINTENDENT. 1 5 Cit)'- of Atlantic City, the $30,000 or $35,000 which it will cost over the original estimated cost of the force main. It is probable, however, that the saving in cost of operation of the proposed plant, compared with the present plant, will more than offset this ad- ditional cost. I therefore recommend the building of a low pressure wooden stave pipe, with the develop- ment of the underground water supply at the Abse- con Pumping Station, the change in location of the pumping station, and the necessary changes in the distribution system, as outlined above. Very respectfully submitted, .. W. C. HawlEy, Superintendent. REPORT OF CONSULTING ENGINEER Atlantic City, N. J. September 3, 1900. To thf. Honorable^ The Board of Water Commis- sioners of Atlantic City^ N. J. : Gentlemen : — In accordance with your request of the 31st ult., to present to your Board and the Com- mittee on Protection of Property and Improvement of City Council a written review of the Report made on the 24th ult., by Mr W. C. Hawley, C. E., Sup- erintendent of the Atlantic City Water Works, in relation to the proposed new water conduit from Absecon Pumping Station to the pumping station at the corner of Kentucky and Mediterranean Avenues, the undersigned begs leave to submit the following : SIZE AND QUALITY OF CONDUIT PIPE. It appears that the diameter of the proposed conduit has been fixed by the Ordinance of May 28th, 1900, at 30-inch, and that the main question now is about the material of which it shall be constructed. As stated in the report, the pipe is to be subjected to more or less internal water pressure, and hence the choice of material is restricted to cast iron, riveted steel plate, and wooden staves held together by steel bands, all of which are well adapted to the purpose under ordinary conditions. In the present case how- ever, the conditions are somewhat unusual, inasmuch as nearly the entire length of the conduit will be ex- posed on the outside to the action of sea water, and that special care must therefore be taken to protect 1 8 REPORT OF CONSULTING ENGINEER. the metal from corrosion ; also that, for the greater part of the distance of five and six-tenths miles be- tween the two pumping stations, the pipe will be located on the salt marsh in close proximity to one or two lines of railway, the heavy traffic on which produces severe tremors or vibrations in the adjoin- ing soft ground. These vibrations tend to loosen the lead joints of cast iron pipes, and to promote leakage therefrom which will ultimately cause settlements and break- age of the pipe itself; and the use of flange joints in such pipes is impracticable on account of their rigi- dity. Furthermore, the present 12-inch and 20-inch cast iron pipes across the marsh exhibit great deterioration after only a few years by external cor- rosion, as pointed out in Mr. Hawley's report ; hence it becomes expedient to provide for the most efficient protective coating in the proposed new work. The application of such a coating, however, will obvious- ly increase the price of the pipe somewhat, and this brings its cost to the figures named in said report. Excluding the trench and the timber foundations that may be required for the heavy pipe in many soft places in the marsh, the cost of the 30-inch cast iron pipe in place may be taken at $4.60 per lineal foot. Riveted steel pipe, made of plates ^ inch thick and with either riveted or flanged joints, is much better adapted to lajnng in the soft marshy soil than the cast iron pipe, and affords much more security against damage from the tremors or vibrations in the vicinity of the railroads. It is, however, open to the serious objection of very rapid corrosion when ex- posed to salt water, and hence the utmost care must be taken in preparing and applying the protective coating. From an extensive experience with such pipe, the undersigned deems it expedient to use plates not less than ^ inch thick, and to require REPORT OP CONSULTING ENGINEER. 1 9 the application of at least two thick bituminous coat- ings on the outside, the second coating to be further guarded against damage during transportation and handling by a wrapping of burlap. All abrasions or injuries to the exterior surface must be scrupulously- repaired before the pipe is covered in the trench, otherwise active rusting by the salt water will make the life of the conduit comparatively short. There are, however, so many chances of injury to the protective coating, and the care needed during the progress of the work to repair such damage is so great, as to render it practically impossible to secure a perfectly continuous protective coating, especially on the lower part pf the pipe which is obviously the most diflB.cult of access ; hence, it is very probable that in spite of all the vigilance that may be exercised in the supervision of the work, more or less corrosion and consequent leakage will soon develop in the con- duit. This may be expected to continue indefinitely, as each exposure for subsequent repairs affords an opportunity to create new abrasions of the coating, and hence the cost of maintenance will probably at- tain a considerable magnitude. The first cost of such a 30-inch riveted steel pipe, laid in place, but not including the trenching and back-filling, may be estimated at $3.60 per lineal foot. Wooden stave pipe, made of sound and clear fir, redwood, or southern pine lumber, and held together by asufiicient number of steel bands to resist with ample safety, both the internal water pressure and the expansive force due to the swelling of the wood from the absorption of moisture, is also well adapted to the purpose under consideration. Many of the advantages and details of constructing such conduits have been set forth so clearly in Mr. Hawley's report that little further need be added thereto here. The only addition which appears to be required is a reference to the protection of the wood against the 20 REPORT OF CONSULTING ENGINEER- attacks of the teredo and other marine animals or or- ganisms. On this point it may be mentioned that a thorough examiaation of a large number of telegraph and telpehone poles which have been set for many years in the salt meadow, has demonstrated that at and below the surface of the ground this wood is still per- fectly sound and entirely free from such attacks ; also, that an inquiry into the habits of the teredo and limno- tia, which are the most active of the marine boring animals, has disclosed the fact that these creatures cannot live in muddy or stagnant water, and hence never penetrate into the soil of such marshes. Fur- thermore, as the wood of the pipe is . constantly satu- rated with fresh water and is kept from exposure to the weather by a thick- covering of meadow sod, its great durability may fairly be regarded as definitely es- tablished. The only question about the life of the wooden conduit is that which relates to the durability of the steel bands which hold the staves together. On this point it may be said that all which can be ad- duced in favor of protective coatings for steel plate and cast iron, applies with even greater force to the protection of the bands or hoops, as it is much easier to secure proper inspection and coating of a small rod than of a larger surface. Every band can be care- fully examined before it is put in place, and can even be immersed in a bath of the preservative mixture immediately before it is applied to the pipe, so as to insure a perfectly continuous coating. The pipe also possesses considerable flexibility, which enables it to be built on blocking at a covenient distance above the bottom of the trench, thereby afifording easy access to the lower side for recoating or repainting all portions of the bands, whereupon the blocks are finally removed and the pipe is lowered into place. No other excuse than gross negligence can be' offered by an inspector if good coating workmanship is not REPORT OF CONSULTING ENGINEER. 21 secured ; and hence if reasonably intelligent and con- scientious men are employed as inspectors, there will be a far greater probability of securing a durable conduit in this case than with either cast iron or riv- eted steel. With respect to the cost of a 30-inch wooden stave conduit, it is proper to remark that the esti- mates of the undersigned are somewhat less than the figures given in Mr. Hawley's report.* Allowing for a reasonable margin of profit, similar to that which was used in the estimates for the cast iron and riveted steel pipes, the prices which are here considered fair for the work, exclusive of the trenching and back-fil- ling, are $3.15 per lineal foot, for such a wooden pipe adapted to withstanding an average internal water pressure of 69 pounds per square inch, and $2.05 per lineal foot for one to withstand an average pressure of only 13 pounds per square inch. The difference in these prices is mainly due to the increased number of bands per lineal foot required to resist the higher pres- sure, and the two pressures mentioned relate respect- ively to the original plan of pumping at the Absecon station directly into the city standpipes, and to the modified plan, proposed by Mr. Hawley, of delivering the water from said station at a height of only about 1 2 feet above the surface of the ground at the City Pump- ing Station, and then repumping it to the required height into the standpipe. It may also be added, that in view of the various uncertainties which affect the strength and durability of both the cast iron and the riveted steel conduit in this case, the undersigned does not deem it expedient to make any material reduction in the thickness or weight of the metal for the pipe which is subjected to the lighter pressure. For t^e comparison of cost with the light pressure wooden conduit, the above * To avoid confusion, the estimates given in Mr. Hawley's report as here prin- ted have been revised to correspond to the prices given by Mr. Kuichling. 22 REPORT OF CONSULTING ENGINEER. named estimates for the cast iron and riveted steel pipes will accordingly be retained. It should fur- thermore be noted that the items for valves, valve houses, special castings, water measuring device, Beach Thoroughfare crossing, extra work at Small Creek Crossings, connections with pumping stations and existing force mains, and connections and exten- sions of the distributing system, are all practically the same in amount for each kind of conduit in Mr. Hawley's estimates, as will obviously be the case , hence the difference in the cost of the entire work is determined almost wholly by the difference in the cost of the 30-inch pipes as given above, it being remem- bered that the required length of such pipe is approxi- mately 26,700 feet. We thus find for the cost of said pipe alone, ex- clusive of the trenching and back-filling, the follow- ing amounts : 26,700 lineal feet 30-inch cast iron pipe laid in place, . . . . . . . ^4.60 ^122.820 26,700 lineal feet 30-inch riveted steel pipe laid in place, 3.60 96.120 26,700 lineal feet 30-inch low pressure wooden stave pipe, 2.05 54-735 26,700 lineal feet 30-inch high pressure wooden stave pipe, 3.15 84.105 From these figures it is seen that the cast iron pipe is the most expensive, and that in comparison therewith, the wooden stave pipe will produce a sav- ing of from $68,085 to $38,715, according as the low or high pressure conduit plan is adopted. In com- parison with riveted steel pipe, on the other hand , the use of the wooden stave pipe will result in a sav- ing of respectively $41,385 and $12,015. It is accord- ingly evident that the wooden stave pipe is by far the most economical material for this conduit, and there is every reason to believe that in respect to REPORT OF CONSULTING ENGINEER. 23 both, serviceability and durability, it will prove supe- rior to either of the other materials. The under- signed therefore agrees fully with Mr. Hawley in this respect As to the size of the proposed new conduit, it may be remarked that the diameter of 30 inches is adapted to the most economical delivery by steam pumping of a volume of about 10,000.000 gallons of water at a uniform rate during 24 hours While this quantity may n,ot be required at the present time, yet judging from the rapid rate of increase of the summer popu- lation, not many years will elapse before it will be needed, and hence the selection of size has been a wise and judicious one. HIGH OR LOW PRESSURE CONDUIT. In the choice between the high and low pressure conduits, a variety of conditions must be taken into consideration. At its best, the long conduit over the salt marsh, and for the greater part of its length in close proximity to lines of railway with a very heavy traffic, and also liable to be more or less deeply covered by sea water several times per year whereby repairs will necessarilly be delayed, cannot be re- garded otherwise than as a relatively insecure struct- ure, which should be subjected to as little internal stress and shock as possible. From this point of view the low pressure conduit must be given the preference. On the other hand, with a high pressure conduit and an abundance of excellent water ob- , tained at Absecop, the pumping station in the city might be abandoned and an appreciable saving in the annual operating expenses might thereby follow. With wooden pipe the difference in cost between the high and low pressure cqnduit is $29,370 accord- ing to the foregoing figures, and at four per cent, interest this sum will produce annually about $1,175. In contrast thereto must be placed the expense of 24 REPORT OF CONSULTING ENGINEER. maintaining the attendants at the low lift pumping station at Absecon, since the expenses of the high lift station, and the aggregate coal consumption, will remain practically the same in either case. This expense will be about $3,375 per year, as both the power developed and the necessary number of men are relatively small ; hence the difference in annual operating expenses will be about $2,200, or in other words, it will cost ab lut $2,200 per year more to maintain the two pumping stations with the low pressure conduit, than a single pumping station at Absecon with a high pressure conduit. It should also be remarked that to this sum should be added the annual interest and depreciation on the cost of the low. lift pumping plant ; but as most of the ma- chinery needed for the purpose can doubtless be adapted without much additional cost from that which is already on hand at Absecon, it will hardly be fair to make said sum much larger than $2,500. The question thus presents itself in the follow- ing form : — Is it worth $2,500 per year to have the greater security from accident to the long conduit which the low pressure plan provides ? In view of what has already been said on the subject, and the further fact that so many of the buildings in the city are of wood and therefore inflammable, it seems to the undersigned that it would be eminently prudent to reduce the fire risk as much as possible by mak- ing the conduit as safe as possible, and this can best , be accomplished by adopting the low pressure plan. The additional annual expense is not large, and will doubtless be more than compensated by a reduction in the insurance rates if the case is properly presen- ted to the underwriters ; and even if such a reduction is not granted, the knowledge that greater depend- ence can be placed on the water supply which comes from the mainland cannot fail to be of far more value than its additional cost to the property owners of a city REPORT OF CONSULTING ENGINEER. 25 whose continued prosperity is so closely allied .to means for preserving the safety, health and comfort of the visiting public. The undersigned has there- fore no hesitation in recommending the adoption of the low pressure conduit. Aside from the foregoing considerations, how- ever, there are other reasons for adopting this plan. One of these is that there is a strong demand for a purely artesian water supply on the part of many citizens, which has resulted in the continued main- tenance of the city station plant at relatively large cost. Now, with the proposed new conduit, the pro- vision of a large additional quantity of water is di- rectly associated, and as will be pointed out below, there is a good probability that an ample volume of such artesian water can be obtained near the Abse- con station. Should the search for such water be successful, there will then be no reason for the fur- ther independent maintenance of the city station plant, and the supply now obtained at this point may either be treated as a reserve, or it can be operated without expense for attendance by the same set of men who will operate the high pressure station. Another reason is that it will sooner or later result in the provision of a large storage of water within the city for use in case of accident to the standpipe. By the first, there can be no reduction in annual operating expenses ; and by the second, a greater safety against a shortage of water will be secured. An ample reserve supply of potable water is of enor- mous value to every large community, and it is obvious that the nearer such reserve is to the con- sumers, the better, especially if there is any doubt as. to the stability of the conduit. Furthermore, if adequate storage capacity is provided at the City Pumping Station, it will be possible to reduce the expenses for attendance at the low pressure station by employing a single set of men for about nine 26 REPORT OF CONSULTING ENGINEER. months of the year, and two sets during the summer. A saving of at least $i.ooo per year can thus be effec- ted which represents the interest at 4 per cent, on a capital of $25,000 ; and this capital can according- ly be expended in securing storage without adding to the previous annual charges. COST OF RESERVOIR. With respect to the quantity of water to be stored at the City Pumping Station, it may be said that the same should be at least equal to one day's consump- tion during the greater part of the year when the temporary population is absent. Assuming the per- manent population at 40,000 and allowing a con- sumption of about 80 gallons of water per head per day, we have total daily use of 3,200,000 gallons. This quantity may be increased somewhat during large conflagrations, so that a storage capacity of about 3,500.000 gallons should be provided, which represent a volume of about 470,000 cubic feet, con- tained in three rectangular basins of concrete mason- ry each 130 feet long. 100 feet wide and 12 feet deep, or in four circular steel tanks 104 feet diameter and 14 feet deep. The estimated cost of such storage capaci- ty is about $50,000, including wooden roof, by either plan of construction ; and the space required therefor is a lot 365 feet long by 175 feet wide, which is the size of the lot now owned by the city opposite the pumping station at the corner of Kentucky and Mediterranean Avenues. QUANTITY OF WATER. As already stated in the foregoing, the proposed 30-inch conduit from Absecon is adapted to the econ- omical delivery of 10,000,000 gallons per day, and the question now arises whether this quantity of water can be permanently obtained at the locality. The minimum delivery of the present wells and REPORT OF CONSULTING ENGINEER. 27 canal at Absecon may be estimated at about 5,000,000 gallons per day, although this limit has not yet been reached ; and from the deep wells at the City Pump- ing Station a probable ultimate minimum of i ,000,000 gallons per day may be anticipated, as it is not likely that a further extensive development of this subter- ranean source will be attempted. The total probable minimum dry weather yield of the two sources to- gether is therefore about 6,000,000 gallons per day. Hence, to satisfy a prospective maximum demand of 10,000 000 gallons per day, an additional daily supply of at least 4,000,000 gallons must eventually be secured from the mainland at Absecon. On the other hand, if the subterranean supply at the City Station is treated as a reserve, as previously suggested, then an additional supply of 5,000,000 gallons per day must be obtained from the mainland. To secure this further quantity, it will be neces- sary either to provide adequate storage on Absecon Creek, and thus deliver into the city mainly surface water, or to undertake an extensive development of the underground supply which appears to be available. In view of the growing general prejudice against surface water supplies, and the large cost of storage, the former course should not be adopted until it has been demonstrated by proper exploration of the sub- soil that a satisfactory ground water supply cannot be secured. From the experience gained with the ex- isting wells, ho-w^ever, there is good reason to believe that a large additional quantity of ground-water can be obtained at moderate depth and cost in the vicinity of the Absecon Pumping Station ; hence it is recom- mended that steps be taken to determine experiment- ally, whether such is the fact. The methods of making this examination have been fully discussed with Mr, Hawley, and need not be recited here. If an adequate volume of ground-water is found the entire problem will be solved in the most economical 28 REPORT OF CONSULTING ENGINEER. manner, and the independent operation of the City Station plant can be permanently discontinued, or considered as a reserve. In relation to the new pumping engine that may be required with the increased supply, little further need now be said, except that if centrifugal pumps are selected for the low pressure station, they should be of the best possible design, as otherwise a good reciprocating pump will give a higher duty. So far as can now be determined, the efficiency of the former class of pumping machinery is on the whole a little higher than that of the latter at the average low lift which will here prevail ; but if the engines are operated at higher speed for a shorter time each day after the aforesaid reservoir capacity is provided in the city, it will probably be found on further in- vestigation that the reciprocating pumps are best suited to the work. As to the cost of securing a large additional un- derground supply at the Absecon Station, nothing very definite can now be given, as everything depends on the location and magnitude of the water bearing stratum. From such data as are available, it is pro- bable that tubular wells of at least six inch diameter must be driven to a depth of about 90 feet below the surface of the ground, in order to penetrate suflS- ciently far into the saturated subsoil, and that the permanent minimum yield of such a well will be about 150,000 gallons per day. On this basis, not less than 33 such wells will be required to secure the desired 5,000,000 gallons per day; and as it will doubtless happen that some of the number will occa- sionally become choked and require cleaning, it will be prudent to estimate that fully 40 wells must be provided. Furthermore, the distance between the wells may be taken at about 200 feet, so that the ^gg^^g^te length of piping required in the whole system will be about 12,000 feet, of which perhaps REPORT OF CONSULTING ENGINEER. 29 2,000 feet will be 36 inclies in diataeter, and 7,000 feet will be 6 inclies in diameter, while the remainder will be of various intermediate sizes. With all the necessary appurtenances, such a system of under- ground water supply will cost, under favorable cir- cumstances, from $40,000 to $50,000, exclusive of the pumping plant, and hence, before it is undertaken, a thorough study of the subsoil should be made. Concerning the development of surface water storage along the course of Absecon Creek, the data now at hand are too meagre to warrant the formula- tion of any valid conclusions. The drainage area of the stream above the head of the existing water wtarks canal is about 14 square miles, and with an average rainfall of about 45 inches thereon, it is reasonable to assume that the required volume of storage water is available. The best places and costs of impound- ing, however, must yet be determined, and this can be done only after proper surveys have been made. Respectfully submitted, E. KUICHLING, Consulting Engineer. REPORT TO THE : Hon. Samuel H.^/lsH bridge : lyiayior^Qf the C^itj^ 6i^ Philadelphia- ON THE EXTENSION AND IMPROVEMENT OF THE AT^ER Supply OF THE 0TY OF PHILADELPHIA BY RUDOLPH HERING,*,/ JOsiPH: M. WILSON SAMUEL M. GRAY ' COMMISSIONERS PHi1.APEl.PHi4 189a ■ " CONTENTS PAGE Introdtjction 6 Population , 6 Existing Conditions 7 Delaware River 7 Schuylkill Eiver 10 Present Works 16 Pumping Systems 17 Pumping Stations, Maobinery, eto 19 Bepairs Immediately Needed at Works 22 Cost of Operation 25 Water ConsuDiption 27 Deficiency of Local Supply 31 Condition of Eeservoirs 32 Use of Beservoirs 33 Storage 33 Sedimentation 34 Distribution 34 Quality of Water 37 Future Supply 38 Character of Present Supply 40 Projects Specially Investigated 41 Filtered vs. Mountain Water r 41 Quantity of Water Available 44 Artificial Purification 47 Water Filtration 47 Filter Plants and Various Projects Examined 55 Becent Beports on filtration Experiments 62 Projects Presented 68 Mountain Water Supplies 73 Filtered Water Supplies 74 RrauME AND Conclusions 82 LIST OF PLATES. Plate I : Watersheds of Delaware and Lehigh Rivers, Perkiomen and TohickoQ Creeks, and the Aqueducts. Plate II : Profile of Delaware Aqueduct, Water Gap to Kintnerville. Plate III: Profile of Delaware Aqueduct, Kintnerville to Philadelphia. Plate IV: Profile of Lehigh Aqueduct, White Haven to Aquanchicola Creek. Plate V : Profile of Lehigh Aqueduct, Aquanchicola Creek to Treich- lersville. Plate VI : Profile of Perkiomen High-level Aqueduct. Plate VII : Profile of Perkiomen LciW-level Aqueduct. Plate VIII; Typical Aqueduct S-eclions. Plate IX : Locations of Filter Plants and Mains recommended for Im- mediate Relief. Plate X : Belmont Filter Plant. Plate XI : Roxborough Filter Plant and Queen Lane Filter Plant. Plate XII : East Park Filter Plant. Plate Xlli : Torresdale Filter Plant. Plate XIV : Plan and Sections of Typical Slow Filter. Plate XV : Details of Sand Washers and Regulating Apparatus. PlaseXyi: Growth of Populations of Various Cities. Percentages of Increase of Populations of Various Cities, by decades, with Probable Rates of Future Increase for the City of Philadelphia. Per capita Water Consumption of Vari- ous Cities. Probable Increase of Population of Phila- delphia. Plate XVII : Effect of Meters on the Water Consumption of Various Cities. OFFICE OF THE COMMISSION ON THE EXTENSION AND IMPROVEMENT OF THE WATER SUPPLY OF THE CITY OF PHILADELPHIA Philadelphia^ September ISth, 1899. Hon. Samuel H. Ash bridge, Mayor of the City of Philadelphia. Sir : — In accordance with a resolution of the Select and Common Councils of the City of Philadelphia, adopted April 20th, 1899, authorizing your Honor to select and employ three experts as a Commission to consider and re- port upon the question of the improvement and extension of the Water Supply for the City, acting in conjunction with the Director of the Department of Public Works and the Chiefs of the Bureaus of Water and Surveys, your Honor was pleased to name for this purpose, the under- signed, Messrs. Rudolph Hering, Joseph M. Wilson, and Samuel M. Gray, all of whom, after being duly advised, made acceptance of their appointment. At a conference with the Commission, held on May 9th, the various questions to be considered were very fully discussed, and your Honor outlined what was de- sired, afterwards incorporating the instructions in a letter as follows : — [Copy] OFFICE OF THE MAYOR PHILADELPHIA Messrs. Rudolph Hering, Joseph M. Wilson AND Samuel M. Gray, Commission to Investigate the Extension and Improvement of Philadelphia's Water Supply. May 18, 1899. Gentlemen : — The Director of the Department of Public Works, Mr. William C. Haddock, has informed me that you desire the remarks which I made at our conference on Tuesday of last week, as to the scope of your work, reduced to writing, with a view to the incor- poration of the same, in your report. It gives me pleasure to send you herewith as nearly as I can recall, what I said upon that occasion. The resolution under which your appointment was made passed the Select and Common Councils on April 20th of this year, and is as follows : " Whereas, The quality of water furnished by the " municipality is such, as to require purification by filtra- " tion or otherwise, and " Whereas, There are on record in the Bureau of " Water authoritative and exhaustive plans, surveys and "reports heretofore drafted and made by various com- " missions, engineering experts and departmental officials " dealing with the water supplj"^ and upon which no " action has been taken. Therefore be it "Resolved, by the Select and Common Covmcils of the "City of Philadelphia, That the Mayor be, and is hereby "authorized and directed to select and employ three "experts to act in conjunction with the Director of the "Department of Public Works, and the Chiefs of the "Bureaus of Water and Survey, to take. up the question " of the immediate improvement and extension of the "water supply, provided that a preliminary report shall "be made by said experts within sixty days of their "appointment, and the final report shall be presented " not. less than three months thereafter, so that it may be 3 " presented to Councils immediately after the summer " recess in September. SfC 3fC 57C S^C 3fC 5fC 0J[i 3^ *JC 5fC In outlining what was desired, I took the foregoing reso- lution as the basis for instructions. Your work naturally divides itself into three great problems, First, — What is necessary for the immediate betterment of our water system. Secondly, — If the remedy be filtration, what is the best method, and Thirdly ,^In what direction is it most desirable to extend our present supply, so that for years to come, the water problem may not give anxiety to our people. Your first duty will be to take up the question of the immediate relief of present conditions. I would request that you visit all of our pumping stations and reservoirs and make a thorough investigation of everything apper- taining thereto. I would ask you to take into consider- ation and compare the waters of the two rivers, ascer- taining, if possible, and approximating the decrease in the flow of the SchuylkiU River. ^,, It would also be well to consider the Schuylkill Valley as a region of great industrial activity, where manufac- turing enterprises must increase in the years to come, with the danger of pollution of the Schuylkill, the natural drain of the Valley, as a consequence. Is it, therefore, wise to continue the Schuylkill as the principal source of our supply ? Secondly : — If your investigations show that filtration is the remedy for the conditions which confront us, I would then ask that you recommend a system, calling attention, however, to the fact that slow sand filtration is generally recognized as the best. I would ask a careful estimate as to the cost of installation of a plant, sufficient to filter the entire supply of the city, with the annual cost for maintenance per million gallons daily, this esti- mate 10 include depreciation and renewals of plant. In the cost of the installation, I would ask that the proposed sites of the filtering beds be specified, together with the approximate cost per acre. I would ask you to take into consideration whether the severity of the temperature of the weather in Philadelphia, during the winter months, would not necessitate the erection of roofs over the filter- ing beds, and how much additional this would cost. It is also a question whether filtered water could re- main in the subsiding basins any length of time, without deterioration. It is but proper for me to call your atten- tion, officially, at this time, to the constant and increasing growth of our municipality and the extension of its busi- ness enterprises, and would ask you to take into consid- eration, in connection, with filtration, whether that system could be extended to accommodate within a reasonable degree, the people of Philadelphia for the next half century. . ■ The third and final duty which devolves upon you under the resolution is to look forward to the- future. Personally, I would like to see a lasting solution of the water question. For forty years this city has been con- fronted with a problem, both as respects quality and quantity. Measures which brought about temporary, relief alone have been instituted, but the question has never ceased to be a vexatious burden to succeeding gen- erations. I would ask you to consider the present con- sumption of water, its increase and how it probably will increase in the future, and to recommend what means should be introduced to give the City of Philadelphia an ample supply, without unnecessary restrictions, for at least fifty years to come. Your suggestion, I might add, should be broad enough to provide for a subsequent extension of the system pro- posed, so that the water problem could be solved for at least a century. Should you be able to cover this vast problem, so as to make a final and definite report, to be laid before Coun- cils when they assemble in the Fall, you will have per- formed a magnificent work, not only for this generation but for generations yet unborn. Your task is no light one. It has puzzled the best engineering brains for one hundred years or more, and if you solve the question, it will not only redound to your honor, but increase the health of the community, advance all its varied interests, and be a monument forever to the credit of Philadelphia. It will strengthen and stimulate our people in all their various lines of activity, and place us far in advance of all other cities on this continent, if not in the world. Very respectfully yours, [Signed] SAMUEL H. ASHBRIDGE, Mayor. These instructions, presenting to us a problem of vast proportions, both as to time and scope, placed upon us a great responsibility, in view of the few months at our dis- posal. As your Honor, however, kindly granted us im- mediately all the assistance we required, it has been pos- sible to cover the ground with such care that we can now report our conclusions on at least the essential features of the problem, with the fullest confidence that they present the means of providing a supply of water for the City of Philadelphia equal to the best, and capable of being se- cured at a reasonable cost. r At meetings had from time to time with your Honor, as occasion required, various niodifications in the details of procedure were brought up and discussed, in further amplification of the subject. All of these modifications have been duly considered and acted upon. The resolution authorizing our appointment required that, in considering the subject placed before us, we were to act in conjunction with the Director of the Department of Pubhc Works, Mr. William C. Haddock; the Chief of the Bureau of Water, Mr. John C. Trautwine, Jr., and the Chief of the Bureau of Surveys, Mr. George S. Web- ster. These gentlemen have aided us by furnishing de- sired information during the entire investigation, and we take pleasure in acknowledging our indebtedness on this account. Special mention in this respect should be made concerning the assistance rendered by the Chief of the Bureau of Water, who supplied us in ready and conveni- ent form not only with the material already available in his office, but also with information prepared at our re- quest and collected by him elsewhere. We desire also to acknowledge our indebtedness to the many engineers, in this country and abroad, who have furnished us with valuable data. The shortness of time allotted to us made it advisable, in some instances, to draw upon the special experience of certain engineers in other cities. For instance, to verify our estimates of cost for the aqueducts from the Blue Mountains and for the filter plants to be located in this city, we obtained the co-operation of Mr. Charles S. Gowen, Principal Assistant P^lngineer of the Croton Aque- duct Commission, New York, and of Mr. William B. Fuller, late Principal Assistant Engineer of the Albany Filter Plant. Finally, we desire to mention, with credit, the services of Mr. James H. Fuertes, Civil Engineer, who with much skill, untiring devotion and unusual dispatch, collected for us most of the' information we required. A preliminary report of progress was submitted on July 3rd, in accordance with your instructions, and we have now the honor to present our final report. INTRODUCTORY. It is proper that we preface our report with a l)rief review of the present works, since you have asked our suggestions as to the measures which should immediately be taken respecting the improvement of the existing plant. Population. The City of Philadelphia contains at the present time about 1,300,000 inhabitants. It is difficult to make a reliable estimate of the future increase in population, as much depends upon local conditions. After a careful study of the rate of increase in the past, and a compari- son with the growth of other cities, a conclusion has been reached. Plate XVI illustrates how this was obtained. One diagram shows the actual growth of the city to 1890^ another shows the percentage of increase of population by decades, including the probable percentage of future in- crease, and a third diagram shows the probable population as far into the future as the year 1950. This last diagram gives, for the year 1925, a maximum of 2,250,000 and a minimum of 1,900,000 persons, and for the year 1950 a maximum of ;5,300,000 and a minimum of 2,700,000 persons. We have thought it proper to provide at present for a population of 1,300,000 persons and also to make esti- mates of cost for works serving 3,000,000 persons. Existing Conditions. The present supply is taken from |.he Delaware and Schuylkill rivers within the geographical limits of the city, about 95 per cent, of it from the Schuylkill, and the remaining 5 per cent, from the Delaware. All of the water is pumped, normally to reservoirs, artificially con- structed on elevated sites, from which it is distributed through cast iron pipes. Certain elevated suburban dis- tricts are supplied by high-service pumping stations, drawing from the reservoirs. Delaware River. The Delaware river rises in the southeastern part of the state of New York, about 190 miles above Philadel- phia, and for about 130 miles of its course, before reach- ing Philadelphia, it forms the boundary between the states of Pennsylvania and New Jersey. The upper part of its watershed is mountainous and very sparsely settled. At the Delaware Water Gap it breaks through the Blue Mountain, and from that point downward its course becomes generally less precipitous, until, below Trenton, it is a tidal estuary, flowing through alluvial flats, gener- ally under a high state of cultivation, but with no very large towns. The area of the water-shed is about 8,600 square miles. The principal affluent of the Delaware, above Philadel- phia, is the Lehigh, which flows through the anthracite coal fields of the same name, and reaches the Delaware at Easton, 75 miles above Philadelphia. The area of the Lehigh water-shed is 1332 square miles. The largest towns upon the banks of the Delaware, above Philadelphia, are Easton, Pa., 14,500 population, 75 miles above Philadelphia ; and Trenton, the capital of New Jersey, 57,500 population, 30 miles above Philadel- phia. But little manufacturing is carried on along the banks of the Delaware above Philadelphia ; but the Lehigh val- ley is a very important iron manufacturing district. The principal source of pollution of the Delaware, as a source of water supply for Philadelphia, is the city itself. The river, within the city limits, being tidal, the water taken is subject to pollution from the city's sewers, several of which discharge large quantities of sewage at points within a mile of the present intake, both up and down stream. The Delaware receives from the Lehigh, at Easton, the drainage of the Lehigh anthracite region, and, at the same time, passes through a broad limestone belt, traversed also by the Schuylkill. Below Trenton it receives the wash- ings of the alluvial plain through which it flows. Its vol- ume, however, is so large that the effects of these influ- ences are seldom very strongl)' marked. The water of the Delaware, at Philadelphia, is quite soft, forming no scale in boilers. During the drought of 1895, the water of the Delaware, 9 opposite Market street, Philadelphia, was sensibly brackish at high tide, but no such effect was noticed at the pump- ing station at Lardner's Point. The Delaware river water above the Water Gap is of fair quality. The water from its mountain tribu- taries, between the Gap and Bushkill, is equal to the best of mountain waters, but the river itself brings down the pollutions from Port Jervis and other towns situated on its banks above. Below the Gap the river is aug- mented by tributaries, and particularly by the Lehigh river, which brings considerable pollution from Easton, Bethlehem, Allentown and other cities. Further down, the river, by its increase of size, becomes relatively less polluted, but it receives the sewage from a number of towns as far down as Philadelphia. Owing to the large volume of flow, the dilution of the impurities is much greater than is the case with the Schuylkill water. The Delaware water contains only about one-third the amount of sulphate of lime found in the Schuylkill water, which accounts for the superior softness of the former, and, therefore, its advantage for use in boilers. The Lehigh river, below White Haven, is not only much contaminated with sewage, but also, for a considerable portion of its length, with coal refuse. The territory east of the river between the Lehigh Gap and White Haven, and the entire water-shed of the river above White Haven, furnish water free from natural and artificial impurities, and, like the tributaries of the Delaware above mentioned, would provide an excellent mountain water supply. The Tohickon creek flows into the Delaware at Point Pleasant, and Big and Little Neshaminy creeks enter the Delaware river below Bristol. The waters of these streams would, require storage and filtering. Lack of time pre- vented a reconsideration of these projects on the basis of filtering their supply. They were fully described in the 10 annual reports of the Bureau of Water from 1883 to 1886. The Tohickou creek furnishes a better water than either of the Neshaminy creeks, but, in our opinion, with the light that is now available regarding the sanitary dan- gers of raw water obtained from populated districts, and with the known efficacy of filtration, we do not now con- sider these streams available sources without a prior filtra- tion. It may be that they can be made available in the future, by filtering their water and delivering it into the reservoirs of the City by gravity at an elevation of 170 feet above city datum. The same may be said of the lower Perkiomen sources. Schuylkill River. The Schuylkill river, which joins the Delaware at the southern extremity of Philadelphia, rises in the anthracite coal region of Schuylkill county, Pennsylvania, about one hundred miles aboV^e Philadelphia. In its generjally south- easterly course, it intersects most of the ridges and geo- logical formations traversed by the Delaware below the Water Gap. The principal towns on the Schuylkill are : Distance above Fbiladelpbla. Miles. Population, 1890. 80 60 36 26' 22 14,120 53,660 1.3,280 8,500 Pottstown Phoenixville 19,790 The water-shed area of the Schuylkill is about 1,915 square miles. At Port Clinton, the Schuylkill breaks through the Blue Mountains, forming a gap similar to the Delaware Water Gap. Below that point it intersects generally the 11 same ridges and the same geological formations traversed by the Delaware. The Schuylkill is tidal as far up as Philadelphia, where the flow of the tide is stopped by the Fairmount Dam. The range of tide in the Schuylkill at Fairmount, below the dam, is about 6 feet. Fairmount dam furnishes water power for the seven turbines at Fairmount station. These are the only water power works employed by the city. Many years ago the navigation of the Schuylkill was improved by the construction of a series of dams and canals, forming a slack water navigation extending from Philadelphia to Pottsville, a distance of about 100 miles. Its navigation is still in use from Philadelphia to Port Clinton, a distance of about 77 miles. It is under the control of che Philadelphia and Reading Railway Com- pany, lessee of the Schuylkill Navigation Compan3^ The available storage capacity of all the existing pools above Philadelphia, Is 563,t)00,000 cubic f^et,- or about 4,200,000,000 gallons. With the exception of the lowest, or Fairmount Dam, which was built by the City of Philadelphia, all the works were built by the Schuylkill Navigation Company, in the early part of the present century. The Schuylkill flows through a densely populated re- gion, and its banks are dotted by important manufactur- ing towns. The anthracite coal mines pour into the river large quantities of sulphuric acid, which, however, is com- pletely, neutralized by the subsequent passage of the river through the limestone beds which it intersects, so that the water, upon reaching Philadelphia, has a hard or alkaline reaction, and it gives considerable trouble by the formation of scale when used in the boilers. In 1885 the city, for the better protection of the water 12 of the Fairmount pool, constructed an intercepting sewer along the east bank of the river, and this sewer now re- ceives the discharge from the numerous and extensive mills from Manayunk within the city limits, and pollu- tion from many other sources, and discharges it into the river below the Fairmount Dam, or below the pools from which the city pumps its water. The ordinary low. summer flow of the Schuylkill river at Philadelphia is about 200,000,000 gallons per day, and the extreme minimum flow should, perhaps, for safety, be taken as low as 150,000,000 gallons per day. When the Schuylkill river was originally selected, the population of the city was small. Its buildings were grouped along the Delaware river, extending hardlj' half way across to the Schuylkill river, which then furnished an excellent supply. Few towns existed along its banks, and the water was almost uncontaminated by sewage, by the output of coal mines or by other impurities. Time has changed all this. For some years the development of the coal regions caused no perceptible injury to the water, although a considerable quantity of acid from the mines mingled with it ; thanks to the limestone beds, already mentioned. Of late years, however, water is being very extensively employed in the coal breakers as an aid in the separation and cleaning of the coal, and the mountains of culm, which have been accumulating for years about the mines, are now being worked over by a similar process, with the result that a large amount of coal dust is washed down the river to the city, especially in freshets. The progressive settlement and cultivation of the coun- try have brought another increasing trouble. The plough- ing up of the soil allows the rains to. wash quantities of mud into the streams, until the water, originally clear and palatable, is now often most objectionable, being re. pulsive in appearance and unpleasant to the taste. 13 A vital trouble, however, has arisen from the great in- crease in population and from the almost unrestricted disposal of much of the sewage and the refuse of manu- factories, these matters entering directly into the streams, until the water, in its present condition, has become dan- gerous to health and unfitted for domestic use. In 1884-5 a sanitary survey of the Schuylkill valley was made by the Water Department, and the reported facts and figures indicate that this pollution was serious even then, or fifteen years ago. Fairmount Park was originated and laid out for the purpose of controlling the banks of the river and protecting the water from pollution. Since then, further efforts have been made for the improvement of the conditions within the park, mainly by the construction of the intercepting sewer along the east bank of the river. But there is at present no adequate protection against pollution of the river beyond the City line, and matters remain very much as they were years ago. An investigation of the river reveals contaminations of the most abominable kind. On one of our recent visits we saw at least a dozen privies discharging their contents directly into the Schuylkill river between Conshohocken and Norristown.* A systematic examination of the Schuylkill water, un- dertaken two years ago by the Bureau, has since been regu- larly continued. Samples bave been taken twice a week at the intake of the Spring Garden Pumping Station, and a determination has been made of the organic and mineral constituents. The results furnished give informa- tion of practical value on the subject of purifying the water for the city's use. A comparison between the samples taken from the in- take, from the reservoir, and from the outlets of service * See also : A joint report of the Philadelphia and Fennsylrania Boards of Health on the sources of pollution of the Schuylkill river from Philadelphia to Beading, 1898. Illustrated. 14 pipes, shows a gradual diminution in the amount of or- ganic matter in the water ; in other words, an improvement after leaving the river. Observations in regard to this matter have almost uniformly given the same results, due, no doubt, to sedimentation, to oxidation, or to both. When the river is frozen over, the water is generally clearer and brighter than at~other times, and the amount of organic matter is less. This would lead to the infer- ence that a large amount of impurity is usually received from the washings of the surface of the country from rain water. The amount of suspended matter in the Schuylkill and Delaware waters has been determined by Mr. James H. Eastwick, of the Bureau of Health. The results are given in Appendix I to this report. The greatest amount of suspended matter is that reported on March 13, 1899, when it reached 1,026 parts per million parts of water. While this amount appears excessive, and represents very muddy water, yet it does not equal that found in the Ohio river water, which is stated by Mr. G. W. Fuller, in his Cincinnati report, to have been as great as 2,333 parts per million, or over twice the amount which is considered very objectionable in the Schuylkill river. This large amount of suspended matter indicates that at times a filtration of the Schuylkill water must neces- sarily be preceded by sedimentation, to remove most of the suspended matter, and that at periods of great tur- bidity even a coagulant would be of great service and economy, for the purpose of more rapid and thorough pre- cipitation. The tables show that the Schuylkill river clears up rather rapidly. For instance, the suspended matter was reduced from 1,026 parts to 130 parts per million in seven days. ■15 At low stages of the Schuylkill river,. 80 parts of alka- line sulphates per million have been found in the water, indicating the degree of its permanent hardness. During freshets the quantity diminishes to about one-third of this amount. Samples of very turbid Schuylkill water, treated with sulphate of alumina, have shown a greatly accel- erated sedimentation. The mineral constituents of a sample of the Schuylkill water, taken at the intake of the Spring Garden station, October 29, 1898, were : Sulphate of Lime and Magnesia 80 parts per million. Carbonate of Lime and Magnesia ' 60 " " Chloride of Sodium 10 " " 150 " Those of a sample taken on March 13, 1899, were : — Sulphate of Lime and Magnesia 24 parts per million. Carbonate of Lime and Magnesia 38 " " Chloride.of Sodium 4.64 " " 66.64 " These analyses give approximately the extremes in mineral constituents, the first having been taken during a long drought and the second after heavy rains. The principal tributaries of the Schuylkill river, above Reading, on account of their contamination from the coal regions, are not available as a water supply. One of its lower and larger tributaries, the Perkiomen creek, in the upper 120 square miles of its water-shed, furnishes a good mountain water. Its water is rendered but slightly tur- bid by rains, and it is comparatively free from sewage pollution. The water from lower branches of the Perki- pmen creek, such as the West Swamp creek and the North East Branch is highly discolored after rains, and also much more polluted, on account of the larger population residing on their watersheds. 16 Present Works. The plant of the Philadelphia Water Works, one of the largest in the world, comprises : (1) Thirty-seven pumping engines, with a rated total pumping capacity of 399,040,000 gallons daily. Of these pumps, thirty are operated by steam, and seven (turbines) by water power. (2) Eleven reservoirs, two stand-pipes and three tanks, with a combined capacity of 1,417,860,000 gallons. (3) A system of more than 1,250 miles of water mains, varying in diameter from 4 to 48 inches, together with the necessary valves, fire hydrants, etc. (4) Fairmount and Flat Rock dams, the latter belong- ing to the Schuylkill Navigation Company. The pumps vary in capacity from 250,000 to 30,000,000 gallons each per 24 hours, and are contained in ten pump- ing stations, five of which obtain water directly from the Schuylkill river and one tVom the Delaware river ; four are high-service pumping stations. The five stations located on the Schuylkill river are the Fairmount, where pumping is done by water, Spring Garden, Belmont, Queen Lane and Roxborough, where pumping is done by steam. The station on the Delaware river is at Frankford, where pumping is done by steam. The four high-service stations are at Belmont, Mount Airy, Chestnut Hill and Roxborough. Of the five Schuylkill stations, four, viz. : Fairmount, Spring Garden, Belmont and Queen Lane, take their water from the lowest or Fairmount pool, formed by the Fairmount dam, while the remaining, or Roxborough station, takes its water from the Flat Rock pool, formed by the Flat Rock, dam, belonging to the Schuylkill Navi- gation Company. 17 Pumping Stations. Fairmount Pumping Station. — The station contains seven turbine wheels, with pumps having a combined capacity of 33,290,000 gallons in 24 hours, lifting the water into Fairmount reservoir. The new or lower house contains : No. 1 turbine with 2,000,000 gallons daily capacity. No. 3 turbine with 5,330,000 gallons daily capacity. No. 4 turbine with 5,330,000 gallons daily capacity. No. 5 turbine with 5,330,000 gallons daily capacity. The old or upper house contains : No. 7 turbine with 5,100,000 gallons daily capacity. No. 8 turbine with 5,100,000 gallons daily capacity. No. 9 turbine with 5,100,000 gallons daily capacity. All the pumps excepting Nos. 1 and 3 are so connected that they can pump also to Corinthian Reservoir, having a slightly higher elevation. Nok 1 and 3 can pump only into Fairmount Reservoir. Spring Garden Pumping Station. — This station contains forty-four boilers and nine steam pumping engines, giving a total daily capacity of 170,000,000 gallons. When the water in the river is clear, No. 8 Worthington engine pumps directly into the Queen Lane distribution system. The old or upper house contains the following pump- ing engines : No. 5 Southwark, with 20,000,000 gallons daily capicity. No. 6 Simpson, with 10,000,000 gallons daily capacity. No. 7 Cramp, with 20,000,000 gallons daily capacity. No. 8 Worthington, with 10,000,000 gallons daily capacity. No. 11 Gaskill, with 20.000,000 gallons daily capacity. The new or lower house contains : No. 9 Worthington, with 15,000,000 gallons daily capacity. No. 10 Worthington, with 15,000,000 gallons daily capacity. No. 2 Holly, with 30,000,000 gallons daily capacity. . No. 3 Holly, with 30,000,000 gallons daily capacity. 2 18 Belmont Pumping Station. — At this station there are nineteen boilers and four Worthington duplex pumping engines, aggregating 38,000,000 gallons daily pumping capacity. The water pumped is delivered into Belmont reservoir. The house contains the following pumping engines : No. 1 Worlhiijgton with 5,000,000 gallons daily capacity. No. 2 Worthington with 5,000,000 gallons daily capacity. No. 3 Worthington with 8,000,000 gallons daily capacity. A rough wooden shed contains : No. 4 Worthington with 20,000,000 gallons daily capacity. Queen Lane Pumping Station. — Here there are twenty- four boilers and four Southwark vertical triple expansion engines, with single acting pumps, having a total daily capacity of 80,000,000 gallons. Roxhorougli Pumjnn^' Station. — At this station there are nineteen boilers and three pumping engines, with a total capacity of 24,500,000'gallons daily : No. 1 Sonthwark with 12,000 gallons dally capacity. No. 2 Worthington with 5,000,000 gallons daily capacity. No. 3 Worthington with 7,500,000 gallons daily capacity. The Worthington pumps Nos. 2 and 3 are duplex and lift directly into the old reservoir, and, when this is full, into the new reservoir. The No. 1 Southwark is a cross-compound bell-crank engine with vertical steam and horizontal water end, and usually forces its water directly to Germantown and Mount Airy reservoirs. Some of the water flows into Mount Airy reservoir and is pumped into the high-ser- vice mains. It is probable that the pumps do not pump over 75 per_cent. of their rated capacity. The city has contracted for four new Worthington 19 pumping engines, each to be of 5,000,000 gallons daily- capacity. Frankford Pumping Station. — This pumping station has twelve boilers and three pumping engines with an aggre- gate capacity of 42,000,000 gallons daily. The pumps are : No 1 Cramp with 10,000,000 gallons daily capacity. No. 2 Wetherill with 10,000,000 gallons daily capacity. No. 3 Southwark with 22,000,000 gallons daily capacity. The Southwark is a vertical cross-compound engine. It was tested by a Venturi meter and by a weir measure- ment, showing less than 5 per cent, loss by slip. The Cramp engine is vertical cross-compound and is in good order. It has not been tested, but it is stated that the slip is not more than 2 per cent. The Wetherill is a horizontal Corliss cross-compound engine. When tested by a Venturi meter and by a weir, it indicated less than 2J per cent, loss by slip. rThis engine is more sensitive to a variation of pressure in steam or water than the other two engines. With a loss of pressure in steam of five pounds or an increase in pressure of water of but a few feet, the engine will slow down. Belmont High-service Pumping Station. — This station has two pumps, No. 1 Worthington, with 2,000,000 gallons daily capacity, and No. 2 Snow, with 500,000 gallons daily capacity, and it has four boilers. The Worthington engine was bought in 1869. A con- tract has recently been made for a new 5,000,000 gallon Worthington engine. These pumps can take water either from Belmont reser- voir or from the mains leading to it. They pump directly into a standpipe near by. Roxborough High-service Pumping Station. — It is located near the old reservoir, and contains a Worthington pump- 2iU ing engine of 5,000,000 gallons daily capacity and four boilers. It takes water either from the old or from the new reservoir, as the case may be, and pumps it into a standpipe near by, from which it flows to the water tower at Chestnut Hill. The pump at this station was first put in service in 1871 at Otis street wharf for the Kensington Water Works, to pump Delaware water into the Lehigh reser- voir. The city has contracted for a new 5,000,000-gallon pumping engine. Mount Airy High-service Station. — This station adjoins the Mount Airy reservoir, and contains four boilers and three pumping engines, having an aggregate capacity of 3,000,000 gallons daily, as follows: No. 1 Davidson, with 1,000,000 gallons daily capacity. No. 2 Davidson, with 1,000,000 gallons daily capacity. No. 3 Knowles, with 1,000,000 gallons daily capacity. These pumps take their water from Mount Airy reser- voir and deliver into the water tower at Chestnut Hill. Chestnut Hill High Service Station. — This station con- tains one Knowles pumping engine, with a capacity of 1,000,000 gallons; one Worthington pump, with a capacity of 500,000 gallons, and one boiler. A list of the City reservoirs and standpipes, with their capacities, depths and elevations, is given in Appendix II. Pumping Systems. The Frankford station, on the Delaware river, forces its water through two parallel mains, 4J miles long, to the Wentz Farm reservoir, at an elevation of 167 feet, which supplies Frankford and other districts in the north- eastern portion of the city. The Lehigh reservoir, at Sixth street and Lehigh avenue, is filled by gravity from the Wentz Farm reser- voir, but owing to its low elevation, it is only occasion- ally used for distribution. 21 Fairmount station, with its seven turbines, pumps into the Fairmount reservoir, close by, with an elevation of 94 feet, and into Corinthian reservoir, distant about one- half mile, with an elevation of 120 feet. These reservoirs supply a small district comprising the lowest levels iu the city, and the elevation of Fairmount reservoir is so slight that, although usually kept full of water, it is now used only occasionally for distribution. Spring Garden Pumping Station, about half a mile above Fairmount, pumps about one-half of all the water consumed by the city, sending it into the largest, or East Park, reservoir, about one-half mile distant, with a capa- city of 689,000,000 gallons. This reservoir supplies the greater part of the city proper. A portion of the supply from Spring Garden goes also to the small Spring Garden reservoir, for the exclusive supply of Girard College, which is in the vicinity of the reservoir. > The Belmont works pump the entire supply of West Philadelphia, sending it into the Belmont reservoir, dis- tant about one mile, with an elevation of 212 feet. From this supply a special 12-incli main crosses the river for the exclusive supply of the City Hall, at Broad and Market streets. The Queen Lane Pumping Station, with the reservoir of the same name, distant one mile, was constructed for the relief of what was formerly known as the "direct pumpage district," which was supplied from the Spring Garden station. The Queen Lane system, although not yet completed as originally designed,^ is now in full ser- vice, and direct pumpage is now used only occasionally, and never when the river is in relatively bad condition. The Roxborough station, drawing from the upper, or Flat Rock pool, pumps into the old and new Roxborough reservoirs, for the supply of Roxborough, Manayunk, Chestnut Hill and Germantown. 22 The nominal pumping capacity of all the works pump- ing from the rivers is nearly 381,000,000 gallons, but the fact that the pressures throughout the city are low, and that incipient water famine exists at many points, show that the actual pumpage capacity is barely equal to the demand. Repairs Immediately Needed. Fairmount. — The roof of the upper house leaks freely, so that it is impossible to protect the pumps and ma- chinery from injury by rust. The internal appearance of the house is so unsightly that it is deemed proper to exclude the public from it. The roof over the lower pump-house will shortly need renewal. The forebay leading from the river to the flumes which conduct the water to the turbines is obstructed by a deposit of sand and mud in that j portion where the velocity of the water is reduced. Th6 form of the forebay, as seen in plan, should be changed as proposed by the Bureau of Water. Spring Garden Pumping Station. — Both bell-cranks of No. 5, Southwark Engine, have been broken and renewed, and both pump chambers are now badly cracked in spite of patches applied for the purpose of preserving them. In No. 7 Cramp engine the pump cylinders are cracked and should be repaired. The large new Holly engines, Nos. 2 and 3, are crippled by the cracking of their pump chambers, five of these having cracked in No. 2, and six in No. 3. Contracts have been awarded for replacing six of these chambers, and provision should immediately be made for replacing the remainder. The fly-wheel of No. 11 HoUy-Gaskill engine fits its shaft imperfectly, and it is therefore difficult to hold it in place. This defect should be remedied, and the dia- phragms and valves of the pumps should be renewed where necessary. 23 Each of the new Holly pumps Nos. 2 and 3 is con- nected directly with the river by two 48-inch pipes. All the other pumps take their water from a forebay which is supplied by one 10-foot conduit and two 48-inch pipes. This forebay is partially filled with sand, which inter- feres with the operation of the pumps, and has to be re- moved at considerable expense, besides forming a bed for the growth of long grass, which, during the summer, requires that four men be kept constantly at work keep- ing the forebay clear. New conduits should at once be built through this forebay, and the latter then filled up, as recommended by the Bureau of Water. Belmont Pumping Station. — The diaphragms and steam valves of Worthington Engines Nos. 1 , 2 and 3 are in poor condition and should be repaired. No. 4 Worthington High-duty Engin,e, since its re- moval from Belmont to Spring Grarden in 1894-9^, has remained unprotected except by a rude frame shed erected by employees of the Bureau. Designs for a new engine house have been prepared. A suitable house for this engine should immediately be constructed. Queen Lane Pumping Station. — Owing, probably, to the admission of air through the joints of the long suction main, the pump chambers in the four pumps at this station have been cracked. Some of them have been replaced by new pump chambers. The remaining ones should immediately be replaced, and the suction mains should at once be lowered so as to discharge by gravity into a well placed as nearly as possible directly under the pumps. The coal shed and tunnel for the proper supply of coal to the station, designed by the Bureau of Water in 1895, should at once be constructed in order to save the 24 present extra cost of about 23 cents per ton for handling the coal. Roxborough Pumping Station. — No. 1 Southwark en- gine has long been in a precarious condition, and within the last few weeks it has finally become so fractured as to place one-half of it out of service. Inasmuch as proper repairs to this engine would be expensive, we recommend that it be abandoned when it reaches a condition forbidding its further use. This station has long been unable to cope with the de- mands upon it. With the recent breakdown of the large Southwark en- gine, the conditions have become even more critical, and a new 3,500,000-gallon Worthington pumping engine has been purchased. Contracts have been awarded for four new Worthing- ton pumping engines, each of 5,000,000 gallons daily capacity. When these are installed, the a,nnoying con- ditions in the district will be materially relieved, and it will then be possible to make needed repairs to the present Worthington pumps. Frankford Pumping Station. — Owing to the insufficiency of the distribution system by which the water is delivered from the Wentz Farm reservoir, this station has, at present, a surplus of pumpage capacity and the engines are generally in good condition. High Service. — At Belmont and at Roxborough high- service pumping stations, the operations have long been carried on under precarious conditions, owing to the fact that each has had only one pumping engine of capacity sufficient for the requirements of the district, but con- tracts have recently been awarded for a new 5,000,000- gallon pumping engine at each station. Mount Airy Pumping Station. — There are no special re- quirements at this station. 25 Chestnut Hill High-service Station. — The two small pumps at this station, when in operation, take water from an adjacent well, fed from a reservoir, which is supplied partly by springs and rain water, and partly by overflow from the pumpage of Mount Airy and Roxborough high- service stations. Proposed Frankford High-service Pumping Station. — Contracts have recently been awarded for the construc- tion of a high-service pumping station adjoining the Wentz farm, or Frankford reservoir, for the purpose of supplying the village of Fox Chase, distant about two and one-half miles, and the intervening district. A 3,000,000 gallon pumping engine has been ordered for this station from the Holly Manufacturing Company, of Lockport, N. Y. Reserve Pumpage Capacity. — If the present system of pumpage and distribution is to be- maintained, it is highly advisable to furnish additional'^machinery at the several pumping stations, except the Frankford station, in order, not only that there may be sufficient capacity for the demand, but also a surplus in case of break-down or needed repairs. Cost of Operation. In Appendix IV is given a table, in itemized form, showing the earnings and expenditures of the Bureau of Water for each year of the ten years from 1889 to 1898, inclusive. In column "e," of this table, are given the expendi- tures upon pumping stations and reservoirs, including salaries and wages of all employees at pumping stations and reservoirs, fuel, lighting, repairs to boilers and ma- chinery, ordinary repairs to reservoirs, and all other ex- penses, for operation and maintenance and incidental to buildings, grounds and reservoirs. 26 It is witli the figures given in this column that our estimates of the annual expense in each of our several projects should- be compared. The other columns repre- sent also expenditures for general maintenance, operation and extensions. If the entire expenditure, necessary for the City's water supply, is desired, the figures in these columns should be added to our figures, which represent only the cost of delivering water into the reservoirs. In making such comparison, however, it must be borne in mind that in our estimates we have included interest on the cost of construction and an allowance for depreciation, items which are not included in the Bureau's figures in column "e." Deducting these, we find our estimates lower than the expenses reported by the Bureau. This difference may be accounted for by two facts, viz. : (1). In our estimate we have not included maintenance of stations and grO|Unds, and certain other general items include^ in the Bu^reau's figures. (2). We have estimated upon a somewhat less con- sumption than the quantity now actually pumped, believ- ing that with reasonable provisions for preventing waste the consumption can presently be reduced below our estimates. The cost of maintaining the high-service stations and the annual expenses of the distribution system, of the construction and repair shop, and of administration, are common to all projects. Hence we have not included them in our estimates, these being intended, primarily, to afford a means of comparison between different projects. For the same reason, and although we have included the cost of laying such mains as will be necessary for bringing the water from the new sources, or from the filter stations, into the distributing reservoirs, we have not included the annual cost of extensions of the distri- bution system. 27 Water Consumption. Quantity Furnished in Philadelphia. — The latest annual report of the Bureau of Water, that for 1898, gives 102,241, 835,372 gallons as the total amount of water pumped during the year, including the double pumpage for high service stations. It states that the " average daily pump- ing was 272,670,777 gallons " and, estimating the popu- lation of the City at 1,400,000, the consumption was 196.2 gallons per capita per day. The Bureau freely admits that, in spite of allowances on account of slip and other defects in the pumps, these figures, based on the plunger displacements of the pumps, exaggerate the actual quantities of water pumped. Re- cent observations, made by the Bureau of Water with the Venturi meter, at certain of the pumping stations, indi- date that the actual average daily pumpage in 1898 was not more than 220 million gallons, dr 169 gallons per capita per day. We recommend that Venturi meters be placed on all the pumping mains, in order to measure the actual quantity of water pumped at each station. The records of the Bureau of Water, based upon plunger displacement, with allowance for slip, etc., show a con- sumption of water per capita per day of only 36 gallons in 1860, 54 gallons in 1870, and 68 gallons in 1880, during a period when water was lavishly used by householders in daily washings of pavements, etc., now much less gen- eral. A gradually increased and more universal intro- duction of sanitary appliances, closets, lavatories, baths, elevators, etc., accounts for some of the recent increase in the consumption of water. When it is found from the records that the per capita consumption in 1890 had mounted to 132 gallons, and in 1897 to 212 gallons, it is evident that the figures do not represent the actual use 28 per capita per day, but that there is added an unneces- sarj"^ waste. Quantity Used Elseivhere. — On Plate XVI. a diagram shows the quantity of water consumed in various cities between the years 1860 and 1898, per capita per day, as compared with that consumed in Philadelphia. The figures in the vertical line to the left give the per capita consumption for the years indicated on the horizontal line at the foot of the diagram. The consumption of water per capita, as deduced from the total population and the total consumption, does not always furnish a true statement of the average amount actually used by each individual, particularly when cities are newly supplied with water, as many of the residents may, in the earlier days of the water plant, be supplied from other sources, and in such cases, a computation based upon the total population gives, of course, too low a consumption per capita. As the number of consumers increases, the error diminishes. Philadelphia is represented on the diagram by a heavy line. Buffalo, N. Y., and Washington, D. C, are the only cities exceeding Philadelphia in the per capita rate, and it is noticed that the Buffalo consumption has been decreasing quite rapidly since 1895, due to some reduc- tion of waste. Some cities show remarkably uniform re- sults, for instance, Montreal, Quebec, and Glasgow, in Scotland. Many of these cities have introduced meters, the effect of which will be discussed further on. Quantity Required in Philadelphia. — Careful estimates of the amount of water required per person, on the most liberal basis, give results considerably below the probable present consumption. The investigations of the Bureau show that there certainly is a large' amount of unnecess- ary and preventable waste throughout the city. After giving this subject serious study we have decided 29 that about 150 gallons per capita per day, or a total of 200,000,000 gallons per day, would be an ample and even liberal allowance for the actual requirements of the present population. Our estimates of cost are, therefore, based on this amount of water being required at present. In the year 1950, when the poi)ulation will probably have increased to three million inhabitants, the supply presumably needed will be 450,000,000 gallons per day. Meters. — No restriction should be placed upon the use of water required for health, comfort and cleanliness; nor should a part of the community be encouraged to deprive another part of its full quota of water. We are therefore emphatically of the opinion, and strongly urge, that all practicable means should be adopted to secure a fair and equitable distribution of the City's water. We know of no better means to this desirable end than the introduction of water meters, not only for all business properties and manufacturing establishments, but also for such private consumers as are found, by the Department of Public Works, to be carelessly wasting water from the public supply. This remedy is available and simple, and it has been already adopted in many cities with entire satisfaction. Plate XVII shows graphically the decrease in con- sumption per tap in a number of cities where meters have been introduced. In 1880, the City of Milwaukee, Wiscon- sin, had only 26 meters and the daily water consumption per tap was 1,781 gallons. In 1898 it had 22,036 meters, with a daily consumption per tap reduced to 644 gallons. About 70 per cent, of all private buildings, all railway sta- tions, all business properties and manufacturing establish- ments were metered. There remain only 30 per cent, of consumers whose supplies are not metered, and yet the amount generally taken through this 30 per cent, for domestic purposes, equals in amount the whole quantity 30 of water delivered by the meters to the remaining 70 per cent. Notwithstanding the extravagant waste through the unmetered connections, the total consumption per capita per day was reduced from 220 to 80 gallons. Many other cities show similar results. Our plate shows several of them. The City of New York, which certainly requires as much water as Philadelphia, consumed in 1898, for the boroughs of Manhattan and Bronx, 121 gallons per capita per day, for a population of about two million inhabit- ants. It had in use over 35,000 meters in these two bor- oughs, covering, it is said, every place where water was used extensively for other than domestic purposes. The report of the Commissioners for the City of Pitts- burgh advises very strongly the introduction of meters in connection with their new supply. The consumptibu per tap for Detroit, as given on Plate XVI, commenced to -decrease after metel-s came into use, and in 1898 it was only 730 gallons per day as against nearly 950 gallons in 1888. In 1888 the Sun- day waste is said to have been from 50 to 60 gallons per capita and the average waste for the entire j'^ear was thought to have been from 30 to 45 gallons per capita per day. A system of inspection was established and this appears to have been very effective. A careful record was kept of the condition of the plant, but the per capita waste was still about 6 gallons daily. It is stated that at present practically all manufacturing, business, muni- cipal, public and semi-public consumers and about 4,000 private families have meters. We earnestly recommend the introduction of meters for the City of Philadelphia with perfect confidence that the private consumer is given full and ample use and enjoy- ment of all water for his needs and comforts, at no greater cost, and probably, in many cases, even less cost 31 than the present rates impose. The meter is not proposed to increase the revenue, but to prevent one citizen from depriving another one of his rightful share of water. A private corporation would introduce meters at an early day if not restricted by law, and would at the same time encourage consumption in every way. The lack of a sufficient supply of water, in various parts of the city, is due either to a deficiency of distribut- ing pipes, to the lack of pressure from the reservoir, to the want of pumping machinery, to a waste of water, which reduces the head, or to two or more of these causes com- bined. The remedies are apparent. Deficiency of Local Supply. There is hardly a district in the city in which some portion is not suffering more or less from want of water. The trouble is greatest during, the summer months when more water is needed than at other times. In some locations, as for instance in portions of the Belmont dis- trict, it is difficult for days in succession to obtain water even on the second floor of the residences. The natural supposition of the inhabitants is, that the trouble is caused by a lack of reservoir capacity, but this is not the case. It is due to the several causes mentioned above. Four-inch pipes have been found to be nearly filled by incrustation, and six-inch pipes, laid in 1834, have been found incrusted to the extent of one-half inch in thickness. The only means of remedying this trouble is to lay larger distributing mains. In several cities corroded mains are reported to have been successfully cleaned with scrapers. The operation has cost about two-thirds that of laying a new pipe, and it obliges the shutting off of the supply during the time of cleaning. Cleaning the pipe leaves the interior sur- face in a condition to more readily corrode than before. In some cases such cleaning may be advisable, but it will more generally be preferred to allow the incrustation to continue as long as possible and then to lay a new pipe. If pipes are laid of sufficiently large size, in the beginning and thoroughly protected with a proper coating, they will give little or no trouble in this respect. The shortness of the time at our disposal forbade any attempt to investigate in detail the condition of the city's elaborate system of distribution ; and we are therefore, of course, unable to make specific recommendations as to which portions of it should be immediately relaid, or where additional mains should be placed, excepting in such cases where there appeared to us no doubt whatever. Condition of Reservoies. Of the three large reservoirs. East Park, Queen Lane and new Roxborough, the last two have, during the past four years, been re-lined with asphalt. East Park has been in full service ever since its completion, ten years ago, and Queen Lane ever since its re-lining. Owing to inability of the pumps to keep pace with the demand, the new Roxborough reservoir has never been filled ; in fact, it has recently been necessary to shut it off from the distribu- tion, to prevent its being entirely emptied. None of the three reservoirs gives evidence of material leakage. Opportunity has been lacking for systematic observa- tion of the behavior of the smaller reservoirs. Except Wentz farm, which is not in good condition, they are in fair order and repair, and holding their intended quantities of water. A hasty observation of the behavior of Wentz Farm reservoir indicated a very moderate amount of leakage ; but it was impossible to ascertain 33 positively that some of the pumpage did not enter the reservoir through partly-open stops. As the reservoir has but one basin, it cannot be thoroughly repaired with- out being put out of service. The coping of the retaining wall at Queen Lane reser- voir should be finished, and a new watch-house should be built. The northwest corner of the new Roxborough reservoir has an undesirably steep slope. This should be remedied by carrying the slope across Port Royal avenue. The Lehigh (or Fairhill) reservoir is too low to be advantageously used at present. With its height in- creased, it would make a useful clear-water reservoir. At Wentz Farm, a clear-water reservoir should be con- structed on the lot (belonging to the city) adjoining the present one on the east, and between it and the Second street road. The existing structure should then be con- verted into a similar reservoir. The Corinthian reservoir, likewise, may eventually be converted into a clear-water reservoir, when the demand for its capacity occurs. Uses of Reseevoies. Reservoirs for municipal water supplies are required for the purposes of storage, sedimentation, and distribution. Storage. It is well known that the flow of a river or smaller water course varies with the rainfall, that it is greatest shortly after rain storms,' and least at the end of a long season of drought. In order to best utilize the flow of a streaiii, it is, therefore, necessary to provide storage reser- voirs in which the flood water is collected and stored, and from which it can be drawn out to augment the flow of the stream when this is low. 3 34 Storage reservoirs are also used for the purpose of re- taining a supply of comparatively clear river water to be furnished to the city when the natural flow of the streams is very turbid, as, for instance, during the first wash after a storm. Sedimentation. When using surface waters, particularly from streams running through territory that is readily eroded, and, therefore, charged with much suspended solid matter, it is desirable to allow the water to come to comparative rest for a short time in order that the suspended matter may settle, and, thereby clarify the water. Settling reservoirs are used for this purpose in either of two ways : By one the water is merely checked in its velocity, but flows through continuously ; and by the other the water comes to an absolute rest, and is discharged intermittently. The time necessary for the water to remain in the settling reservoirs, whether by the continuous or intermittent sys- tem, depends largely upon the nature and quantity of its suspended matter. It is found that in most cases the dep- osition which takes place during twenty-four hours is practically sufficient, inasmuch as by far the greatest deposit usually results within this period. Distribution. Owing to the fact that the consumption of water in a city varies from hour to hour, and that it is greater dur- ing day-time than during night-time, and greater during some hours of the daj'^ than during others, it is nece«sary to provide a reserve within the city sufficient to balance the irregular draughts that may occur. This irregularity of draught is caused mainly by the domestic use of- the water ; but a conflagration or the bursting of a water main 35 would also affect it. The larger the city, the less notice- able, comparatively speaking, is the eflfect due to the draught for a large fire or to the breaking of a pipe, be- 'cause the regular domestic consumption is then large in proportion to the other draughts. For this reason, dis- tributing reservoirs in large cities may provide for a smaller proportionate daily reserve than those in small cities. In the former, one-quarter to one-half day's sup- ply is ample. Distributing reservoirs in different parts of the city are used also for the purpose. of maintaining approximately even pressures in the pipe system. Reservoirs in Philadelphia. — In this city reservoirs have been used for all three of the above purposes. It was de- sirable to provide a sufl&cient amount of storage, to enable the reservoirs to supply the citizens while the Schuylkill river was running very muddy, and thus somewhat lessen the turbidity of the water as finally turned into the pipes. Incidentally, the reservoirs were also efficacious for the purpose of allowing sedimentation to proceed during the comparatively quiescent state of the water while stored. Thirdly, they were used for the purpose of balancing the irregular draught during the different hours of the day. At the present day, the people are no longer satisfied by a mere lessening of the turbidity of a city's water supply through sedimentation. Perfectly clear, and practically pure water is now demanded. Storage reservoirs are, therefore, no longer necessary for ordinary river or lake water, unless they are used for sedimentation to be fol- lowed by filtration. Storage reservoirs,, for the purpose of compensating the yearly flow of the streams, would be required for this city 36 only ill the event of the supply being taken from the com- paratively small streams in the mountainous districts. If the Delaware and Schuylkill rivers are to be used, at points near the city of Philadelphia, the former at least,' owing to its large size, constitutes its own storage reservoir, and, therefore, no special structures are needed here for the specific purpose of equalizing the seasonal flow. Sedimentation or settling reservoirs have been wanted in this city only because the waters of the Delaware and Schuylkill rivers are now used in their raw state, and, par- ticularly the latter, are very muddy after rain storms. If clear mountain water is used, they are not required. If a filtered supply is obtained from these two rivers, sedimen- tation reservoirs are required to give the water a prelimi- nary clearing. Distributing reservoirs within the city are wanted only for the purpose of providing for the. variation in the daily draught, and they may, therefore, be comparatively small. Clear-water reservoirs, a term used when filtered water is supplied, are identical with distributing reservoirs so far as their purpose and their sizes are concerned. Inas- much as filtered water, like spring water, is apt to permit of a rank growth of algse and similar plants on exposure to sunlight and air, these reservoirs are generally covered to avoid a deterioration of the water. Such a covering is usually of masonry, rather than of wood or iron. The for- mer has the advantage of keeping the water cooler, and also for this reason is less likely to induce the growth of minute vegetal organisms. In the case of a filtered supply, the Schuylkill and Delaware rivers form the storage reservoirs, like those' afforded by the lakes at Chicago and Detroit. There are no artificially built reservoirs within those cities. The only use for clear-water reservoirs in Philadelphia is 37 for the purpose- of providing for the irregular daily con- sumption, for accidents to the mains, fires, etc. It is, however, necessary, when the sizes of the reservoirs are limited to half a day's supply, to have a sufficient re- serve of pumping capacity in case of accidents to any of the machinery. A surplus of mains and machinery for pumping and for distribution affords as effective protection as does a large surplus capacity of clear-water reservoirs, and is generally less costly and more serviceable. Quality of Water. Standard of Purity. — A water for the City's supply should have a bacterial purity fully up to the best recog- nized standard. It should be clear, palatable, and free from chemical and organic pollutions. These qualities can be obtained by using either natural waters which are free from organic and mineral pollutions, or by artifici- ally purifying waters already polluted. With regard to natural waters, one of the first steps to- wards maintaining their purity is to reduce their pollu- tion by every possible means, and principally by enacting and enforcing stringent laws against pollution. Legislation. — In the State of Pennsylvania the statutes governing the question of preserving the purity of streams are somewhat deficient. Attention may be called to the good results achieved by the statutes enacted in Massachusetts.* The State Board of Health should have supervision of all sources of domestic water and ice supply, with author- ity to inspect, to make examinations and analyses, and to enact and enforce regulations for the purpose of prevent- ing pollution and securing proper sanitary protection of all sources of water supply for cities. It should have * Manual for the use of Boards of Health of Massachusetts, and contaiDlng statutes relating to the public health, Boston, 1889. 38 authority to appoint agents to enforce the provisions of statutes and those of its own regulations. The Board should have authoritj', and it should be its duty, to cause investigations to be made and to prohibit the pollution of any water course by any city, town, vil- lage or habitation, and to require such modifications in any operation, or plant, as it may deem necessary, for the abatement of the nuisance. While legislation can do much towards lessening the pollution and thus improving the quality of natural water courses, it cannot wholly eradicate the dangers due to contamination. There will still remain surface drain- age from cultivated lands, treated by the ordinary pro- cesses of agriculture, the use of legitimate fertilizers, man- ures, etc., and undetected defilements of a minor character, which cannot be controlled. It is well-known that, with certain diseases, such as cholera and typhoid fever, a A'ery slight pollution by excreta from the patient, may produce widespread and most serious results in any community using the water directly as it comes from the streams. The epidemic at Plymouth, Pennsylvania, after the Cen- tennial Exposition, is one of many cases which have de- monstrated this fact. During the late war strong evidence has come from the Surgeon General's office in support of an already well- known fact that winged insects may carry disease germs from one place to another and thus may infect surface water supplies. FUTURE SUPPLY. Earlier Studies. — The subject of obtaining a pure and ample supply of water for the City of Philadelphia is one that has been before the public for nearly half a century. As early as 1856 suggestions were made for securing better water, and, in 1858, Mr. H. M. P. Birkinbine, Chief Engineer of the Water Department, drew attention to the 39 Wissahickou Creek, the Delaware and Lehigh rivers at Easton, and the Schuylkill river above Reading. In 1864 a reconnaissance was made of all the streams around the city, within a radius of forty miles, and a gravity supply recommended from the Perkiomen creek. In 1867 a special Committee of the Park Commission made a report, concluding that the Schuylkill river could be relied upon for many years if proper means were taken to guard it from pollution. In 1875 a Commission of Engineers was appointed to consider the entire subject of present and future water supply, with special reference to immediate needs. Owing to the lack of information at their disposal, this Commis- sion made no recommendation as to the question of future supply. Agitation of the subject was continued by the Water Department and by private citizens, until, in 1882, the imminent prospect of a water famine resulted in the appointment of another Board of Experts. They reported not only that' a marked deficiency existed in the capacity of tlie plant for the required supply, but also that com- plete- and accurate surveys should be made of all avail- able sources from which a future supply might be obtained, and that a thorough investigation should be made of the increasing pollution of the water of the Schuylkill river, with the possibility of its control by engineering works and legislative enactments, so as to restore it in some measure to its " pristine condition of comparative puritj^ and wholesomeness." The result was the appointment of a coips of engineers under Col. (now General) William Ludlow, Chief Engin- eer of the Water Department, by whom topographic, sani- tary and hydrographic surveys were made, important data collected, and maps and approximate estimates of cost prepared for a supply from all available sources. All 40 of the information thus secured is contained in the annual Reports of the Water Department of this city. We have availed ourselves of the data collected by the previous Commissions, and of the surveys, reports, etc., on record in the Water Bureau, particularly of the data contained in the reports made to Col. Wm. Ludlow by Mr. Rudolph Hering in 1883-6, and we have made ex- amination of the various areas of country comprised within the watershed lines of the tributaries of the Schuylkill and Delaware rivers in Pennsylvania, includ- ing the Lehigh river. We have considered all available sources of supply and have investigated the question of the pollution and purification of streams. We have also examined the present plant of the Water Bureau, in- cluding the reservoirs, the pumping stations and the distribution service. The possession of still other data v/ould have been desir- able, regarding conditions such as the turbidity and pollu- tion of the water of both rivers at different seasons, the tidal distribution of sewage in the Delaware river, particu- larly at spring tides and when the upland flow is a minimum. The allotted time was too short, however, to obtain such information. Nor would it have brought out any new facts to change our main conclusions, be- cause it affects only details, which can be considered later. Character of Present Supply. The water supplied to some parts of the city is scarce in quantity, and in all parts inferior in quality. As already stated, the pressures throughout the city are generally lower than they should be, and many important districts are almost entirely without a supply, notwith- standing that several of the pumping stations are driven to their utmost to keep pace with the demand. The quality of the water is also very far below what is 41 now considered a proper standard, the water being not only exceedingly turbid after storms, but also subject to serious sewage pollution. The typhoid rate of the city is unusually high, and this condition is no doubt charge- able, in great degree, to the sewage pollution at all times present in the water. The deficiency of the supply, as to quantity, can be remedied by diminishing the waste, or by increasing the puii^page. In many cases the supply to buildings is restricted by the lack of capacity of the distributing mains, but there are instances where, if the condition of these were im- proved, the draught would exceed the capacity of the pumping engines. We have elsewhere recommended the adoption of efficient measures for the reduction of the waste. The defects in the matter of quality can be reme- died either by abandoning the present sources of supply and adopting purer ones, or by applying to the water taken from the present sources well-known methods of purification. Projects Specially Investigated. The projects which we have investigated in detail may be grouped under two principal heads : Mountain waters, supplied in their natural state ; and filtered waters, sup- plied from the Delaware and Schuylkill rivers. The impounding or storage reservoirs may discharge directly into aqueducts or into the beds of the streams below them. In either case the water is received in other storage reservoirs further down stream, from which aque- ducts convey it to the city. Filtered vs. Mountain Water. Wliere ample supplies of relatively pure water are ob- tainable at sufficient elevations and within short distances 42 of the community to be supplied, it will usually be found best to take advantage of them ; but where, as in our case, these sources are found at long distances from the city, it is necessary to estimate very carefully, and to bal- ance still more carefully, the relative costs and advantages of different methods. A gravity supply obviates the heavy operating ex- penses incident to a supply by pumpage, and thus natu- rally commends itself at first sight, but it may readily happen that the interest on the cost of construction of the gravity supply considerably overbalances the saving due to this consideration. To utilize a. gravity source of supply in our case re- quires not only the construction of long and expensive aqueducts, but also that of large and numerous impound- ing dams on the various small streams which would be taken as sources. These dams are necessary in Order that the heavy winter and spring flows may be saved and made to compensate for the droughts of summer ; thus regulating and rendering more nearly uniform the avail- able yield of the stream throughout the year. An advantage of the pumpage over the gravity .system consists in this, that the former is capable of indefinite extension In' small additions, whereas, when the capacity of an aqueduct has been fully taxed, a second one, usually of at least equal capacity, must be built. In comparing the relative advantages and disadvan- tages of a mountain and a filtered water supply, it must be borne in mind that a filtered water supply is ordinarily susceptible of gradual and indefinite extension, as the demands upon it increase; whereas the construction of a gravity system, for a growing community, requires an outlay much in advance of requirements. It is true also that, owing to the greater length of time required for the construction of a gravity system, large sums of money 43 must be invested long before the system cau be put into operation. The adoption of any project for bringing mountain water to the city by gravity, at sufficient elevation to flow into our reservoirs, involves, of course, the abandonment of the pumping stations supplying these reservoirs. In our own case, another consideration to be borne in mind is that the sums represented by the present value of the pumping plants would be lost in the case of the construction of a gravity supply. In comparing the relative advantages of filtered and mountain water supplies, it is important to bear in mind the lengths of time which would probably be required for their installation. It is quite safe to say that the comple- tion of all plants of any one of the slow filter systems herein suggested, could be accomplished within three years ; whereas the construction of any of the mountain water systems would probably require not less than seven or eight years. In bringing mountain water to the city, there would always be a question as to its absolute purity, because there is no guarantee against an accidental pollution. Nor is there a guarantee that the water, coming from territory sometimes densely covered with forests, would not, in the late summer, have a slight vegetal taste such as we find in most supplies from similar sources. In view of recent progress in the methods of water puri- fication, and of the growing demand for better water, it seems not at all improbable that water procured from the Blue Mountains might in the future require filtration before being delivered to the city, thus adding mate- rially to the expense of the project. The New York supply, although not coming from the mountains, is de- rived from a territory which is carefully protected against pollution, but it is an almost annual occurrence that, in 44 the summer, the water has a vegetal taste. A former Health Officer of New York City, Dr. Jenkius, is on record as saying that the New York water would no doubt eventually have to be artificially filtered in order to remove this taste and a slight turbidity. Another advantage of a filtered water supply, lies in the fact that, in case it should ever, in the future, be found necessary to change the source of supply — as, for instance, to abandon the Schuylkill and take filtered water from the Delaware — the loss in money would be less than if a mountain source had been used and a purification of such water iiad been found necessary. In cases where the issue was doubtful, as to yield of water, or as to cost of construction and operation, we have, as a rule, given the benefit to the mountain water supply. Quantity of Water Available. As to the quantity of water obtainable from the sources at command under present conditions, it is self-evident that the minimum flow is all that can be relied upon throughout the year in any stream without reservoir storage, and the minimum flow usually occurs at a time of year when water is most needed. Minimum' Flow of Schuylkill. — On the Schuylkill river are a number of reservoirs which have been constructed by and which belong to the Schuylkill Navigation Com- pany. The company uses these reservoirs for the purpose of supplying water for the maintenance of its navigation and for power at certain points. The City cannot de- pend upon the use of this impounded water to supply its need in time of drought, as the Navigation Company is under no legal obligation in this respect, although it has on several occasions acceded to the City's requests for assistance. 45 Various opinions have been given as to the extreme minimum flow of water in the Schuylkill river. After considering these we have decided that it would not be safe to rely upon taking from the river in times of drought, more than 150,000,000 gallons per day. This quantity may be less than the minimum flow, but even if the City had a plant for purifying the water, we do not consider it safe or proper to provide for using the entire flow, par- ticularly at a time when the relative pollution of the river is greatest. Yield of the Delaware River. — The Delaware river from its mouth to Trenton, N. J., is a tidal estuary. The mini- mum flow at the Water Gap, situated about 100 miles above Philadelphia, is estimated at about 700,000,000 gallons per deij, only one-half of which, or 350,000,000 gallons per day, could be appropriated by the City of Philadelphia. At this point it should be stated that while the City has the right to one-half of the flow, it. has no right to cause injury to any of the riparian owners, resulting from a material lowering of the water level during a drought, the exposure of shoals, and the recession of the low water line from where it is at present. We have in- cluded in our estimates of cost no allowance for such damage, nor for diminishing any water-power rights be- low the intake. Yields from Watersheds with Storage. — The upper Perki- omen creek comprises a drainage area of 120 square miles and, with storage reservoirs for which satisfactory sites exist, this source may yield a supply of 90,000,000 gallons per day. The available safe yield of the Tohickon, the Big and Little Neshaminy creeks, according to the recent reports of the Water Bureau, are : Tohickon, 61,000,000 gallons per day, and Neshaminy, 83,000,000 gallons per day. 46 The upper Lehigh river drains an area of 377 stjuare miles, while Big and Aquanchicola creeks, to the east of the river, south of Mauch Chunk, drain an area of 165 square miles. The watersheds of all these streams com- bined include 542 square miles, which, with storage, can be depended upon for a supply of 360,000,000 gallons per day. The .sources of the Delaware river above the Water Gap, from which good' mountain water can be obtained, cover a drainage area of 430 square miles, and, with storage, will furnish 260,000,000 gallons of water daily. Quantities available. — The quantities available, there- fore, from the several sources mentioned are : — Mountain Water, Unfiltered. Lehigh river, including Big and Aquanchicola creeks 360,000,000 gallons per day. Upper Perkiomen 90,000,000 " " Sources near Delaware Water Gap.. 260,000,000 " " Total 710,000,000 " " Water Requiring Filtering. Schuylkill river at Philadelphia 1-50,000,000 gallons per day. Delaware river at Philadelphia Practically unlimited. Delaware river at Water Gap 350,000,000 gallons per day. Tohickon and Neshaminy creeks, with storage 144,000,000 " " Perkiomen creek above Schweuks- ville, with storage 160,000,000 " " It is of course practica,ble to increase the minimum flow of the Schuylkill river by the storage of water on its afflu- ents, particularly on the Perkiomen creek. As the pol- lution to which the water of the Schuylkill river is sub- jected will be greater, in spite of all legal restrictions, than the pollution of the Delaware river, which has a smaller resident population, we have deemed it rather 47 better to increase the city's supply by taking water from the latter stream, than by artificially increasing the mini- mum flow of the former. We have been guided in this decision also by the question of cost and by the existing rights of the Schuylkill Navigation Company. Artificial Purification.. When water, to be used for a domestic supply, has become contaminated, it should be artificially purified by filtration, preceded by sedimentation where neces- sary. This method of purification has been in success- ful use in Europe for many years, and its use is growing in this country. The investigations and experiments of the Massachu- setts State Board of Health, which have extended over a number of years, have placed the subject of water puri- purification upon a scientific basis, and it is possible now to efi'ect any desired degree of purification with a certainty of results which, previous to such investigations, was im- possible. * Water Filtration. As we have already stated, it is rarely, if ever, that water obtainable in large quantities from natural sources can be used for domestic purposes in its natural condition with absolute safety. The very existence of such quanti- ties of water generally involves the co-existence of a popu- lation more or less dense, with the corresponding certainty of more or less serious pollution ; and, even where a supply is normally of a high degree of purity, as in the oft- quoted case of Plymouth, Pa., we are never free from the menace of accidental and temporary pollution, such as decimated that unfortunate town. Hence, it is becoming more and more the conviction of water-supply engineers that proper regard for the health of the community de- 48 maads the artificial purification of all surface waters, how- ever promising the sources from which they are drawn. Nature's process of filtration in the production of spring and well waters has long been understood in a general way, and its artificial imitation, on a small scale, is prob- ably as old as history itself. "Within the last half century the same process has been extensively applied to the pu- rification of the large volumes of water supplied to cities. Loudon furnishes the most noteworthy example of this, and the system there adopted is still the one most gen- erally employed. In that system, the water is, if necessary, first allowed to remain at rest in sedimentation reservoirs, in order that it may free itself of its grosser mechanically suspended im- purities, and is then allowed to filter slowly through beds of sand. Until within a very few years, the sole function of this process seems to have been regarded as consisting in the removal of the mechanically suspended impurities and the consequent improvement in the appearance of the water ; the turbid water of the Thames, for instance, being con- verted into a bright and sparkling liquid, probably quite as attractive in appearance, and as palatable, as the finest spring water ; and it was freely conceded, even by the ad- vocates of the process, that, in the language of an author- ity, the micro-organisms contained in the water " could pass a hundred or a thousand abreast through the inter- stitial spaces of ordinary sand as used for this purpose," and hence that " while filtration certainly clarified, it could not purify " — while it removed the visible dirt, " it could not remove the bacteria." During recent years, however, the investigations of biologists and the sanitary results of filtration have clearly demonstrated its very important usefulness in the true purification and sanitation of the water — efficient filtra- 49 tion commonly removing 98, 99 and even more per cent, of the bacteria existing in the water. The consequence of modern discoveries is a complete change in the accepted standard of purity of water. Whereas, previously, clearness and a satisfactory chemical analysis were considered sufficient evidence of wholesome- ness, we now insist, also, that a certain permissible maxi- mum number of bacteria — usually placed at 100 per cubic centimeter — shall not be exceeded. Hence, while the science of filtration may be said to be still in its infancy, it cannot truly be said that " filtration is only an experiment." One of the most striking instances of the efficiency of filtration in checking the spread of disease is the well- known case of Hamburg and Altona, in Germany. These cities, side by side on the banks of the Elbe, both take their water supplies from that stream, the Altona intake being placed below the point where the sewers of Ham- burg discharge into the Elbe the sewage of nearly 800,000 persons. The two cities are practically one, the line of demarcation being, at most, a narrow street. In the winter of 1892-3, when the cholera visited Hamburg and when the deaths from that disease, in Hamburg, which used the Elbe water unfiltered, reached 1,250 per 100,000 of population, Altona, which used the same water filtered, had but 221 per 100,000. The boundary line between the two cities can be clearly traced, upon a map on which are plotted the cases of this disease, by the large number of such cases on the Hamburg side of the line, and their nearly complete absence on the Altona side. The few cases which appeard in Altona were generally traceable to the use of the Hamburg water, by transient Adsitors to the other city. Filtration is found to remove, not only disease germs, but also the unpleasant vegetal taste which often charac- 4 50 terizes the water of small streams during summer and autumn. it has also been well established that, in the system already mentioned, the sanitary work is done, not always directly by the sand itself, but, in the case of continuous filters, rather by matter deposited from the water upon and within the sand, which thus serves merely as a me- chanical suppoi't for what may be termed the true filter, the deposit of " bacterial jelly," or, in German, the " SchmutzdecTce," which we may freely translate into " dirt-cover." It is not to be supposed that all, or even most, of the bacteria found in ordinary surface waters, are hurtful. On the contrary, many of them are beneficent; and it is, indeed, upon the operations of these, that the biological processes of purification upon and within the sand filter largely depend ; but it is practically impossible to dis- criminate between the beneficient and the harmful bac- teria, and hence the removal of the hurtful or pathogenic bacteria, brought into the water by sewage pollution, re(|uires that the depopulation of the water be made as complete as possible. It is stated, upon good authority, that more than twenty million people in Europe are now being supplied with water filtered by slow-filters, and the number of persons thus supplied is annually increasing. The first filter of record appears to have been constructed about seventy years ago. The slowness of operation of the system now being considered, requires an acre of ground space for every two or three million gallons filtered daily ; so that, for a daily supply of 200 million gallons, from 70 to 100 acres would be required for the filter beds alone, in addition to that which might be required for the sedimentation basins. To obviate the necessity for acquiring so much land. 51 American inventors have sought to take advantage of the method, known to the ancients, of using alum or some other coagulant, to hasten the formation of the true filtering medium, as well as to expedite, in other ways, the entire process. Besides, the " Schmutzdecke" sometimes requires sev- eral days for its formation, and, during this time, the water is but imperfectly filtered, and should be allowed to run to waste. Again, the filters naturally become clogged with sediment, and require cleansing usually every few weeks, and this cleansing is a slow process, throwing the filters out of use for a still further time. The result of these efforts is the so-called "me- chanical" filter, which consists usually of a tank, from ten to twenty feet in diameter, and containing a sand filter bed. Either at or prior to its admission to the filter, the water receives a small quantity of alum, or other coagulant, the proper behavior of which depends upon the presence, in the water, of some base, such as lime. This base unites with the sulphuric acid of the coagulant, thus setting free the alumina, which, in the form of aluminum hydrate, settles slowly through the water, carrying down with it much of the impurity of the water, while the new sulphate formed by this process is deposited with it. Another distinguishing feature of the " mechanical " process consists in the arrangement for cleansing the filter. This consists .(I) of a set of rakes, set in revolutoin by machinery when required (whence the term " me- chanical ") and (2) the reversal of the normal current of water, the water already filtered being forced backward through the bed, not only facilitating the revolution of the rake, but carrying with it most of the impurities deposited upon and within the bed by the water pre. viously filtered. 52 For convenience, we apply the term " slow " to the sys- tem first described, and represented most prominently by the great filter beds of London, Berlin and Hamburg, and sometimes called the "English" system; and the term " rapid " to the so-called " mechanical " or recently called "American" type of filter. Our reason for selecting the terms " slow " and " rapid " is that they designate the most important distinguishing feature. The former sys- tem allows from 6 to 12 cubic feet, the latter from 200 to 300 cubic feet of water daily to pass through one square foot of filter surface. By virtue of some operation not yet thoroughly ex- plained, rapid filters appear to be able to secure equally as satisfactory bacteriological results as the slow filter, although filtering at from thirty to fifty times the rate. In other words, for a quantity of water requiring from thirty to fifty acres of filter beds by the slow process, one acre of surface of rapid filters would suffice, provided the conditions of the given case were equally favorable to the two systems. It is now generally recognized that, as a rule, the slow system is best adapted to waters containing relatively little suspended matter, although they may be highly pol- luted by sewage, and the rapid system to the treatment of highly turbid but less seriously polluted waters, or in those cases where, as in certain manufacturing processes, clearness is the first consideration, and wholesomeness of less or no account. As a matter 'of course, the rapid filter commends itself especially for those cases where suitable ground for the- large slow filter bed is not practi- cally available. The functions of the rapid filter, however, are by no means confined to the mere clarification of the water. It is also a very efficient purifier. Hence, it has found ex- tensive and successful application, especially in this coun- try, for the purification of the water supplies of towns and cities. In these cases, usually a considerable number of the filtering tanks, or " units," are installed side by side, in connectio)! with suitable machinery for operating the revolving rakes, and with appliances for the admixture of the coagulant and the regulation of its quantity. Where lime, or its equivalent, is deficient in the water in its natural state, it must be added artificially, in order to insure the necessary decomposition of the coagulant. We have already given the results of investigations to determine the effects of the use of coagulants upon the wholesomenesS of water and upon their availability for use in boilers. The hardness of the Schuylkill water adapts it to the use of the rapid system, with its necessary employment of coagulants, the lime in the water acting favorably in the decomposition of the coagulant, and it is bur opinion that the use of the coagulant would not materially, if per-' ceptibly, increase the hardness of the water. With slow filters, a coagulant would be used only when the river runs very muddy, as happens only occasionally even with the Schuylkill water ; and we doubt whether it would ever be required with the much softer and less turbid water of the Delaware. If precisely the proper quantity of coagulant could be applied, it would all be decomposed, and all of the lime in the water would unite with all of the sulphuric acid of the coagulant. Hence, none of the coagulant could pass out with the filtered water. The only efi'ect, in this res- pect, would probably be the diminution of the " tempor- ary" hardness (that due to carbonate of lime) and an .' increase of the "permanent" hardness (that due to sul- phate of lime). The maximum quantity of sulphate of alumina used in the rapid filter rarely exceeds two grains per gallon, 54 and it is often much less. The Rhode Island Board of Health, for instance, has stated that 0.6 grain per gallon is sufficient. Sedimentation in reservoirs is accomplished in two ways. lu one of these the water is allowed to pass through the basin continuously ; in the other it is ad- mitted and drawn off intermittently. By the continuous method, the water enters at one- side of the basin, and its velocity very greatly decreases as the water flows to the other side, whence it is drawn off near the surface. The reduction of velocity permits the gross particles of sus- pended matter to subside. By the intermittent method the water enters the reservoir generally with a greater velocity than in the continuous method, but it is then shut off, comes practically to rest, and remains at rest for a sufficient time to allow the suspended matter to settle. The clear water is then drawn ofl". Both methods have their advantages and disadvantages ; and, in the lack of sufficient information regarding the quantity of suspended matter in the water furnished. to this city throughout the year, it is impossible to say which of the two methods would be the better one to use in this city. The estimates of cost presented are, however, in our opinion, sufficiently liberal to cover approximately either case. In the case of the Schuylkill river water, and when settling reservoirs are used for its preliminary treatment, it may be necssary to add alum to the water at times when suspended matter is in very large amount or when it is very fine. Our opportunities of observation do not enable us to state to what extent such a treatment would be neces- sary in the case of Philadelphia, but we think it would hardly be necessary on more than from ten to twenty days in the year. In the case of slow filtration, the use of a coagulant would be found advisable, only during, or just after, 55 heavy freshets. At such times the amount of alkaline sulphates naturally in the water are approaching a mini- mum, the organic matter being then most diluted. Any increase in these sulphates due to the decomposition of the natural carbonates by the use of sulphate of alu- mina, would in all probability not make the total per- centage of alkaline sulphates as high as during a drought, when it approaches a maximum. The use of a coagu- lant during freshets, therefore, could make no appreciable difference in the quality of the water. Miter Planis and Various Prajeets Examined. Sundry methods of filtration, purification, sterilization, etc., have been presented to us for examination, and a hearing has been given to those proposing or suggesting them. Wilminffton Filter. — A visit was made to tlie wiiter purification plant at Wilmington, Delaware. The system there adopted is based on a treatment of the water with iron, a subsequent thorough aeration and an upward fil- tration through a bed of 20 inches of gravel and 18 inches of sand. There are five filter beds in use, each 16 by 125 feet, filtering at a daily rate of over 40,000,000 gallons per acre (about 133 cubic feet per square foot) per day. The ordinary cleaning of the beds is accom- plished by reversing the current and washing the ma. terial in place with both air and water. The iron treatment is secured by means of a series of revolving bundles of small iron rods, suspended trans- versely in a long trough through which the water flows. A small portion of the iron is removed by attrition and oxidation, and acts like the ijon used in the Anderson process. No repeated analyses of the water had been made before and after this treatment, and no data were available to establish its real efficacy. This fact, and tlie 56 lack of success of the similar Anderson process, rendered a further investigation inadvisable. Albany Filter. — We next visited the new filter plant at Albany, N. Y.,» designed by Mr. Allen Hazen. At that time it was not quite completed, but it has since been put into operation. It is the largest filter plant now in use in the United States. It consists of an open sedimenta- tion basin holding 16,000,000 gallons, and 8 filter beds, each 0.7 acre, with a total area of 5.6 acres, and a total filtering capacity of about 15,000,000 gallons per daj% being at the rate of about 3,000,000 gallons per acre (about 9 cubic feet per square foot) per day. The filter beds are covered with groined concrete arches ; and all appurtenances necessary or advisable for effective operation, such as regulators to control the rate of filtra- tion, sand washing apparatus, a bacteriological labora- tory, etc., are provided. The filters have apparently been built with great care and excellence. The filtered water is pumped- into an uncovered distributing reservoir in the western part of the city. Poughkeepsie Filter. — We also' visited Poughkeepsie, where a filter plant has been in existence for more than 27 years and is still in active use, filtering about 1,600,000 gallons per day, or, with a population of 23,000, about 70 gallons per capita. This plant was designed and erected under the super- vision of the late James P. Kirkwood, who had investi- gated the filtration systems of Europe in the interest of the City of St. Louis, Mo. It is located on the east bank of the Hudson river and the original plant comprises a settling basin, 25 by 50 feet by 12 feet deep, a filtering basin. 150 by 200 feet Ijy 12 feet deep, and a filtered- water basin, 28 by 88 feet by 17 feet deep, with an inter- mediate chamber, 6 feet by 88 feet by 16 feet deep. The filtering materials are 24 inches of coarse broken stone, 57 24 inches of gravel of varying sizes and 24 inches of sand. The filtered water is pumped up to a large uncov- ered distributing reservoir on a hill back of the city. This reservoir has a capacity of 12,000,000 gallons or about 7 days' supply. In 1896 an additional filter was constructed, doubling the capacity of the plant. It is fed from the old settling basin, the water discharging into a delivery well and thence to the old filtered-water basin. All of the basins are uncovered, except that for filtered water, which also was originallj'' open. So much trouble was experienced from the growth of algte that in 1X91 it was covered, and it has since given no further trouble. The fact that the filters are uncovered has caused much diffi- culty in operating them. In summer the growth of algae at times almost stops their action, and in winter the frost causes difficulty in cleansing the beds and keeping them in proper working condition. On account of the expense an attempt was made, in 1874, to substitute simple subsidence in the distributing reservoir, but after a trial this was abandoned. The an- nual report of 1878 states that "the consuixiers accustomed to drink filtered water will accept nothing else, nor will they consider any circumstances or complication of cir- cumstances as offering any excuse for the non-use " of the filters. We were informed that the filters are operated only for three or four days in the week. The rate of filtration was about 4,500,000 gallons per acre (about 14 cubic feet per square foot) per day, until the construction of the addi- tional filter in 1896. With the latter in use, the rate is now about 2 J to 3 million gallons per acre (about 9 cubic feet per square foot) per actual day of operation. The filters are usually cleaned at intervals of from one to five weeks, but sometimes not for two months, depending upon 58 the condition of the raw water, the accumulation of algae, etc. With good management, the plant seems to have pro- duced satisfactor}' results, even under adverse conditions. Reading Sewage Filter. — During our investigation of the Schuylkill region, we made an examination of the sewage filtering plant at Reading, erected under the pat- ented system of Mr. John .Jerome Deery, President of the Pennsylvania Sanitation Company, Mr. Deery having presented plans for filtering the water of Philadelphia by the same method. The plant has been in operation about three years. It is designed to act first as a strainer, then as a filter, passing, as we understand, about 10,000,000 gallons per acre per day. Much reliance is placed upon aeration and sunlight to purify the water. Repeated examina- tions of the filtrate at Reading since the plant has been in operation, revealed faint sewage odor and color. The process as proposed for Philadelphia would not by itself yield a safe drinking water, if judged by, established principles and the results of experience. Maignen Filter. — Our attention was called to the Maig- nen method of-water purification, and a special examina- tion was made of a model plant of experimental filters in operation. The new and special feature of the Maignen system is the use of an asbestos film resting on top of the sand of the filter bed, to take the place and perform the useful office of the dirt cover (Schmutzdecke) on the bed of the ordinary filters, in retaining bacteria. The water passes through Mr. Maignen's experimental filters at the rate "of about 10 to 12 million gallons per acre (about 30 to 37 cubic feet per square foot) per day. It is clear, and, according to the analyses made by the chemist of the Company, also nearly free from bac- teria. It is stated that in all the examinations made of 59 the effluent water, the bacteria have never been found to exceed one hundred to the cubic centimeter, the stand- ard Hmit adopted by the German Imperial Board of Health. While the models show good results, and while the treatment of the problem lias been carefully consid- ered by the inventor, experiments have indicated the ex- istence of a troublesome feature, in connection with the asbestos film, due to the collection of air bubbles below the same and a consequent interference with.the required percolation of water. A remedy has been suggested by the inventor, but, until the system has been successfully used on a sufficiently large scale, we cannot recommend it for the City of Philadelphia. It could, however, at any time be readily added to any of the usual slow filtering plants, if its usefulness were established, and thus obviate their otherwise necessary areal extension pari passu with the growth of the cit3\ New York Filter MaiMfaduriag Company. — A confer- ence was held with the New York Filter Manufacturing Company, represented by Messrs. Samuel L. Morison, General Manager, and Edmund B. Weston, Consulting Engineer, to consider the method of filtration employed by that Company, and the cost of constructing and main- taining such a plant. This Company uses a system of rapid filtration, the rate being 100,000,000 gallons per acre (about 300 cubic feet per square foot) per day. They claim to have made an advance in their latest apparatus, by joining the com- partments for sedimentation and filtration in such a way that they can dispose of from 76 to 80 per cent, of impurities by sedimentation, before the water reaches the filter. The water is first treated with a coagulant (sul- phate of alumina) and then passed through a settling basin, where it is subjected to a rotary movement which 60 hastens the precipitation of the hydrate of alumina and other matter in suspension ; thereby shortening the necessary time for sedimentation, and obviating the use of much larger settling basins. A design for a unit filter was shown us, consisting of a cylmdrical tank, 26 feet outside diameter, and appurte- nances, giving an inside filtering area of 452 square feet. The water is first treated with the required amount of sulphate of alumina from a supply tank, by means of an automatic feed, known as the old Warren Chemical Feed, which secures the delivery of a quantity of alum in exact proportion to the quantity of water entering the filter, and which is provided, also, with means of regula- tion to meet the requirements of the varj'ing character of the water. Having received the proper amount of this coagu- lating solution, the water enters a lower basin in the bottom of the cylindrical filter tank in a tangential di- rection through a deflecting, elbow.' This gives it a rotary motion about the centre of the tank, which motion mate- rially facilitates coagulation. When filtering at the rate of 100,000,000 gallons per acre (about 300 cubic feet per square foot) per day, the water is detained in this lower basin for the period of about one hour, although in con- tinuous passage to the filter bed above. During this time a large amount of the small particles of matter in suspen- sion, with bacteria, etc., are gathered together in larger masses, many of them being of sufficient weight to fall to the bottom of the settling tank, about 75 per cent, of such matter being removed. The fresh supply tends to the outer edge of the tank by the centrifugal force due to rota- tion, and the water remaining longest under the action of the coagulant gradually reaches the centre. The water is then fed from the centre of the tank by an upright central pipe to the filter bed above. ■ 61 The water is distributed upon the filter, and deposits thereon matter remaining in suspension, together with the lighter particles of hydrate of alumina not already deposited in the settling tank below. The water passes through the accumulating film, and then through four feet of sand to the screen system. There are about 1800 separate screens, made of aluminum bronze, laid as a pavement on the floor of this upper tank to prevent the sand from escaping. Through these screens the water passes to a central manifold, and then to the controller. The controller, designed by Mr. Weston, is one of the most important improvements in this type of filter. A filter bed, when clean, operates more rapidly than after" it becomes clogged with deposit, the speed becoming less and less until a condition is reached when washing is required. The controller regulates, this speed automati- cally, determining the number of gallons per minute that shall pass through the filter, and thus making its action uniform. When a certain height of 'water is reached and the available friction head has been entirely consumed, ■ the sounding of an automatic signal shows that the time for washing has arrived. To wash the filter, clear water is pumped back through the pipes leading to the screens in such quantities as to give an upward velocity through the sand about three times as great as the velocity of filtration. It raises the entire bed in the form of quicksand, leaving the sand grains com- pletely in suspension. At the same time a mechanical agitator is started, consisting of a series of iron rods suspen- ded from arms revolving over the filter bed. The rods penetrate the bed nearly to the bottom. This facilitates . the loosening of foreign matter in the bed, and both this matter and the wash waier are carried off over the edge of the tank forming the filter proper, down the annular space between this and an outside tank, and thence to the sewer. 62 Information was given us as to the experience of this filter company in treating Schuylkill and Delaware river waters. Private filter plants of their installation were mentioned, of which three were on the Delaware river connected with sugar refineries and one of 5,000,000 gal- lons on the Schuylkill river. Sterilization by Boiling. — Our attention was called, by Mr. John Forbes, of the Waterhouse-Forbes Company, to an ingenious method of sterilizing water by boiling. B}' a simple contrivance, two vertical currents of water are kept flowing past each other iu contiguous passages, so that their heat is equalized by convection. The water, upon reaching the top of one column, is heated to the boiling point, and then, in descending, gradually imparts its heat to the water rising in the first column, and finally escapes with a temperature only 2 to 5 degrees higher than the original temperature of the water. The water is thus actually boiled with a very small expenditure of heat. In our judgment, this process, irrespective of finan- cial reasons, falls short of solving the question before us, inasmuch as it does not remove tlie turbidity of the water. Recent Reports on Filtration Experiments. Reports have recently been made by Filtration Com- missions of Louisville, Cincinnati and Pittsburgh, contain- ing the results of investigations and experiments which have been valuable to us. The character of the waters of the Schuylkill and tlie Delaware rivers is not necessar- ily the same as that of the waters of the Ohio and Alle- gheny rivers. . The latter vary considerably in quality, but there is much that is common' to all river waters. As it was impossible, within the limited time at our dis- posal, to undertake any investigations with experimental filters at Philadelphia, desirable as it would have been, we were obliged to depend largely upon experience gained 63 elsewhere, and we have taken all possible advantage of the above mentioned reports. The investigations at Louisville and Cincinnati were reported. by Mr. George W. Fuller, Consulting Expert, and those at Pittsburgh by Mr. Allen Hazen, Consulting Engineer, with Mr. Morris Knowles, Resident Engineer. Reference must be made to the reports for details. Louisville. — At Louisville seven different types of fil- ters were investigated : the -Jewell, the Warren, the West- ern gravity, the Western pressure, the Harris magneto, an electric system. Palmer and Brownell water puri- fier, and the MacDougall Polarite system. Daily tests were made from October 21, 1895 to August 1, 1896. The area of 85,000 square miles, comprised within the water-shed of the Ohio river, exhibits wide extremes in geological formation. The population above Louisville is 4,500,000, of which 1,575,000 are in 220 towns and cities, with an increase of 15 per cent, in six years. With groat and rapid changes in the character of the river water, depending upon the section of country from which freshets come, (and these freshets are occasionally very heavy,) the problem was not a simple one. At no time is the water entirely clear. The ratio between maximum and minimum weights of suspended matter was found to be as great as 5311 to 1, and the color varied from light grey to dark red. As a rule the water had a slight odor, which, however, was occasionally quite pronounced, in fall and spring musty, sometimes aromatic and resinous. In spring after rains it had a vegetable odor. The sedi- ment consisted at times of large amounts of fine particles of silt and clay, requiring weeks to settle. Individual particles were sometimes as small as 0.00001 inch in diameter. Mr. Fuller states, in his general conclusions, that " it is proved conclusively that the general method embodying 64 subsidence, coagulation and filtration is most suitable for the proper and economical purification of the Ohio river water at this city (Louisville). With regard to the use of coagulants, it may be stated in unqualified terms that their use is imperative for this water, because for at least six or ten weeks in the spring and early summer, the Ohio river water contains such large quantities of, fine clay particles, many of which are smaller than bac- teria, that clarification and purification without coagula- tion would be impracticable if not impossible." Further investigations were conducted from April to July, ] 897. In the final summary and conclusions there are some points that may here be noted. When the water is high and muddy it is not specifically injurious. When it is low and comparatively clear it is most to be feared. Filtration, preceded by subsidence, is the correct method, and the use of coagulants is imperative. An effluent water, free from turbidity, could be secured with an English (slow) filter plant at a net rate of about 1.5 million gallons per acre (4J cubic feet per square foot) daily, but there was a marked indication that fine clay was passing into the sand-layers, necessitating a cleansing at frequent intervals. A preliminary subsidence was neces- sary, and sulphate of alumina was found to be the most suitable coagulant. There were times when coagulation, in conjunction with subsidence, could be employed to advan- tage. The experiments with American (rapid) filters in- dicated that by taking advantage of a preliminary subsi- dence, the amount of sulphate of alumina could be held at from 1.5 to 2 grains per gallon, with an annual average of 1.75 grains. With regard to the use of water in steam boilers, from filters, using a coagulant, more incrusting constituents were found than in the raw river water, although their annual average amount contained in the filtered water was only 65 about 60 per cent, of the quantity normally present in the river water during the fall months. The effect of add- ing coagulant would be largely if not wholly offset by the removal of the suspended matters. Compared vvith the waters of other cities, that of Louisville would be classed as a satisfactory boiler water. Cincinnati. — Mr. Fuller's investigations at Cincinnati are of later date. He says : " This work had for its ob- ject an investigation in the practicability of the method proposed by the Engineer Commission of 1896, by which, as a part of the extension and betterment of the municipal water supply, the Ohio river water should be partially clari- fied by plain subsidence for several days and then filtered, as suggested, through filters of the English type." " Owing to the fact that the Ohio river water at Cincinnati differs materially in its character for about six months in the year from those waters where this method of purification has long been successful, it was recommended both by the Board of Expert Engineers of 1896 and the Chief and Consulting Engineers of the present Board that, before proceeding further, sufficient reliable data should be ob- tained with reference to the exact local conditions." Ex- perimental filter plants were erected and operated, chemi- cal and bacterial analyses were made, weight of suspended matter determined in special samples, etc., etc. "Specific difficulties and complications in the treat- ment of the local river water by plain subsidence and English filtration " were encountered and there were "features wherein plain subsidence for three days and English (slow) filtration failed essentially in the clarifi- cation and purification." To prolong the average period of plain subsidence " beyond about three days would not be practicable on the ground of cost." The use of coagulants was there- fore found to be imperative at certain periods on account 66 of the fine clay particles contained in the water. Mr. Fuller further states that "so far as present knowledge upon this subject goes, there is only one way in which these clay particles can be removed, and that is to apply a chemical which shall aggregate them into flakes or masses, so that it is practicable to remove them subse- quently by subsidence and filtration." To assist the process of filtration at times of heavy freshets, a coagulant was introduced into the settling basins " only when economical provisions for plain sub- sidence are incapable of preparing the turbid water adequatelj' for filtration, and in such amounts thai the water going upon the English (slow) filters may be properly and readily filtered." He found it essential to allow the coagulated matters suspended in the water to subside in the settling basins so that they would not rapidly close up the pores of the sand layer at the sur- face. "That is to say, the water applied to the English (slow) filters must be substantially free from coagulated masses of clay." Mr. Fuller finally concluded that "it is practicable to clarify and purify the Ohio river water in a satisfactory manner by either the modified English system (slow filtration with accelerated sedimentation) or by the American (mechanical or rapid) system." He recom- mends the adoption of the latter system as " somewhat cheaper," and giving substantially the same quality of filtered water. This decision is of course based on local conditions. Pittsburgh. — A report upon the investigations made by the Pittsburgh Commission was issued in January last. Personal visits were made by its members to the experi- mental plants at Louisville and Cincinnati and several of the members also visited filter plants in Europe. The City of Pittsburgh is supplied with water from the 67 Allegheny and Monongahela rivers, principally from the former, and the water is objectionable on account of the mud it carries and because of its pollution by sew- age. " Elaborate experiments extending over a period of time of sufficient length to show the effect of filtra- tion upon the water of the Allegheny river in all seasons and at all stages " were carried on, and it is stated in the report that the " investigations show the entire feasi- bility of so treating the water by several methods as to remove both the mud and the deleterious vegetalile growths contained therein." Of the various methods of filtration examined, "two have proved themselves efficient, the method of me- chanical (rapid) filtration and the method of sand (slow) filtration" "The latter has yielded upon the whole somewhat better results than the formex." " As is fully- put forth in the report of the Experts employed by the Commission, etc., the method of sand (slow) filtration not only yields a supply of water free from mud and objec- tionable bacterial life, but also furnishes a supply of water of a quality adapted to mechanical purposes, suited to the uses of industrial establishments." These con- clusions, like those reached in Cincinnati, are based on local conditions. The use of meters was also investigated and strongly recommended, it having been concluded, after careful in- vestigation, that the city was wasting "more than twice as much water" as it had any use for. The Pittsburgh (Commission had special experiments made to determine the adaptability of filtered water to use in boilers. It was originally intended to experi- ment at the city's pumping station, but difficulties oc- curred in regard to this, and three, new 2r)-horse-power boilers were loaned to the Commission. These were operated respectively with the effluent from the sand 68 filters, with the effluent from the mechanical filters and with uufiltered river water. The report states that the boiler using unfiltered river water was found in the best condition, and, although considerable scale and sediment were deposited, the deposit was soft, adhered loosely and could easily be washed off and removed. The other boilers showed about the same results, although, in the one in which water from the mechanical or rapid filter was used, -the rivets were badly corroded and a thicker incrustation formed in the tubes than in the one taking its water from the slow filter. ■, Regarding all the methods of purification, the report concludes " that filtration of the Allegheny river re- moves the mud and insoluble matter which would,, by depositing, cause the boilers to be frequently cleaned and washed out. The incrusting properties which remain, while they may not make scale as quickly or as thick as if greater amounts of other material (mud, etc.) were present, yet, when the deposit is formed, it is hard, of a character which gives it the name of 'porcelain scale,' and difficult to remove except by tools." Filtration is no longer an experiment. All filter works, so far constructed and properly operated, have demon- strated their efficiency beyond any question. In comparing the slow with the rapid filter, it' should be borne in mind that any accidental disturbance in the process of filtration is likely to interfere with the purifica- tion of the water nearly in proportion to the rate of fil- tration. Projects Presented. "We present, for your consideration, several projects for the radical improvement of the entire water supply of the city, viz. : A. — 200,000,000 gallons daily of mountain water, from 69 the tributaries of the Upper Lehigh and from the Upper Perkiomen, delivered by gravity into Queen Lane reser- voir. An 8-foot aqueduct, extending from Big creek, on the Lehigh river, to Treichlersville, on the Upper Perkiomen, carries the impounded waters of the Upper Lehigh tribu- taries into one of the reservoirs on the Perkiomen creek. A dam at Green Lane, on the Perkiomen, impounds the combined waters of the two streams, which are carried thence by a 12-foot high-level aqueduct to the Queen Lane reservoir. B. — 200,000,000 gallons daily of mountain water, from the tributaries of the Upper Delaware, near the Water Gap, delivered by gravity into a new reservoir to be con- structed near Twelfth street and Olney avenue. A 1 4 -foot aqueduct extends from the Delaware Water Gap to the proposed new reservoir at Twelfth street and Olney avenue, conveying thereto the impounded waters of the Upper Delaware tributaries, which it receives through several feeders at the Water Gap. C. — 450,000,000 gallons daily of mountain water, from the tributaries of the Upper Lehigh, and from the Upper Perkiomen, delivered by gravity into Queen Lane and East Park reservoirs. The Lehigh aqueduct extends from White Haven, on the Lehigh, to Treichlersville, on the upper Perkiomen, and the combined waters of the Lehigh tributaries and of the upper Perkiomen are carried from the impounding reservoirs at Green Lane through a 12-foot high-level aqueduct to Queen Lane reservoir, and through a 12-foot low-level aqueduct to East Park reservoir. D. 450,000,000 gallons daily of mountain water, viz. : 225,000,000 gallons delivered by gravity from the tribu- taries of the upper Delaware, near the Water Gap, into the new Olney avenue reservoir, and 225,000,000 gallons 70 delivered by gravity from the tributaries of the upper Lehigh and from the upper Perkiomen, into Queen Lane reservoir. The 14-foot Delaware aqueduct extends from the Dela- ware Water Gap to the Olney avenue reservoir, as in Project B, and the 8-foot Lehigh and 12-foot high-level Perkiomen aqueducts extend from Big Creek, on the Lehigh, to Queen Lane reservoir, as in Project A. E. 700,000,000 gallons daily of mountain water, from all of the above-named sources, delivered by gravity into the new Olney avenue reservoir, into Queen Lane reser- voir and into East Park reservoir. This project, forming, in fact, a combination of the preceding projects, is suggested merely to indicate what provision may be made in case the consumption of water should continue increasing as it has done in the past. But, if proper precautious are adopted to prevent unnec- essary waste, so large a quantity as 700,000,000 gallons daily will not be required even fifty years hence. Indeed, it is likely that 450,000,000 gallons per day will meet all requirements at that time. F. 200,000,000 gallons daily, of filtered Schuylkill and Delaware river water, from slow filter at Torresdale, Queen Lane, Roxborough and Belmont and from rapid filters at East Park, delivered by pumpage into existing reservoirs. The Delaware water filtered through slow filters at Torresdale, and the Schuylkill water filtered through slow filters at Roxborough, Queen Lane and Belmont, and through rapid filters at East Park, is pumped into the existing reservoirs. G. 450,000,000 gallons of filtered Schuylkill and Dela- ware river water daily from the same system enlarged. H. 430,000,000 gallons of Delaware river water daily, filtered by rapid filters at Portland,, and delivered by gra- vity into the new Olney avenue reservoir. Mountain Water Supply* SOUECB. Daily Supply. Gallons. Cost of Aque- ducts, storage, etc. Cost of Distri- bution to City Reservoirs. Total Cost. Annual Cost of Operation and Maintenance. Cost per Thou- sand gailont. Cents. per Perklomen creek and Lehigh river tribu- taries 200,000,000 ■150,000,000 150,000,000 $32,090,000 64,590,000 78,630,000 $1,320,000 2,150,000 4,5So,O00 $33,410,000 66,740,000 83,185,000 $1,205,000 2,480,000 2,923,000 1.65 per Perklomeu creek and Lehigh river, with tributaries ~ aware river tributaries near the Water Gap, Upper Perkiomen creek and Lehigh river 1.51 1.78 Slow Filter Supply. SOUBCE. Daily Supply. Gallons. Cost of Filters and Accessory Works. Costof Mains to connect Torres- dale plant with East Park Distri- bution System. Total Cost. Annual Cost of Operation and Maintenance. Cost per Thou- sand gallons. Cents. ,000,000 gallons daily from the Schuylkill river i 50,000,000 gallons daily from the Del- 200,000,000 450,000,000 $9,453,591 05 23,174,691 51 $1,520,000 00 10,980,000 00 $10,973,591 05 34,154,679 51 $1,227,373 35 2,971,801 26 ,000,000 gallons daily from tha Schuylkill river; 300,000,000 gallons daily from the Eapid Filter Supply. Source. Daily Supply. Gallons. Cost of Filters, Aqueducts and Accessory Works. Cost of Distri- bution to City Eeservoirs. Total Cost. Annual Cost of Operation and Maintenance. Cost per Thou- sand Gallons, Cents. ,000,000 gallons daily from the Delaware river, 450,000,000 450,000,000 430,000,000 $62,642,747 00 73,326,052 00 16,798,376 00 $5,320,000 00 5,320,000 00 6,120,000 00 $67,862,747 00 78,645,052 00 21,918,876 00 $3,239,379 21 3,170,805 46 8,108,606 OO 1.97 1.93 1.89 ,000,000 gallons daily, mountain water from storage above Water Gap and 190,000,000 gallons daily from the Delaware, filtered at Portland. ,000,000 gallons daily from the Delaware, fll- * The cost of deliTering 700,000,000 gallons daily of mountalD water into the city reservoirs would be S116,585,000, and the annual expense of operation and maintenance would be S4,310,000. 71 The rapid filters at Portland are supplied with water pumped at that point from the Delaware river, the supply of which, in dry seasons, is to be augmented by a maximum of J 00,000,000 gallons daily of impounded mountain water from the tributaries of the upper Delaware near the Water Gap. Two 12-foot aqueducts carry the filtered water from Portland to the new reservoir near Olney avenue. J. 450,000,000 gallons daily of Delaware river water, filtered by rapid filters at Torresdale, and delivered by pumpage into existing reservoirs. The rapid filter plant at Torresdale is supplied with Delaware river water pumped at that point. After filtration the water is pumped to existing reservoirs. K. 450,000,000 gallons of water daily deliverel by gravity into the new Olney avenue reservoir, viz : — 190,000,000 gallons of Delaware river water taken and filtered by rapid filters at Portland, supplemented by 260,- 000,000 gallons of mountain water brought to Portland by gravity from tributaries of the Delaware river near the Water Gap. The rapid filter plant at Portland is supplied with 190,- 000,000 gallons daily, pumped from the Delaware at that point, and the filtered water' delivered into the aqueducts there is supplemented by 260,000,000 gallons daily of mountain water brought from the tributaries of the upper Delaware through an aqueduct. Two 12-foot aqueducts carry the combined mountain water and filtered Delaware water to the new reservoir near Olney avenue. The costs of construction, operation and maintenance of these various projects are set forth in the table opposite. MuUica River Project. — In addition to the Schuylkill and Delaware rivers and their tributaries, attention was drawn to another source, viz. : a locality some thirty miles 72 southeast of Camden, in the State of New Jersey, compris- ing the area drained by the MuUica river and its branches, and of this some investigation was made. This source appears to offer peculiar advantages in some respects, namely : contiguity to the city, abundance and comparative purity of the water, with but little danger of pollution in the future, and reasonableness in the cost of constructing the works. But the acquisition of the right to draw upon this water supply in another state would require legislation, and this would certainly involve con- siderable delay, if, indeed, the necessary authority could be obtained at all ; a matter which, in the opinion of the Law Department of the city is, at least, doubtful. In the event of adverse action, the whole question of water supply for your city would revert to its present status, with nothing accomplished. We, therefore, abandoned the consideratioji of the subject. The method to be adopted for distributing, the water into the city's reservoirs will depend upon the amount required, and also upon whether it is brought from the mountains or is taken from the Schuylkill and Delaware • rivers near the city. If the water is brought from the Perkiomen and Le- high watersheds, it can be delivered by gravity into Queen Lane reservoir (238 feet above city datum), and into all the other reservoirs of the city, excepting those at Roxborough and Mount Airy, into which the water must be pumped. For the present consumption it would be necessary to build but one aqueduct from the Lehigh watershed into the Perkiomen valley, and thence to the citj^ A consumption of 450,000,000 gallons daily would require the building of another aqueduct, from Perkiomen creek to the city, which would deliver its water into East Park reservoir. If the water is brought from the Delaware Water Gap, 73 it would be delivered by gravity into all the reservoirs of the city, including a new one proposed near Olneyville, excepting the Queen Lane, Roxborough, and Belmont reservoirs, into all three of which the water would have to be pumped from the level of the aqueduct (170 feet above city datum). If the water is obtained near the city from the Schuyl- kill and Delaware rivers, it will require pumping into all of the reservoirs. Mountain Water Supplies. In our estimates of gravity supplies we have given preference to aqueducts of masonry, where these were practicable. In other situations, such as the crossings of deep valleys, we resort to the use of steel pipe. Where mountains or hills are encountered, tunnels are fre- quently necessary. Where steel pipes are used, it is preferable to lay seve- ral of smaller size side by side rather than to use one large pipe of nearly equal cross-section, because, in the former case, the result of a break is far less disastrous. The building of a gravity or aqueduct system in- volves the construction not only of intakes and gate- houses where the water leaves the reservoirs, but also of gate-houses at different points, with arrangements for shutting off, upon occasion, any portion of the pipe. The Delaware river aqueduct begins at the mouth of the Bushkill creek. Pike county, follows the western side of the Delaware river as far down as Point Pleasant, and thence takes the most available course to the city. It collects the water from the various mountain creeks, which is to be stored in large reservoirs to equalize their yearly flow. The Lehigh-Perkiomen aqueduct begins at White Haven on the Lehigh river, follows this river down to 74 near Slatington, and thence takes the most direct course to the Perkiomen watershed, which it reaches through a tunnel terminating near Treichlersville. It collects the water froin the upper Lehigh river and from several mountain creeks, the water of which is stored in large reservoirs for equalization. These waters mingle with those of the watershed, and together they are taken into two aqueducts at Green Lane, one being a high-level and the other a low-level aqueduct, which convey the water to the city. The following plates illustrate the projects for mountain water supply : Plate I. Plan of watersheds of Delaware and Lehigh rivers, and Perkiomen and Tohickon creeks, with aqueducts. Plate II. Profile of Delaware aqueduct. Water Gap to Kintnersville. Plate III. Profile of Delaware aqueduct, Kintnersville to Philadelphia. Plate IV. Profile of Lehigh aqueduct. White Haven to Aquanchicola creek. Plate V. Profile of Lehigh aqueduct, Aquanchicola creek to Treichlersville. Plate VI. Profile of Perkiomen high-level aqueduct. Plate VII. Profile of Perkiomen low-level aqueduct. Plate VIII. Typical aqueduct sections. Filtered Water Sv/p-plies. The slow filters are all designed for an average rate of 3,000,000 gallons of water per acre of effective area (about 9 cubic feet per square foot) per day. The number of filter beds erected at each site at first would be only for present demands, and each plant could, be increased thereafter from time to time, as found necessary, by additional filter beds, 75 ample ground having been reserved for this purpose, ex- cept in the case of the Queen Lane. The area available for slow filters at Queen Lane is limited, so that provision cannot be made at that site to filter more than 58,000,000 gallons per day, although the amount used in that district will hereafter be con- siderably greater. This deficiency will be made up from East Park, and for that purpose high-service pumps at East Park will be required. A rapid filter plant has been adopted at East Park. In considering the Schuylkill and Delaware rivers as sources of supply to be filtered for the city, we have de- cided upon the following main points : a. To utilize and adopt the present plants as far as possible and to the best advantage. b. To use the Schuylkill water for the districts of Bel- mont, Roxborough and Queen Lane, with such surplus as may remain of the limited 150,000,000 gallons supply per day, for East Park. c. To abandon the reservoir at Fairmount, which is now in use only for about seven months in the year, and to connect the turbine pumps with Spring Garden station, so that they may be placed in service whenever the sup- , ply of water will allow, thereby relieving the steam plant of a corresponding amount of work. d. To abandon the Corinthian reservoir. e. To retain the Fairhill reservoir, which, although not now designated for use, will hereafter undoubtedly be found valuable as a centre of distribution for filtered water and can be so. adapted by modification and covering. /. To adopt slow filtration for Belmont, Roxborough and Queen Lane districts, and rapid filtration for such remaining portion of the Schuylkill water as is delivered at East Park. g. To establish a slow-filter plant on the Delaware 76 river below Torresdale, from which all the water not sup- plied from the Schuylkill will be obtained. h. To make use of the present reservoirs, whenever possible, for sedimentation and for the storage of filtered water. j. To allow at least 24 hours for sedimentation, and to provide storage capacity for one-half day's supply of filtered water. k. To cover all storage reservoirs for filtered water. I. To cover all filters. It is, of course, eminently desirable that the water sup- plies for filtration should be as free from impurities as possible, so as to reduce to a minimum the duty on the filters ; and every effort should be made, by legislation and otherwise, to prevent the pollution of streams ; yet such water as exists to-day, in the Schuylkill and Dela- ware rivers at the City of Philadelphia, can be purified by filtration and rendered wholesome and fit, for all do- mestic purposes. Within the city limits it is possible to locate the filter plants at places where the water supplied them will not be subject to direct sewage polution. A point can be selected on the Delaware river within the city limits, but above such direct contamination, and the present intakes on the Schuylkill are well situated in this respect. The locations and conditions of existing pumping stations and reservoirs are such that it is advisable to continue the use of the water in this river up to a quantity equal to its minimum flow, at least so long as the present plant can be made serviceable. For additional supply, and for fu- ture extensions, the Delaware is the proper source, and in time it is not impossible that the whole supply may come from that river. In order to ascertain the suitability of certain sands, obtainable in the vicinity of Philadelphia, for use in 77 filter plants, we have had mechanical analyses made of a number of samples. The results, which are given in Appendix V, indicate that there will be no difficulty in obtaining suitable material for the purpose. If the annual rates remain the same, the surplus earn- ings of the Bureau of Water would, to all appearances, be sufficient to pay for the continual extension of the plant as required by the .growth of the city. Owing to the improvements constantly being made in the operation of filtration plants, it is probable that our estimated cost of filtration will be found, in the future, to have been too high, rather than too low. It will be noticed that the estimated cost of filtering on the Delaware is slightly less than on the Schulkill. When the present reservoirs are converted into settling reservoirs for use prior to the filtration of the water, it will be necessary, in some instances, to re-adjust the water intakes and outlets, so as to accomplish the highest possible degree of sedimentation during the time that the water is paesing through the reservoir. It is advisable that filters and clear-water reservoirs be covered or roofed, to prevent the formation of ice on the surface and to protect the filtered water from pollution by the dust in the air which carries the seeds of lower life. There is abundant evidence of the deterioration of filtered water, or of spring water, kept in open reservoirs. In covered reservoirs, the water is also cooler in summer, than when exposed to sunlight. There is an erroneous idea that sunlight and air are advantageous to stored water. The contrary has been frequently demonstrated, and everyone appreciates the excellence of spring water, which issues, so to speak, from the bottom of a large natural filter, without having been exposed to either sun- light or air. There are both chemical and biological reasons for these facts. 78 The sljw-filter plants contemplated in our recommen- dations are similar, in general arrangement, to those of London and of Hamburg, and to the recently completed filter plant at Albany, N. Y. The latter is the largest fil- tration plant in this country. The following plates illustrate the plans for the several filter plants and show how it is proposed to utilize the present reservoirs for subsidence and for filtered-water reservoirs. Plate IX. Locations of filter plants and mains recom- mended for immediate relief Plate X. Belmont filter plant. Plate XL Roxborough filter plant and Queen Lane filter plant. Plate XII. East Park filter plant. Plate XIII. Torresdale filter plant. Plate XIV. Plan and sections of typical slow- filter. Plate XV. Details of sand washers and regulating apparatus. We have said that we consider it inadvisable dur- ing dry years to obtain a greater amount of water from the Schuylkill river than 150,000,000 gallons per day. A provision for supplying the city with 200,000,000 gallons daily, therefore, requires 50,000,000 gallons a daj'^ to be obtained from the Delaware river ; and all future increase of supply is assumed to be taken from this river. We have selected the neighborhood of Torresdale as the site for the new pumping station oh the river because the present site at Lardner's Point will, in our opinion, not be suitable in the future, on account of the several large sewers now delivering, or which will soon deliver, a large amount of sewage into the river in that neighborhood. From data at hand and from our estimates of the growth of tlie city, we have made the distribution of the total daily quantity of water required as follows : 79 For a supply of 200,000,000 gallons per day : Belmont station -.■27,000,000 gallons dai Eoxborough station 15,000,000 gallons dai Queen Lane station 58,000,000 gallons dai Spring Garden station 50,000,000 gallons dai Torresdale station 50,000,000 gallons dai Making a total of. 200,000,000. gallons dai It is proposed to add, at Belmont station, one new pum ing engine of 20,000,000 gallons daily capacity. For a supply of 800,000,000 gallons per day : Belmont station 37,000,000 gallons dai Roxborough statipn 25,000,000 gallons dai Queen Lane station 58,000,000 gallons dai Spring Garden station 30,000,000 gallons dai Torresdale station 150,000,000 gallons dii Milking a total of 300,000,000 gallons dai It is proposed to erect, at East Park filter plant, t'\ 12,000, 000-gallon pumping engines to parhp from Ee Park reservoir into the Queen Lane ■ district, in order supply the deficiency between the amount pumped ( rectly at Queen Lane station and the consumption Queen Lane district. For a supply of 450,000,000 gallons per day : Belmont station ^55,000,000 gallons dai Eoxborough station 37,000,000 gallons dai Queen Lane station 58,000,000 gallons dai Spring Garden station Torresdale station..... 300,000,000 gallons dai Making a total of 450,000,000 gallons dai If the future water supply is obtained from the rive and filtered, it will be necessary to make a few chang in the pumping machinery and reservoirs. At Fairmount, the reservoir is too low for a prop service, being only 94 feet above tide. Also it is ine 80 pedient to filter the water at this station, and for these reasons we have recommended the abandonment of the Fairmount reservoir. The Spring Garden reservoir; would be of no use in the new apportionment and might be abandoned as a reser- voir, unless retained for the use of Girard College. At Belmont, the present reservoir, with slight altera- tions, could be used as a settling reservoir. A new 20,000,000-gallon pumping engine shouldbe added to the station at the Schu^'lkill river, to be used as a reserve. By limiting to 150,000,000 gallons per day the amount of water to be obtained from the Schuylkill river, it be- came necessary to re-apportion the amounts to be supplied to each station, as it was evident that at the present time more than the above amount is actually pumped from the Schuylkill river. We found it to be more econom- ical, therefore, to limit the amount of water to be sup- plied to the Queen Lane district from the Schuylkill river, and to furnish the deficiency hereafter from the Delaware river. The quantity thus supplied to the Queen Lane district from the Schuylkill river is to be secured from the East Park reservoir by a new pumping station at the East Park filter plant, with proper engine capacity to pump from this resevoir into the Queen Lane district. A portion of the Queen Lane reservoir is to be con- verted into a clear-water reservoir discharging into the City mains. The East Park reservoir, being verj"^ large, will not only serve as a storage reservoir for Schuylkill water, but also for the excess delivered in the future from the Delaware river. A part of this reservoir is to be con- verted into a clear-water reservoir delivering into the city mains. The new Roxborough reservoir is to be kept in use, but a part of it is to be converted into a clear-water reservoir. 81 The Mount Airy reservoir will be used, and the old Roxborough reservoir may be temporarily put out of use. The Fraukford reservoir will be converted iuto a clear- water reservoir. The Lehigh reservoir can be temporarily placed out of use, and eventually converted into a clear-water reservoir if found necessary. In assigning the quantity of water to be supplied to the several districts, and the capacity required of the filtration plant for each, consideration has been given to their probable relative growth and increase in population, as some districts, particularly those of suburban character, will undoubtedly show a much greater annual increase than others. The lower levels in the Roxborough district, now sup- plied from the new reservoir, with its great elevation of 414 feet, could be more economically supplied from the Queen Lane reservoir if proper mains were laid for the purpose. Indeed, a portion of the lower Roxborough dis- trict is already supplied by a main which taps the Queen Lane pumping main near the station. In connection with the gravity supply from the Dela- ware River, a new distributing reservoir near Olneyville, at the point of discharge of the conduit, is proposed. The cost of this reservoir is estimated at $1,000,000. A new reservoir at Belmont, adjoining the present reservoir, for the sedimentation of raw water, may be re- quired when the present consumption has been materially increased. It probably need not be more than half as large as the reservoir recently proposed. As the demands of the Belmont, Roxborough and Queen Lane districts increase, the surplus of the Schuyl- kill water delivered at East Park during minimum flow will gradually diminish; and this deficiency, together with what will be required for increased consumption all 6 82 over the city, is to be supplied from the Delaware through the Torresdale filter plant. When the Schuylkill is flow- ing above its minimum, which will be during most of the year, the supply will be ample to keep the East Park plant in full service as well as the others. To abandon completely at this time the present Schuyl- kill plants would mean the abandonment of much valu able pumping machinery and other works, and also a loss of time in making the change. This change would re- quire not only that a large additional plant be in oper- ation on the Delaware before the Schuylkill plant could be removed, but also the laying of large and costly mains to bring the water to the city. Upon the completion of the proposed Torresdale pump- ing station, the Frankford station would be abandoned. The standpipes at Belmont, Roxborough and Chestnut Hill will be kept in use, but will be supplied with filtered water instead of with raw water. RESUME AND CONCLUSIONS. We now desire to re-state briefly what has been stated at length in the preceding pages, and to present the con- clusions derived from our examinations. The deplorable condition of the City's water supply, which it is sought to remedy, is due to the pollution of its sources, to the lack of effective pumping machinery, and to the insufficient capacity of the distributing system. The question of first importance is the source of supply, and to this nearly all of our thought and time has been devoted. Most of the water is now obtained from the Schuylkill river, within the city limits. Five pumping stations take 83 from it about 200,000,000 gallons daily. One pumping station is located on the tidal estuary of the Delaware river at Lardner's Point, and supplies about 15,000,000 gallons daily. The Schuylkill water is being polluted at many points from its source down to the city line. Beginning with the mine waters, the coal dust and some sewage from the upper parts of the water-shed, the pollution is increased below by the sewage of cities and villages situated along the river and its chief tributaries, by the manufacturing refuse and by the surface water from agricultural dis- tricts, all of which render the water sometimes turbid, unpalatable, impure and dangerous to health. The Delaware water at Lardner's Point is less turbid after rains than the Schuylkill water ; it is also softer and less polluted. Its flow is many times larger. While this water is, therefore, now somewhat better than the Schuylkill water, the growth of the city, the newly-built or projected sewers above and below the intake, and the tidal oscillation of the water, tend to a continually increasing pollution also of the water taken from the Delaware river. It, therefore, becomes imperative, either to select a new source of supply or to improve the present one, so that it will become thoroughly satisfactory to the citizens both as to quality and quantity. The first project requires the bringing of Blue Mountain water to the city ; the second requires a thorough filtration of the Schuylkill and Delaware waters taken within the city limits. A de- cision as to which of these alternative projects is the better one must be based on the quality and quantity of water to be supplied and on the cost. It was, therefore, necessary first to make certain pre- liminary assumptions, then to make designs for both projects, and to ascertain the cost of construction and 84 operation. 'I he assumptions as to population, and as to quality and quantity of water are as follows : The present population, to be supplied from the city's pipe system as soon as practicable, is taken at 1,300,000 persons. The population to be held in view in the de- sign for new works is assumed at 3,000,000 persons. It was considered that the waters collected from the affluents of the Delaware and Lehigh rivers in the Blue Mountains, and from the Upper Perkioraen creek, could be used in their natural condition. While these natural sources are the best obtainable at a reasonable cost, and while their average standard of purit}' is high, it must be remembered that a guarantee against an occasional and temporary pollution of the water by disease germs from man and animals, cannot be given for such large and exposed water-sheds. Nor can an occasional taste, due to vegetal matter, be entirely avoided. The alternative source of supply is the water of the Schuylkill and Delaware rivers, within or near the city limits, artificially purified to the required standard. The purification is obtained bj^ filtering the water through sand ; no better and cheaper method is known. The progress made in this country and in Europe in ascertaining the. laws of the mechanical and biological process of filtration, and the practical success obtained in filtering water for many years in large cities of Europe, confirm and warrant the conclusion that this method of purification can furnish this City, from both rivers, with water that will be clear and palatable, and will conform to the best bacterial and chemical standards. When the raw river water carries much suspended matter with it, this must be allowed to subside, as a pre- liminary to filtration, so as to lengthen as much as practi- ticable the time between the filter cleanings. Settling reservoirs are therefore essential as preliminaries to the 85 filtration of tlie water of tliese two rivers. In order to secure the greatest practicable efficiency, the filter plant must not only be built with skill, and be provided with the best means for regulating the flow, and for cleaning the sand, but it must also be carefully operated by trained men, in accordance with the daily condition of the river water and of the filters. The quantity of water required for city consumption depends on local conditions. In some cities much less water is used than in others. The quantity with which Philadelphia has generally been credited, is somewhat mis- leading, due to the absence of proper measuring appliances; as a matter of fact, it is less than appears on the records. There is also, in this city, an undoubted waste of water, the amount of which cannot now be accurately determined, and.wliich confers no benefit whatever, either to persons or property, or for street or sewer cleaning. It, therefore, subjects the citizens at large to an entirely useless expen- diture, which should be stopped at the earliest practicable moment. We consider that, at present, a daily supply of 200,- 000,000 gallons, being 150 gallons per capita, is a very liberal allowance. We recommend that this quantity of pure water be immediately provided for. At the same rate, a population of 3,000,000 persons will require a daily supply of 450,000,000 gallons. Comparative estimates of cost have been made for eventually supplying these quantities. In order to indi- cate the legitimate outcome of an extravagant use of water, we have made a further estimate of cost for sup- plying the city daily with 700,000,000 gallons of moun- tain waters. The Blue Mountain water projects deliver water to the city reservoirs by gravity. In one, mountain water is ob- tained from the upper Perkiomen creek and from the Lehigh river with its tributaries. In another, mountain water is taken from the Delaware tributaries near the Water Gap. Still another project was considered using the Delaware water at Portland below the Water Gap, but after filtration. Other projects were considered, but were found to possess no special advantages, and were also more expensive. The filtered water project which has been specially considered, is confined to taking water from the Schuylkill and Delaware rivers within the city limits. Two methods of filtration are in common use.; one allows the water to percolate slowly tlirough a bed of sand, while the other allows it to pass through much more rapidly, and, in order to give it the same degree of purity, requires the use of a coagulating substance to prevent objectionable organisms and suspended matter from passing tbro'ugh the filter. The first we have called a slow, and the second, a rapid filtration. Inasmuch as it has been impossible, in the time at our disposal, to make the necessary experiments showing the precise effects of filtering both the Schuylkill and Dela- ware waters, either through slow or rapid filters, it is also impossible now to state which of the two systems would be the more economical. But we know, and can posi- tively assert, from experience obtained elsewhere, that, for the plants which we have recommended, a slow filter sys- tem will not materially differ in annual expense from a rapid filter system. We likewise know that the slow filters, from long experience, and from their successful operation in many cities, can, without question, yield satisfactory results with the waters of the above-mentioned rivers. The rapid filters have only recently been sufficiently de- veloped to command a high degree of confidence in their results under all circumstances. We are of the opinion that for the present supplj', slow filters should be adopted at everj'- station in the city, ex- 87 cepting at the one near East Park reservoir. 'W^e believe that at the latter station a rapid filter plant Avould be more serviceable. A comparison of the estimates of cost shows the follow- ing results : The most economical project for a supply of moiintain water is that taken from the upper Perkiomen and from the Lehigh water sheds. For immediate needs, its cost of construction is |33, 410,000. Its annual cost, for op- eration, interest on investment, and all expenses, to deliver the water into the City reservoirs, is $1,205,000. For a daily supply of 450,000,000 gallons, the total first cost would be $66,740,000, and the annual cost $2,4X0,000. The most economical project for a supply of filtered water is that by which the waters of the Schuylkill and Delaware rivers are filtered within the City limits. Its cost of construction, for present requirements, would be $10,974,000. Its annual cost, for operation, interest and all other expenses, to deliver the water into the City reservoirs, is $1,227,000. For a daily supply of 450,000,000 gallons, the total cost of the filter plant, including special mains from Torres- dale to the centre of the city, would be $34,155,000, and the annual cost $2,972,000. The estimates of cost have shown three important re- sults : 1 . The original cost of any of the mountain water sup- plies is very great for the large quantities of water which the city requirefe. 2. A filtered water supply can be obtained at a first cost which is within the present borrowing capacity of the city, and the plant can be operated at a cost which will, not exceed the probable annual net earnings of the water works. 88 3. The total annual cost of delivering the water into the City reservoirs, by either method, is about the same, and the annual earnings will cover the operation and extension. In conclusion we recommend : 1. The adoption of that project by which the waters of the Schuylkill and Delaware rivers, taken within the City limits, are purified by filtration. 2. The immediate improvement of the existing plant, in accordance with the detailed recommendations of our report. The necessity for the second of these recommendations is manifest. Our reasons for the first are as follows : The entire works can be built for a sum which the City can secure at this time through a loan. A supply of pure water for the entire City can be obtained within a comparatively short time, and the City can thus at an early day be protected against a continu- ance of those diseases which are known to be caused by the present polluted water supply. A filtered water supply, under skillful management,, off'ers a greater security against the effects of accidental pollution of the water, than is possible when the supply is taken from open, unprotected water courses. Filtration can, without difficulty, be made to render the water thor- oughly wholesome. The two large rivers at Philadelphia, or even the Dela- ware river alone, can furnish, at all times, a quantity of water sufficient for a very large city. The foregoing is respectfully presented. RUDOLPH BERING. JOSEPH M. WILSON. SAMUEL M. GRAY. Commissioners. APPENDICES CONTENTS APPENDIX I. Suspended matter in Schuylkill and Delaware Eiver Water ^ 93 APPENDIX, II. Reservoir and Standpipe Data 97 APPENDIX III. Estimates of Cost 101 APPENDIX IV. Annual Earnings and Expenditures of the Bureau of Water facing 124 APPENDIX V. Mechanical Analyses of Sands 124 APPENDIX I. Furnished by Burkau of Health. A Total Residue Precipitated in Parts per Million. Semi-Weehly Observations of Schuylkill River Water. Date. Besldue. Date. Residue. Date. Residue. Date. Residue. 1898. Jan. 3 10. ^?.l'.. 12. July 5 7. Oct. 3 5. 6 3. 7 5. 7 14. 6 8. 10 3. 11 8. 11 17. 10 12. 13 4. 14 8. 14 13 13 12. 17 47. 18 9. 18 22. 17 11. 20 7. 21 14. 21 9. 20 14. 24 183. 25 17. 25 13. 24 15. 27 53. 28 22. 28 27. 27 19. 31 Feb. 3 10. 5. May 2 5 13. 13. Aug. 1 ■ 4 16. 21. 81 Nov. 3 20. 14. 7 2. 9 149. 8 107. 7 lo: ' - 10 n. 12 37. 11 152. 10 8. 14 16. IB 88. 15 41. 14 20. 17 19. 19 23. 18 29. 17 9. 21 545. 2i 51. 22 80. 21 40. 24 3S. 26 28. 25 25. 24 14. 28 2. 30 18. 29 17. 28 15. March 3 6. June 2 28. Sept. 1 .... 10. Dec. I 5. 7 2. 6 12. G 11. 5 .... 416. 10...... 6. 1 9 9. 8 20. 8 71. 14 11. 13 10. 12 16. 12 6. 17 7. 16 2il. ! li 7. 15 14. 21 23. 2'i 17. 19 7. 19 7. 24 .18. 23 16. 2.' 11. 22 52. 28 1-'. 27 13. 26 7. 27 32. 31 50. 31 3.<. 29 7, 29 18. 94 Total Residue Precipitated in Parts per Million in Schuylkill River Water — Continued. Date. Residue. Date. Residue. Date. Residue. Date. Residue. 1899 Jan. Mar. May. June. 5 9. 2 77. 4 15. 26 10. 9 17. 6 325. 8 7, i 29 17. 12 6. 9 27. 11 15. July. 1 3...... 104. 16 5. 13 1026. 15 15. 24 15. 19...... 48. 23 130. 18 20. 27 15. 23 4. 27 23. 22 21. 31 17. 26 403. SO 116. 2!) 18. Aug. 3 31. 30 13. April. 29 8. 4 21. 7 48. Feb. June. 2 5. 6 10. 1 17. 10 12. 6 26. 10 92. 5 9. 14 35. 9 8. 20 21. 8 8. 17. ... 71. 15 .6. 24 21. 12 15. 21 12. 20 38. 27 36. 15 7. ( 24 10. 23 224. May. 19 9. 28 48. 27 • 642. 1 10. ■ 22 22. 31 41. 95 B Total Residue Precipitated in Parts per Million. Daily Observations. Date. Delaware Water AT Laednek's Point. Schuylkill Water AT Spring Garden. By acid. In24hrs. In48hrs. By acid. In 24 hrs. In 48 hrs. 1899. July 19 16. 9.1 7.5 13.5 6.5 7.5 8.6 14.6 9. 6. 5. 10. 13.5 9.9 10.6 17.6 12.6 3.7 5. 12.5 6.5 6.5 6.5 8.5 20 21 '. 22 23 24 10. 15. 16. 10. 14.4 15. 9.5 12.5 10.5 17.5 7.5 17. 9.6 19.6 9.6 9.1 9.1 18.6 14.5 13.5 18.5 16.5 10.5 4.5 23.6 16.6 19.6 22.6 14.6 6.6 7. 10.5 19.5 7.6 7.5 12 25 26 27 28 29 30 31 26. 24.5 2? 41.7 24. 15. 15. 13. l.i. 16. 22. 24. 29.5 22.0 19.5 46. 21. 14.5 15. 15. 14. 16. 24. 21. 29.5 22.5 ■11. 50. 26. 16. 16. l.i. 18. 15. 27. 22. 16.5 18.5 28.5 39.6 49. 23. 33. 27. 28. 19. ISJ. 362. 103. 63. 67. 104. 78. 19.6 16.5 26. 31. 59.5 30. 36. 26. 22. 13. 130. 315. 72. 40. 48. 57. 47. 2 ''9 8 38 4 63 n 34. 39. 8 3-'. 9 27. 10 17. 1 1 144. 12 334. 13 74. 14 20. 29. 29. 31. 16. 24. 22. 2.'. 18. 27. 20. 23. 42. 15 48. 16 71. 17 50. 96 Total Residue Precipitated in Parts per Million. Daily Observations — Continued. Date. 1899. August 18 19 20 21 22 23 ■ 24 25 26 27 28 29 30 31 beptember ].. 2.. 3.. 4.. 5.. 6.. 7.. 8.. a.. lo- ll.. 12.. 13.. 14.. 15.. 16.. Delawake Water AT Laedner's Point. By Acid. 31. 20. 15. 19. 31. 19. 14. n. 14. 9. 19. 17. 13. 18. 12, 14. 12. 14. 13. 15. 17. 20. 15. 19. 14. 17. 14. 18. 20. 14. In 24 hrs. In 48 hrs. 24. 18. 17. 20. 23. 15. 13. 9. 14. 11. 11. 16. 12. 14, 12. 13. 14. 14, 12. 11. 17. 21. 15. 16. 12. 14. 1.3. 19. 21. 13. ScHnYLKiLL Water AT Spring Gardrr, By Add. 47. 21. 23. 16. 19. 23. 19. 14. 11. 25. 55. 44. 69. 53. i5. 26. 16. 30. 32. 22. 30. 25. 28. 28. 22. 23. 25. 23. 22. 17. In 24 hrs; 25. 15. 14. 14. 19. 13. 16. 8. 11. 30. 47. 38. 48. 40. IS. 12. 16. 22. 23. 21. 26. 15. 16. 19. 23. 21. 21. 21. 18. 16. In 48 hrs. 27. 17. 18. 15. 19. 18. 13. 9. 15. 30. 47. 43. 43. 38. 19, 14, 18. 34. 28, 19, 24, 12. 16. 18. 21. 23. 21. 24. 19. IG. 97 ■ » .Sh X ^^ e Q « Z LU g Q- « 0- < ■| f^ oa a d C3 1 "S 1 £) ^ ^ i 1 0) 03 -t^ -tj "S a cd s & &= ^< ti tCl ce .s % "3 M 1 o s s d fl a d S ^ ^ TS a t) Td ^ (« -S o V 0) -S i 3 i 1 d 1 e3 d Cm O o tM o C-H £ O O c C4 o o o iJ o s 3 ■s _^ S ^ "5 >. u d Eh o 1 bn cd 1~, d o T3 1 :3 1 ft a ft i ft -3 d o a ■i ft Ph 1 1 3 1 1 d p. a 1 d § ft tJ fO ^3 S g S d d d £ . Surface igh Wat bove Cit Datum. Feet. tH o o M M CO CO «3 "^ t-- t1< ■^ Oi M (M CO CO «D lo to Cf QO ^ (N CO CO "* " cc "* -* M»= APPENDIX III. ESTIMATES OF COST. Construction, Operation, and Maintenance. APPENDIX III. BASIS OF ESTIMATES. Cost op Pumping, Per million gallons raised one foot high, including coal, labor, oil, waste and supplies, and ordinary repairs ; but excluding interest and depreciation : High-Lift Pumps. For a daily supply of 200,000,000 gallons, 300,000,000 gallons, 450,000,000 gallons, 3.5 cents. 3.25 cents. 3.0 cents. Low-Lift Pumps. For a daily supply of 200,000,000 gallons, 300,000,000 gallons, 450,000,000 gallons, 5.25 cents. 4.875 cents. 4.5 cents. For stations away from railroad sidings add to all above prices 0.5 cent. Cost of^ Filtration, Per million gallons of filtered water, including labor, cost of wash and waste water, lost sand, sanitary analyses of water, chemicals, superintendence, watchmen, or- dinary repairs, and all incidental expenses; but excluding interest, depreciation and cost of pumping water to filters : a„>,„nii,;ii T);.,a» Delaware Hiver, Schuylkill River, ^t Portland. At Torresdale. Slow filters 13.60 13.00 Eapidfilters 4.80 ^3.20 4.00 102 Interest and Depreciation. Interest on cost of works is assumed at 3 % . Depreciation of works is assumed as follows : structures, Apparatus, etc. MasoDry Conduits Covered Masonry Filter Beds Covered Reservoirs Permanent Buildings Cast-iron Pipe Railroad Sidetracks Steel Pipe Air Valves, Blow-offs and Gates on Pipe Lines. Engines and Pumps Boilers. Rapid Filters and Appurtenances Electric Light Plants Tramways and Equipment Iron Fences Telephone Lines Sand-washer Apparatus Regulating Apparatus for Slow Filters Life, in Years. Permanent. Permanent. Permanent. 100 80 80 35 35 30 20 20 20 20 20 10 10 10 . Annuity on One Dollar. .00165 .00311 .00311 .01654 . .01651 .02102 .03722 .03722 .03722 .03722 .03722 .08724 .08724 .08724 Unit Pieces. Aqueducts, Tunnels, Steel-Pipe Lines and Dams. Clearing and grubbing, per acre $100 00 Earth excavation, per cubic yard 30 Borrowed earth, per cubic yard, per 1,000-foot haul 20 Overhaul, 5 cents per 1,000 feet ; limit, 60 cents. Eock excavation, granite, etc., in tunnel, per cubic yard 5 00 Eook excavation, shale and soft rock in tunnel, per cubic yard. 4 00 Kock excavation, open out, per cubic yard 1 20 Shaft excavation, per linear foot, Perkiomen 40 60 Shaft excavation, per linear fool, Lehigh and Delaware 100 00 Dry rock filling over arch in tunnel, per cubic yard 2 50 103 Unit prices, continued. Rubble masonry filling in tunnel, per cubic yard Brick masonry in tunnel, per cubic yard Brick masonry in trench, per eubic yard Arch culvert masonry, per cubic yard Kectangular culvert masonry, per cubic yard Foundation masonry, per cubic yard Betaining walls and cradling, per cubic yard Bubble stone masonry in trench, per cubic yard '. Paving, per cubic yard Portland cement concrete in tunnel 1:3:6 Portland cement concrete in trench 1:3 6 Portland cement concrete 1 : 2J: 4J Portland plastering on arch, per linear foot, 12-foot aqueduct.. Portland cement wash, invert and sides, per linear foot of aqueduct Eiveted-steel pipes, coated and erected, per pound Manholes, each Blow-offs for 72-inch and 80-inch pipes, each... Blow-offs for 60-inch pipes, each Air valves for 72-inch and 80-inch pipes, each Air valves for 6U-inch pipes, each Gate-houses, including gates etc. With 4 lines of 80-inch pipe, each |40,000 00 With 3 lines of 80-inch pipe, each 35,000 00 With 3 lines of 72-inch pipe, each 30,000 00 With 2 lines of 72-inch pipe, each 25,000 00 With 2 lines of 60-inch pipe, each 25,000 00 Railroad crossings, each, extra 2,000 00 Culverts over brooks, each, extra 10,000 00 Telegraph and telephone lines, per mile 600 00 Stone wall fence (two sides), per linear foot 40 Iron fence (two sides), per linear foot 2 00 Dressing and seeding banks, per square yard 08 $5 00 12 00 12 00 15 00 10 00 6 00 5 00 5 50 2 00 7 00 6 00 7 50 26 08 06 15 00 500 00 300 00 150 00 100 00 104 MOUNTAIN WATER SUPPLY. Cost of Construction. Summary. A.— For a daily supply of 200,000,000 gallons from the Upper Perkiomen creek and Lehigh river tributaries. Perkiomen high-level aqueduct, 12-foot diameter $9,490,000 Lehigh aqueduct, 8-foot diameter 6,550,000 Storage, Perkiomen 8,330,000 Storage, Lehigh 7,720,000 $32,090,000 Distribution to Belmont reservoir $245,000 Distribution to Eoxborough reservoir, including pumping station 370,000 Distribution to East Park reservoir 235,000 Distribution to Wentz Farm reservoir 470,000 $1,320,000 Total ,. $83,410,000 B.— For a daily supply of 200,000,000 gallons, from tribu- taries of Delaware river near the Water Gap. Delaware aqueduct, 14-foot diameter $31,690,000 Storage, Delaware 14,850,000 $46,540,000 200,00p,000-gallon reservoir, near Twelfth street and Olney avenue $1,000,000 $1,000,000 Total $47,540,000 C— For a daily supply of 450,000,000 gallons from the UpperPerkiomen creek and Lehigh river, with tribu- taries. Perkiomen high-level aqueduct, 1 2-foot diameter. . $9,490,000 Perkiomen low-level aqueduct, 12-foot diameter... 9,050,000 Lehigh aqueduct 18,700,000 Storage, Lehigh 19,020,000 Storage, Perkiomen 8,330,000 $64,590,000 105 Distribution to Belmont Reservoir from Queen Lane |245,000 Distribution to Belmont Eeservoir from East Park. 155,000 Distribution to Eoxborough Reservoir, including pumping station 525,000 Distribution to Wentz Farm Reservoir 990,000 Distribution 4o East Park 235,000 12,150,000 Total 166,740,000 D. — For a daily supply of 450,000,000 gallons from the Delaware river tributaries, near the Water Gap, from the Upper Perkiomen creek, and from the Lehigh river tributaries. Delaware aqueduct, 14-foot diameter 131,690,000 Storage, Delaware 14,850,000 Perkiomen high-level aqueduct 9,490,000 Lehigh aqueduct, 8-foot diameter 6,550,000 Storage, Perkiomen , 8,330,000 Storage, Lehigh 7,720,000 178,630,000 Distribution to Belmont reservoir $475,000 Distribution to Roxborough reservoir, including pumping station 525,000 Distribution to East Park reservoir from Olney avenue 1,500,000 Distribution to Wentz Farm reservoir from Queen Lane 470,000 Distribution to Wentz Farm reservoir from Ol- ney avenue 350,000 Distribution to East Park reservoir from Queen Lane 235,000 200,0U0,000-gallon reservoir near Twelfth street and Olney avenue 1,000,000 4,555,000 Total $83,185,000 106 Annual Cost of Operation and Maintenance. Summary. A.— For a daily supply of 200,000,000 gallons from the Upper Perkiomen creek and Lehigh river tributaries. Interest on $33,410,000 $1,002,300 Depreciation of works 91,820 Sanitary inspection 8,040 Ordinary repairs 23700 Keepers' wages and pumping 68,490 Sanitary analyses of water 11,000 Total $1,205,350 Say $1,205,000 Cost per 1,000 'gallons for the water delivered into the city reservoirs 1.65 cents. C. — For a daily supply of 450,000,000 gallons from the Upper Perkiomen creek and Lehigh river, with tribu- taries. Interest on $66,740,000 $2,002,200 Depreciation of works 218,450 Sanitary inspection 15,640 Ordinary repairs 50,860 Keepers' wages and pumping 166,650 Sanitary analyses of water 25,000 TotaJ $2,478,800 Say ■ 2,480,000 Cost per 1,000 gallons delivered into the city reservoirs 1 51 cents. 107 D.— For a daily supply of 450,000,000 gallons from the Delaware river tributaries near the Water Gap, from the Upper Perkiomen creek, and from the Lehigh river tributaries. Interest on 183,185,000 $2,495,550 Depreciation of works 198,640 Sanitary inspection 16,620 Ordinary repairs 49,150 Keepers' wages and pumping 140,770 Sanitary analyses of water 25,000 Total $2,925,730 Say 2,925,000 Cost per 1,000 gallons delivered into the city reservoirs 1.78 cents. SLOW FILTER SUPPLY. Cost of Construction. Summary. F.— For a daily supply of 200,000,000 gallons : 150,000,- 000 gallons from the Schuylkill and 50,000,000 gal- lons from the Delaware river, near the city. Belmont filter plant, complete for 27 millions daily $1,802,786 00 Eoxborough filter plant, complete...for 15 millions daily..... 729,099 31 Queen Lane filter plant, complete... for 58 millions daily 2,416,566 30 East Park filter plant, complete for 50 millions daily 1,288,740 89 Torresdale filter plant, complete for 50 millions daily 3,216,398 55 Totals 200 millions daily 19,453,591 05 Mains to connect Torresdale filter plant with East Park dis- tribution system 11,520,000 00 $10,973,591 05 108 G.— For a daily supply of 450,000,000 gallons : 150,000,- 000 gallons from the Schuylkill and 300,000,000 gallons from the Delaware river, near the city. Belmont filter plant, complete for 55 millions daily $3,751,386 00 Eoxborough filter plant, comi5lete..for 37 millions daily 1,782,457 79 Queen Lane filter plant, complete..for 58 millions daily 2,416,566 30 East Park filter plant, complete 1,594,640 89 Torresdale filter plant, complete. ..for 300 millions daily 13,629,628 53 Totals 45e millions daily $23,174,679 51 Mains to connect Torresdale plant with East Park distribu- tion system 10,980,000 00 $34,154,679 51 AxxuAL Cost op Operation and Maintenance. Summary. F.— For a daily supply of 200,000,000 gallons from the Delaware and Schuylkill rivers, near the city. Interest on $10,97M,591.05 $329,207 74 Depreciation of plant 57,916 25 Cost of pumping into reservoirs 566,499 36 Cost of filtering water 273,750 00 Total $1,227,373 35 Cost per 1,000 gallons for the filtered water delivered into the city reservoirs 1.68 cents. G.— For a daily supply of 450,000,000 gallons from the Delaware and Schuylkill rivers near the city. Interest on $34,154,679.51 $1,024,640 39 Depreciation of plant 205,539 65 Cost of pumping into reservoirs 1,216,021 22 Cost of filtering water 525,600 00 Total $2,971,801 26 Cost per 1,000 gallons for the filtered water delivered into the city reservoirs 1.81 cents. 109 RAPID FILTER SUPPLY. Cost of Construction. Summary. H.--For a daily supply of 450,000,000 gallons from the Delaware river at Portland (two 12-foot aqueducts to Philadelphia). Cost complete, including aqueducts, filter plants, all accessory- works and distribution pipes to city reservoirs 167,862,747 J.— For a daily supply of 450,000,000 gallons from the Delaware river at Torresdale. Cost complete, including filter plant, pipe lines and distribu- tion pipes to city reservoirs.: 121,918,376 K.— For a daily supply of 450,000,000 gallons : 260,- 000,000 gallons of mountain water from storage in the Delaware watershed above the Water Gap, and 190,000,000 gallons from the Delaware river filtered at Portland with rapid filters (two 12-foot aqueducts to Philadelphia.) Cost, including storage reservoirs, aqueducts, rapid-filter plant, all accessories, and distribution pipes to city reservoirs $78,645,052 no Annual Cost op Operation and Maintknance. S^immary. H. — For a daily supply of 450,000,000 gallons from the Delaware river at Portland. Interest on $67,882,747 12,035,882 41 Depreciation of plant 222,558 80 Maintenance of aqueducts, storage reservoirs, etc 46,400 00 Cost of pumping into city reservoirs 408,938 00 Cost of filtering water 525,600 00 Total $3,239,379 21 Cost per 1,000 gallons for the filtered water delivered into the city reservoirs 1.97 cents. J, — For a daily supply of 450,000,000 million gallons from the Delaware river at Torresdale. Interest on $21,918,376 $657,551 00 Depreciation of plant 311,670 00 Cost of pumping into reservoirs 1,411,796 00 Cost of filtering water 727,589 00 Total $3,108,606 00 Cost per 1,000 gallons for the filtered water delivered into the city reservoirs 1.89 cents. K.— For a daily supply of 450,000,000 gallons: 260,000,- 000 gallons of mountain water from storage in the Delaware watershed above the Water Gap, and 100,- 000,000 gallons from the Delaware river filtered at Portland with rapid filters. Interest on $78,645,052 $2,359,351 56 Depreciation of plant " 202,465 90 Maintenance of aqueducts and storage reservoirs 63,540 00 Cost of pumping into city reservoirs 323,528 00 Cost of filtering water 221,920 CO Total $3,170,805 46 Cost per 1,000 gallons for the water delivered into the city reservoirs 1.93 cents. Ill SLOW FILTER SUPPLY. Cost op Construction. Belmont Filter Plant. Capacity, 27,000,000 gallons daily. Land $322,500 00 Excavation....'. 31,200 00 Piping, including specials..... 211,022 00 Piping for sand washers 800 00 Drains 1,530 00 Sand washers 2,400 00 13 Filter beds, complete 490,568 00 Pumping machinery 225,000 00 Filtered-water reservoir, (capacity, 15,000,000 gallons) 220,000 00 Electric light plant 10,000 00 Double-track tramway, cars and equipment 5,120 00 Eesidence for superintendent 5,000 00 Proportional part of cost of bacteriological laboratory 3,500 00 Office and store-room 5,000 00 Shelter, lunch-room and conveniences 10,000 00 Fencing 15,(!00 00 Cleaning up, etc.: 9,000 00 ^1,567,640 00 15 per cent 235,146 00 Total $1,802,786 00 ]12 Belmont Filter Plant. Capacity, 55,000,000 gallons daily. Land $322,500 00 Excavation ' fi2,500 00 Piping, including specials 387,205 00 Piping for sand washers 1,200 00 Drainpipe 3,9]0 00 Sand washers 4,200 00 26 Filter beds, complete 974,910 00 Pumping plant at sedimentation basins 424,000 00 Piltered-water reservoir, (capacity, 26,000,000 gallons) 400,000 00 Additional sedimentation basin at Belmont reservoir 100,000 00 Electric light plant , 15,000 00 Additional pumping plant at river station 484,300 00 Double-track tramway, cars and equipment 11,250 00 Residence for superintendent 5,000 00 Proportional part of cost of bacteriological laboratory 6,100 00 Office and store-room 5,000 00 Shelter, lunch-room and conveniences 20,000 00 Fencing 15,000 00 Cleaning up, etc 20,000 00 $3,262,075 00 15 per cent 489,311 00 Total $3,751,386 00 113 ROXBOEODGH FiLTER PlANT. Capacity, 15,000,000 gallons daily. Land $35,000 00 Excavation 19,250 40 Piping, including specials 64,242 00 Water mains for sand washers 1,000 00 Drain pipe 1,130 00 Sand washers 1,200 00 8 Filter beds, complete 303,697 00 Pumping plant 58,400 00 Roofing filtered-water reservoir 106,480 00 Double-track tramway, cars and equipment 2,800 00 Residence for superintendent 5,(^00 00 Proportional part of cost of bacteriological laboratory 2,000 00 Office and store-room 5,000 00 Shelter, lunch-room and coQveniences 10,000 00 Fencing 10,800 00 Cleaning up, etc 8,000 00 $633,999 40 15 per cent 95,099 91 Total $729,099 31 114 RoxBOROUGH Filter Plant. Capacity, 37,000,000 gallons daily. Land $35,000 00 Excavation 38,526 80 Piping, including specials 113,540 50 Water mains for sand washers 1,500 00 Drainpipe 2,510 00 Pand washers 3,000 00 18 Filter beds, complete 680,536 00 Pumping Plant : At river station $174,800 00 Atfilters 87,200 00. 262,000 00 New force main from river station to reservoir 142,500 00 Eooflng filtered- water reservoir 212,960 00 Double-track tramway, cars and equipment 6,000 00 Residence for superintendent 5,000 00 Proportional part of cost of bacteriological laboratory 4,100 00 Office and store-room 5,000 00 Shelter, lunch-room and conveniences 15,000 00 Fencing 10,800 00 Cleaning up, etc 12,000 00 $1,549,973 30 15 per cent 232,496 00 Total $1,782,469 30 115 Queen Lane Filter Plant. Capacity, 58,000,000 gallons daily. Laud $310,000 00 Excavation 66,382 50 Piping, including specials 158,331 50 Piping for sand washers 3,000 00 Vitriiied drain pipe 7,390 00 Sand washers .: 4,200 00 27 Filter beds, complete 1,009,949 00 Pumping plant at sedimentation reservoirs 100,000 00 Roofing filtered-water reservoir (capacity 40,000,000 gallons) 349,169 00 Electric light plant 15,000 00 Double-track tramway, cars and equipment 11,840 00 Residence for superintendent 5,000 00 Office and store-room 8,000 00 Shelter, lunch-room and conveniences 20,000 00 Fencing 15,600 00 Proportional part of cust of bacteriological laboratory 7,500 00 Cleaning up, etc 10,000 00 $2,101,362 00 15 per cent 315,204 30 Total $2,416,566 30 East Park Filter Plant. Capacity, 50,000,000 gallons daily. Bapid-filter plant, complete $347,112 00 Building, including stack 215,000 00 Piping and specials, outside of building ,... 51,850 00 Pumping plant for rapid filters 55,500 00 Roofing filtered-water reservoir 421,182 25 Changing roadways, sodding, etc 25,000 00 Proportional part of cost of bacteriological laboratory 5,000 00 $1,120,644 25 15 per cent 168,096 64 Total $1,288,740 89 116 East Park Filter Plant. No water to be filtered at this station when total city consumption is 450,000,000 gallons daily. Cost of plant as above 11,120,644 25 New pumps to supply deficiency of 32,000,000 gallons daily in Queen Lane district 266,000 00 11,386,644 25 15 per cent 207,996 64 Total $1,594,640 89 TORRESDALE FiLTEE PlANT. Capacity, 50,000,000 gallons daily. Land $284,800 00 Excavation 55,382 80 Piping, including specials 77,014 50 Piping for sand washers 2,000 00 Drain pipe 6,060 00 Sand washers 3,000 00 24 Filter beds, complete 904,236 00 Pumping plant 848,655 00 Filtered-water reservoir at Torresdale 60,000 00 Filtered-water reservoir at Wentz farm 175,000 00 Sedimentation reservoirs 307,420 00 Double-track tramway, equipment and cars 9,700 00 Residence for superintendent 5,000 00 Proportional part of cost of bacteriological laboratory 7,000 00 Office and store-room 8,000 00 Shelter, lunch-room and conveniences 20,000 00 Fencing .5,000 00 Sidetrack from Pennsylvania Eail road -8,000 00 Cleaning up, etc 10,000 Oq $2,796,868 30 15 percent 419,530 25 Total $3,216,398 55 in TORRESDALTS FiLTER PlANT. Capacity, 300,000,000 gallons daily. Land $284,800 00 Excavation 314,505 80 Piping, including specials 504,354 50 Raw-water conduit 224,964 00 Filtered-water conduit, including manholes, gate houses, etc... 77,975 60 Water mains for sand washers 13,000 00 Drain pipe..., 41,648 00 Brick drains ; 46,226 00 Sand washers 21,600 00 138 Filter beds, complete 5,159,087 00 Pumping plant 2,325,620 00 nitered-water reservoir at Torresdale 322,500 00 Filtered-water reservoir at Wentz farm 528,000 00 Sedimentation reservoirs 1,712,420 00 Double-track tramway, cars and equipment 54,800 00 Residence for superintendent 5,000 00 .Proportional part of cost of bacteriological laboratory 32,300 00 Office and store-room 30,000 00 Shelter, lunch-room, and conveniences 100,000 00 Fencing 15,000 00 Sidetrack from Pennsylvania Railroad 8,000 00 Cleaning up, etc 3u,000 00 $11,851,850 90 15 per cent 1,777,777 63 Total $13,629,628 53 118 MOUNTAIN WATER SUPPLY. Cost of Construction. Storage. Reservoirs, Intake-dams, Gmnecting Pipes and Accessory Works. Perkiomen watershed $3,330,000 00 Lehigh watershed : Above Big creek 10,650,000 00 Big and Aquanchicola creeks 8,370,000 00 Delaware watershed 14,850,000 00 20 per cent, of the area of each reservoir is assumed to- be stripped 12 inches deep. Perkiomen High-Level Aqueduct. Green Lane to Queen Lane Reservoir. Diameter, 12 feet. Capacity, 225,000,000 gallons daily. Miles. Total cost. Cost per foot. Description. Aqueduct.... Tunnel Steel Pipe... 14.00 6.30 7.80 83,105,00U 2,170,000 4,215,000 $42 20 65 30 101 40 Slope, .000167. Slope, .000167. 4 sections of four 80-inch pipes ; slope, .0003. Bth section— three 72-incli pipes; slope, .001. Totals... 28.10 ?»,490,000 Average cost per foot, $64 00 119 Perkiomen Low-Level Aqueduct. Green Lane 1o East Park Reservoir. Diameter, 12 feet. Capacity, 225,000,000 gallons daily. From To Miles. Total cost. Cost aer foot Description. Green Lane... Frederic Frederic Frederic Wissahickon Frederic Wissahickon Wissahickon Wissahickon East Park reservoir ... 4.89 18.73 4.83 1.08 2.37 81,226^000 4,845,000 1,600,000 755,000 fi25,000 $47 50 49 00 62 80 132 50 50 00 Two 72 -inch pipes; slope, .0025. Aqueduct, 12-foot diameter; slope, .000167. Tunnel, 12-foot diameter; slope, .000167. Three steel pipes, 72-inch dia- meter ; slope, .001. One 72-inch steel pipe, one 60-inch steel pipe ; slope, .0033. Totals 31.9 $9,050,000 Average cost per foot, $53.3' 120 Lehigh Aqueduct. White Haven to Aquanchicola Creek. Section. From To Miles. Total cost. Cost per foot. I. II. III. IV. White Haven... Muddy Run Bear Creek Big Creek.. Muddy Run Bear Creek Big Creek Aquanchicola... 8.24 C.91 8.33 6.35 81,660,000 2,050,000 2,860,000 2,830,000 $38.20 56.10 65.00 84.50 29.83 S9,400.000 J59.70 (Average) Aquanchicola Creek to Reservoir near Treichlersville. Section "V. Miles. Total cost. Cost per foot. Description. Masonry Aqueduct 20.2 5.09 3.39 0.95 85,345,000 1,605,000 2,230,000 120,000 850.10 59.60 124.50 24.00 12-foot aqueduct ; slope, .00015. 12-foot aqueduct ; slope, .00045. Four 80-inch pipes; slope, .001. Steel Pioe 29.63 89,300.000 839.40 (Average) NOTES. Section I : Capacity, 200,000,000 gallons daily. 14 300 feet^ two 5-foot pipes; slope, .0055. 29 200 feet, one 10-foot aqueduct ; slope, .0005. Section II: Capacity, 250,000,000 gallons daily. 25 100 feet, two 5-foot pipes ; slope, .008. 700 feet, two 6-foot pipej ; slope, .004. 10 700 feet, 8-foot tunnel ; slope, .002. Section III : Capacity, 300,000,000 gallons daily, 34 800 feet, two 6-foot pipes; slope, .005. 9 200 feet, 10-foot tunnel 5 slope, .001. Section IV : Capacity, 350,000,000 gallons daily. Three 80-ineh pipes ; slope, .002. Section V : Capacity, 350,000,000 gallons daily. 121 Delaware Aqueduct. Water Gap to Portland. Diameter, 12 feet. Slope, .0003. Capacity, 260,000,000 gallons daily. Length, 5.86 miles. Cost, complete, $2,112,540. Portland to Point Pleasant. Diameter, 14 feet. Slope, .0001 «7. Capacity, 260,000,- 000 gallons daily. Total cost. Length. Cost. Feet. Miles. Per Foot. Per Mile. $14,220,000 2,200,000 1,730,000 ■ 208,200 29,500 11,200 39.43 5.59 2.12 W8 30 74 60 154 00 $360,000 394,000 810,000 Tunnel Steel pipe - »18,lo0,000 248,900 47.14 »72 90 (Average.l S3S5,O0O (Average.) Point Pleasant to Philadelphia. Diameter, 14 feet. Slope, .000167. Capacity, 260,000,- 000 gallons daily. Total cost. Length. Cost. Feet. Miles. Per Foot. Per Mile. Aqueduct $7,990,000 2,693,000 743,000 130,600 38,400 '4,700 24.73 7.27 0.89 $61 20 70 20 158 50 $323 000 .S7o,ono 837,000 $ll,430,0-)0 173,700 32.89 $6.5 80 (Average.) $347,000 (Average.) 122 Expenditures Necessary to Put the Present Works INTO Proper Condition. Fairmount Pumping Station. River wall at Fairmount forebay $16,500 00 Eoof and improvements to Fairmount western pump house.. 20,000 00 Spring Garden Pumping Station. Repairs to Cramp pumping engine No. 7 5,000 00 Repairs to Holly pumping engines Nos. 2 and 3 16,500 00 Repairs to Gaskill pumpingengineNo.il 6,000 00 Building conduits and filling forebay 25,000 00 Belmord Pumping Station. House for Worthington pumping engine No. 4 and for hous- ing additional pumping engines 25,000 00 Repairs to Worthington pumping engines Nos. 1, 2 and 3 3,000 00 Additional pumping machinery 75,000 00 Queen Lane Pumping Station. Relaying suction mains and building new pump well 35,000 00 Tunnel and coal shed 36,000 00 Poxborough Pumping Station. Repairs to Worthington pumping engines. 1,500 00 Total $263,500 00 As the labor at the stations will be furnished by the De- partment, it is not included in the above estimates of cost. Fairmount and Flat Rock Dams. Without more knowledge of the interior condition of the dams at Fairmount and Flat Rock, it is not possible to approximate the cost of repairing them, and, there- fore, no amount is allowed in the above estimate of cost for this work. 123 Additional Mains for the Distribution System. Belmont district $372,000 00 Koxborough district 535,000 00 Queen Lane district 440,000 00 East Park district 770,000 00 Frankford district 650,000 00 $2,767,000 00 Mains to connect Pairmount pumping station with East Park reservoir, and alterations' in pumps (approximately) 260,000 00 Meters. Repair shop and testing plant |15,000 00 Tools, etc , 5,000 00 Purchase of meters 80,000 00 Total... $100,000 00 APPENDIX V. Mechanical Analysis of Sands. Cape May beach sand Delaware bar sand Gloucester sand— fine Bar sand Bank sand — Kancocas creek..,. Cape May beach sand— fine White Jersey sand White sand — Gloucester , Delaware bar sand — fine Gloucester sand— rough Jersey gravel — Rancocas creek. Effective Size, m. m. Uniformity Coefficient. .38 1.9 .35 2.1 .33 2.1 .30 2.5 .27 2.8 ■ .24 1.5 .23 2.8 .24 3.0 .20 1.8 .35 9.0 .16 4.4 "EflFective size" designates the diameter of a grain of sand, 1(1 per cent., by weight of all the grains of the sample being smaller and 90 per cent, being larger than itself. " Uniformity coefficient " designates the ratio of the size of a grain which has 60 per cent, of all the grains finer than itself to the size which has 10 per cent, finer than itself; if all grains were of equal size this coefficient would be unity. All the above sands would require washing before being used in filter beds — the Cape May beach sands, to remove the salt ; the bar sands, to remove the fragments of bark and finely-divided organic matter ; and the bank sands, to remove the loam and clay mixed through them. The Cape May sand is the best, but most expensive sand. The Delaware bar sand can be secured at a reason- able price, and, with proper selection and washing, would prove well adapted for purposes of filtration in the slow filters. The rough Gloucester sand, with screening and washing, would be valuable both for filter sand and for making concrete. nil 1889 to 1898 inclusive. • Nut Earnikos, or EzcBas OF Gross Earhinos OVER Expenses. "i" Extensions. Total Expenditure. Total. Amount. Per Cent, of Earnings. $708,817 53 $605,668 57 $1,314,506 10 $927,493 75 41.37 712,497 37 280,866 92 993,364 29 1,387,673 41 68.30 781,227 83 749,066 21 1,530,294 04 970,408 69 38.82 814,832 89 658,124 42 1,372,457 31 1,261,998 71 47.91 ,121,555 91 1,471,834 90 2,593,390 81 80,884 43 3.02 ,677,081 03 1,235,775 01 2,912,866 04 —153,225 45 —5.65 1,509,902 97 387,322 23 1,897,225 20 932,631 97 32.97 1,311,338 57 514,272 32 1,825,610 89 1,053,522 37 36.58 1,354,642 90 310,510 31 1,665,153 21 1,306,204 31 43.96 1,360,220 19 135,778 65 1,495,996 84 1,569,669 02 51.22 iry supply and service mains ; meters of all kinds, large valves, stops and snses of purveyors' offices, and lead service pipes laid by the City from main to Expenditures for construction and repair shop include salaries of Superla- lachinists, blacksmiths and all employees at shop ; shop castings^ including br fire hydrants, small stops, stop-box frames and covers, and grate bars ; ine and miscellaneous castings, brass castings used in connection with the wrought iron and steel, other materials, and expenses of pattern shop. Expenditures for office include salaries of Chief and assistants, Chief Clerk ants, Chief Inspectors and nineteen inspectors, four draughtsmen, assistant in listribution, and all other office employees ; also office expenses, stationery and * supplies. Expenditures for extensions include new work for which special appropria- nade ; including, in general, new pumping stations or engines, boilers, reser- bkes, mains, etc. RBPORT OP THB I i TO run Select and Common Councils OP TRS / City ok Readinq, Pa., PBKTAINING TO THB PURIFICATION OP THE WATER SUPPLY BY KI]:vTRATION, — INCtODING - The previous reports of the Board on the subject, the report of Superintendent and Engineer Emil L. Nuebling^ and Consulting Engineer Allen Hazen. Also recent opinions upon Filtration by Prof. Erastus G. Smith, Prof, Wm. P. Mason and Rudolph Herring. :FEBI2."Cr-A.IJ»"Z- 28, 1898. RBPORT OF THE of • • er uinissioiiers To THE Select and Common Councils OF THE City ok Reading, Pa., PERTAINING TO THE PURIFICATION OF THE WATER SUPPLY BY FILTRATION, — INCLUDING — The previous reports of the Board on the subject, the report of Superintendent and Engineer Emil L,. Nuebling and Consulting Engineer Allen Hazen. Also recent opinions upon Filtration by Prof. Erastus G. Smith, Prof. Wm. P. Mason and Rudolph Herring. FEBK.TJ-A.K.ir 28, 18©8- READING, PA.: W. RosENTHAi,, City Printer, 710 Penn Street. OF THE S^oari. of ©y^afer (somrr|Ix^x«>Ionerx^ Relative to the Filtration of the Water Supply. To the HonoraUe the Select and Common Councils of the City of Reading, Pa. : Gentlemen : Following the preliminary report pertaining to the purifica- tion of the city water supply by filtration, submitted to your- honorable bodies under date of Oct. 1 1, 1897, we beg to present herewith. our conclusions upon the subject, together with an ex- tract from the Annual Report of theBoafd for 1805-96 relative- to filtration ; also the report made by our^ Engineer and Super- intendent," Mr. Nuebling, and Consulting Engineer, Mr. Allen.. Hazen, which report includes results of experiments and examinations made with the waters from the various sources, of supply ; also covei's in detail all technical questions pertain- ing to the problem. We here suggest to the reader, who wishes, to study the subject, that before reading the conclusions of the Board, he read first the extracts referred to, the preliminary report, and the report of the engineers. This will afford a clearer insight into the whole subject. The purification of the water supply has been a problem', which has frequently received the consideration of this Board; and until the appointment of j'our Special Committee on Filtra- tion, to urge us to action thereon, we have hesitated to make- any recommendations definitelj' favoring filtration, principally because of the great cost attending both the installation of filtration plants and the expense of operating them as applied to our water works system. Another question considered wasr Is the water supplied impure to a degree to serioiisly endanger the life and health of the citizens ? And another, will che ma- jority of our citizens approve of an expenditure of $300,000.00" for filtration in view of the fact that there is no constant serious- contamination and the waters nearly always clear. In our observation of cities filtering water, it was learned, that scarcely any have had water in its crude state of suck tiniformly good character as that of Reading, and that there =are many large cities using unfiltered water which is contami- nated to a vastly greater degree than any of the waters supplied our people. In recent years, however, the filtration of public water supplies has received greater consideration by water works officials, and while many cities are not filtering the water supply to-day, it is not because it is deemed unnecessary, but because the cost is most frequently a bar to the construc- tion of the necessary plant. Since the germ theory of diseases has attained such prorai- nence as an explanation of the cause of disease, and scientists have conclusively proven its transmission with water into the -human system, we can no longer ignore their claims that im- pure drinking waters are frequent and potent causes of sick- ness and death to the consumer ; and the water works official •of to-day must conclude that his most important duty lies in making every effort to furnish the consumer with water free from dangerous impurities. Nor is the bacterial quality only to be considered. Clear sparkling and palatable water will be •demanded by the people. It is, therefore, to filtration we must look to secure these re- . suits, and however great the cost, it is but proper to consider health and life paramount thereto. We believe that this senti- ment will prevail, and that the next teu years will witness filtration in use in nearly every progressive city in this country. Recognizing the fact that, with the constantly increasing use •of the Maidencreek water, our citizens are exposed to the dangers of typhoid fever (by reason of its large and uncontrollable drain- age area), and that the waters of the Antietam supply will be, by filtration, freed from the occasional objectionable taste and •odor from vegetable growths therein, we recommend that the entire supply be filtered as early as practicable, first the Maid- encreek supply, next the Antietam, the Bernhart and theEgel- man, and we further recommend that the system known as ■ Slow Sand Filtration be adopted as recommended by our Engi- neers, Mr. Nuebling and Mr. Hazen, in their report herewith. In determining upon a system, we have considered first, the patented system of the Penna. Sanitation Co., somewhat simi- lar to that in use for filtering sewage in this city. Second, the mechanical systems of rapid filtration with the use of coagu- lants. Third, Slo\y Sand. Filtration. Each of which is dis- ■cussed in detail in the report of the Engineers and their merits •and demerits fully set forth. The Penna. Sanitation Company submitted plans to filter the .Maidencreek water by providing for a steel or iron structure erected on the inside slope of the embankment and on the bot- tom of the Hampden reservoir, and for the Antietam on the- front of the wall or breast of the dam, designed somewhat similar to the sewage filtration beds built by that com.pany for this city. The representatives of this company have on several occa- sions during the past three 3'ears appeared before this Board to urge consideration of its system to filter the water supply. Up to the time of the appointment of your special Committee we declined to take up the subject with them, because at that time we considered the principles tipon which the system was. designed and constructed as wrong, and not calculated to in- sure a high efficiency in purification, and because its construc- tion was illy adapted to our system, and we doubted whether the steel structure would prove durable under the exposure to air and moisture to which it is subject. After the company had first submitted its plans to your special Committee, this- Board was asked to give them consideration, which we have done very fully and are thus forced to put upon record what before was our private judgment. It has been apparent that this company has some friends in this city who are disposed to favor its system. For this reason we have given the plans submitted careful investigation, so- that the claims made for it would be fully understood, and its shortcomings fully explained. Not only has our own judg- ment been passed, but that of our Engineer, Mr. Nuebling, and Consulting Engineer Mr. Hazen, whose explanations of the systera are fully made in their report, and we ask your honorable bodies to give them careful study. We have also- had other disinterested water works engineers to visit the sewage plant in this city to learn the principles of its construc- tion and operation, and their expressions we can assure you were not those of approval. We are also advised that the o£B- cials of several other cities in this State have been during the past year studying the system of this company, and we have yet to hear of its adoption by any of them. Director Thomp- son in a report on the filtration of the Philadelphia supply,, states he has visited nearly all large filter plants in this country,, (and we know he made an inspection of the Sanitation Co. plant in this city )aDd strongly recommends Slow SandFiltration. The "Engineering News," a journal devoted to the discus- sion of all important engineering problems before the people, in its issue of Jan. 27th, 1898, has published a detailed report upon the sewage plant in this city, and in an editorial upon the system attacks its efficiency and the high rate of filtration ■claimed. The aeration feature upon which the company places "^o much stress is shown to be of no great value, particularly ias a filter for water, and closes the article as follows : "In conclusion we are compelled to say that in our opinion the claims made for this plant are not proven ; that the rate of ■filtration is many times beyond the capacity of the plant to ■perform ; that the expense involved in elevating the upper bed is not warranted by results shown ; and that the City of Reading could have built for less money filter beds that would have done far more work than can ever be expected from these." With due deference to the opinion of the friends of the sys- tem of this company, we regret that their views cannot be sustained by our investigation thereof. We therefore -are obliged to reiterate our previous opinion that if the inter- ests of the city are to be regarded, this system should be con- sidered as an experiment, and no money risked thereon. Several systems of mechanical filtration have also been in- vestigated, principally small plants, and operated by private water companies. It is deemed advisable not to recommend the adoption of any of these systems brought to our atten- tion. They are all costly to maintain if a good bacterial effi- 'ciency is desired. All of the systems have expensive machinery to maintain and in time to replace. We regard the conclu- sions of Mr. Nuebling and Mr. Hazen, relative to their appli- cation to our system, as sufficient reasons for declining to re- 'Commend mechanical filtration. Slow Sand Filtration has been in continued use in this country and Europe for a period of over fifty years. Under all recorded tests, and under all conditions it has shown the highest efficiency. It has been proven the most durable in construction, and its efficiency most successful. Abundant testimony could be submitted on these points. Our observa- tions on this subject lead us to the conclusion that there has not yet been anything devised which will equal it, because its basis is founded upon natural principles ; it filters water as nature does through the earth, and all other systems are only attempts at imitation thereof and therefore make-shifts and largely experimental. The first cost of construction for Slow Sand Filtration is greater vinder ordinary conditions than the mechanical systems, or that of the Penna. Sanitation Co., but in cost of operation and durability this is more than made up. In construction, the filter beds under this system resemble -and have that permanency which is found in storage reservoirs; when once built thej' will require no expenditures for main- tainarice except the cost of scraping .and washing the sand ■from top of the beds. We have through correspondence and personal contact with able and experienced water works men, and experts on filtration, made numerous inquiries, and dis- •cussed this subject, and found nearly all agreed that slow sand filtration is to-day the only system which can be safely recom- mended. Director Thompson of Philadelphia in his recent report to 'Councils, on the water supply, under date of October yth, 1897, says : ' ' From all the information which I have been able to gather .from abroad and in this country, it appears that slow sand fil- tration has proven a success, and is no longer considered an ex- periment. It is the only system in successful practical use for •purifying the water supply of any of the great cities of the world. In several instances, as at Antwerp, for example, where it was originally intended to supplement sand filtration by other methods, sand filtration alone has proved sufiicient for the purpose, and the supplemental methods have been aban- doned. From all the research which I have made in the mat- •ter, I am of the opinion that the City of Philadelphia should adopt a method of slow sand filtration. Although this system is considered more costly in construction, the filter bed, if well designed and properly cared for, is almost indestructible, and involves practically no expense for repairs. The expense of operation is practically confined to the cleaning of the filters, and is much less than that attending any system of mechanical filtration." Jno. C. Trautwine, Chief Engineer of the Bureau of Water, in same report says : ' ' If Councils are determined to select a system without ex- .periment, let it, by all means, be the old-fashioned, slow, or so-called 'natural' sand filtration, which has, at least, demon- strated its usefulness by many years of successful use on an -enormous scale in lyondon and other European cities. ' ' We, therefore, have no hesitancy in saying to our people it is the best, and that if filtration is wanted, it should be adopted .and put in service without delay. In support of our decision .and for the information of our citizens studying this subject we have appended to this report the letters and recent opinions of the following eminent authorities, viz. : Prof. Erastus G. Smith, Prof, of Chemistry of Beliot College, Wisconsin. Prof. W. P. Mason, of the Polytechnic Institute, Troy, -N. Y. 8 Rudolph Herring, G. E., Cousulting Engineer, New York. These are taken by permission from the aforesaid report of Director Thompson on the Philadelphia water supply and filtra- tion, all of which have an important bearing on the conditions existing in our city. They afford valuable information for our citizens on the subject and careful study is suggested. As the Maidencreek and Antietam are first to be filtered, we give here comparative cost thereof ; the cost of the other sources can be seen in the statement of the Engineers. The cost of a Slow Filtration Gravity Plant for the Maiden- creek of 9,000,000 gallons daily capacity, and Antietam of 3,500,000 gallons, at the highest figure is $186,000 for the former and $50,500 for the latter, or a total for the two of $236,500, as against $268,700 for Penna. Sanitation system, and $169,700 for the mechanical or rapid system. The great cost of the Sanitation Co. system here is because it is, as before stated, not well adapted to the existing conditions of these sources, more fully referred to in our Engineers' report. The annual cost of operating the Slow Sand Filters for some years to come, exclusive of interest charges, for the Maiden- creek is $2,500.00, and of the Antietam $2,625.00, or a total of $5,125.00, as against $5,345.00 for the Sanitation Co.'s system and $9,534.00 for the Mechanical system. There is thus shown that for a system of Slow Sand Filtratian for these two sources the cost would be $32,200.00 less than for that of the Penna. Sanitation Co., and $66,800.00 greater than the Mechanical system; and operating $220.00 less than the Sanitation Co., and $4,409.00 less than the Mechanical system If the plan to- pump to the sand filters at Maidencreek be adopted, the cost of construction for Maidencreek and Antietam plants would be $77,700.00 less than that of the Sanitation Co. and $21,300.00 greater than a Mechanical system, and in operating $1,030.00- greater than the Sanitation Co., and $3,159.00 less than the Mechanical system. In these estimates on operation nothing has been figured for depreciation by corrosion, &c., of the steel structure of the Sanitation Company filters, painting and other care thereof. Having now disposed of the problem with regard to the kind of filter, the means to raise money to install the system is the next important question. The total cost to construct slow sand filters for the entire system is as follows : Maidencreek gravity filter fi86,ooo 00 Antietam " " 50,500 00 Bernhart " " 44,90006 Egelman " " 6,900 oc^ Total for all $288,300 60 and the maintenance and operation, including interest charges,, on the same would be as follows : Maidencreek supply $11,800 oq Antietam " 5,150 oq Bernhart " 3.870 00' Egelman " 570 op Total $21,390 00 In our judgment the cost of construction for the Maidencreek and Antietam filters can be provided for only by a loan, which would have to be authorized by the people, and as the people want to have some saj' on this question of filtration, let them decide it bj' voting on a loan of $225,000.00, which amount with the premium received on sale of bonds will install a system of slow sand filtration for these two sources of suppli^ The cost of filtering for the Bernhart and Egelman supplies can be provided for from the surplus revenue after the year 1900. The cost of maintenance, operation and interest will have to be made up, in part, by increasing the annual water rates which will mean the restoration of the fifty cents taken off the hydrant charge a few years ago. Any increase of the city's indebtedness is usually looked upon by many with disfavor, and this Board feels the responsibility it has assumed in making- a recommendation for a public improvement which involves such an increase of the debt. Pure and palatable drinking wa- ter, however, is something in which every crdzen must have some concern, and he should know whether it is of sufficient importance to him to lend his aid in providing means to secure it. Relative to the debt of the Water Department, we may ex- plain that in January 1900 and 1902, $127,500 of the debt matures ; provision has been made to pay this from the reve- nues of the Department ; this would leave the net debt $405,000. In 1910 another loan of $75,000 matures, which is also provided for from the revenues. At that time the net debt will be less than $325,000, and not due until 1920. If the proposed loan of $225,000 is issued, by 1920 $165,000 thereof will have been provided for in the Sinking Fund. The balance of $60,000, added to the $325,000, will leave the net debt in 1920 less than lO ^^8p,ooo, This would be a very light debt on a water vvorks Dl^pt, which will, at that time, be worth over $2,000,000. I^et It bp remembered that the water works system has cost to this (4ate about $1,700,000, and this present debt of $405,000 is y^ry small when we consider that the works was bought but jthirty-three years ago, at a cost of $300,000, and bonds issued for that amount at the time of purchase. This sum of $1,700,- oob represents the original cost of $300,000, the permanent im- provements and extensions to the system since its purchase. Interest, maintenance and operation are not included. If the M^t debt, is but $405,000 and the cost $1,700,000, the Departr ment from its own revenues will have expended for these im- provements about $1,300,000 in thirty-five years, or an average of about $37,000 per annum. To add $225,000 to the bonded debt for filtration, will in our opinion not add any burden to the water consumers, and if they want better water than we have to-day, let them vote for a loan to secure filtration, or make no further complaints about bad tasting or impure water. •■ Cities of the size and progressive spirit of Reading are con- titiually spending large amounts for improvements to their •Water supply, and we can point to many spending relatively much greater sums than Reading. In addition to the places already having filter plants, Albany, N. Y. , has to-day a filtration plant under construction which i^ to cost nearly half a million dollars, while Pittsburg, Pa. , Phi- ladelphia, Pa., St. lyouis, Cincinnati and many other cities have the matter under consideration. Reading will accomplish much if it can secure a filtration system without increasing the debt more than we recommend, viz : $225,000. The argument may be made by some not familiar with the requirements of our water works system, that the total cost of filtration could be paid from the surplus revenues of the de- partment. A careful inquiry into present and prospective needs, forces us to the conclusion to recommend the Loan, for iio large surplus can be maintained unless the increase in water rates be made considerably greater than before mentioned. In 1 898-' 99 nearly $30,000 in each year will have to be placed in the Sinking Fund to meet a bonded debt of $57,500 maturing January 1900, which is unprovided for in the present Sinking Fund ; $20,000 for Section "2" of the Maidencreek Pumping Main and $8,000 for completion of Sedimentation Basins at Antietam. In 1899 $30,000 to the Sinking Fund as above, $20,000 to the Maidencreek. In 1900 the 20-iuch and 16-inch trunk line will have to be laid beginning at Centre Ave. and Amity streets, thence along II 'fthe route laid out on the distribution plan adopted by the Board to Fifth and Pine streets, at a cost of about $150,000.00. The issue of a loan for this trunk line has been discussed by the Board during the past year, but since your honorable bodies have urged filtration by the appointment of a special Commit- ■tee to confer with this Board, it has been decided that it would be the wisest policy to give filtration precedence and ask for a loan for that purpose, and defer the laying of the trunk line for two j-ears and then endeavor to pay for it with the surplus "revenue over a period of five years. If filtration should be put into effect the increased cost result- ing therefrom will add nearly $22,000 to the iriaintenance and interest account, which, with the ordinary requirements always to be provided for, will, we are convinced, prevent any large sum from the annual revenues being used for the installation -of filtering plants at all the sources of supply. In conclusion we wish to say that we have endeavored to -present the subject to you as fully as possible, and in making up our conclusions have had in mind solely the best interest of 'Our city. We beg to add our expression of approval of the efforts and -ability shown by our Superintendent, Mr. Nuebling, in his treatment of this subject as given in the joint reports of him- self and Consulting Engineer Mr. Hazen, and further to add, that in Mr. Hazen the city has had the experience and know- ledge of one whose opinions and recommendations will bear the scrutinj' and command the respect of the best authority on ■filtration. In order to bring the question to an issue without further 'delay, we will present for 3'our consideration, immediately after the organization of the new Councils, the legislation nec- "cessary to provide for the loan, and authority to proceed with the work. Trusting that our recommendation may meet your favor, we •are Very respectfully, Geo. H. Felix, M. Harbster, Frank A. Tyson, F. P. Heller, Board of Water Commissioners. 3leading, Pa., Feb. 28th, 1898. 12 From Slst Annual Report of the Board of Water- Commissioners, 189S-'96. ANTIETAM LAKE SUPPLY. This source of supply remains in about the same conditiom as last reported. The water was in use during the year except at the time of the drought. The low condition of the streams- and having to use the water which had been stored for some time, were cause for the musty or woody taste and smell no- ticed during this time. The reservoir was emptied and fresh water dammed to refill it, after which it was again put to use. Of the lands adjacent to the reservoir and contaminating the water, the Hinnershitz Mill property ; the Daniel Faust prop- erty ; the Chas. Wentzel and Henry Schildt properties have been secured. The appropriation for 1895 for this purpose not being sufiBcient to pay all and receive title at once, they were- secured and paid for in part, the title to be delivered in April, 1896, when the annual appropriation for 1896 would be availa- ble ; possession to be given on July ist, 1896. Until that time, therefore, no work to improve the condition of these properties- and prevent the pollution could be commenced. Plans will be immediately prepared for the improvements calulated to free this water from dangerous contamination, and put into effect as early as possible, For the cost of each of these properties see statement in Superintendent's report. Incident to the purchase of these lands, and before they were bought, there arose in the minds of your Commissioners the question of the filtration of this supply, and its results as com- pared with results to be attained by the ownership of portion of the drainage area. The subject was given thorough consideration and investiga- tion. The facts to be considered in discussing it are these : ist — The source of the supply is from springs in a mountain- ous country, covering a total drainage area of 5.44 square miles, formed into several streams and collected in a storage reservoir of 101,000,000 gallons capacity, and piped by gravity 3 miles- to the city ; the character of the water, such as when free from the contamination from farm lands through which the streams- 13 ipass, makes it a most delightful, palatable and healthful drink- ing water, which with its high elevation and being capable of a "daily service of about 3j^ million gallons, proves it to be a ■very desirable source of supply. 2d — The contamination which this water receives is the •drainage, in times of heavy rainstorms and freshets, from some ■of the barn yards and farm dwellings, located near the reser- voir and along the streams, and causes a pollution of the wa- ter. This, and this only, affects the purity of the water with its resultant dangers to the health and life of our citizens. If we remove the cause of this danger (and prevention is always 1 better than cure), no better water can be had. 3d — ^That at a certain period of some years this water de- velopes a musty, wood}' taste and smell. This condition is found in many water supplies having their source in moun- tain springs and stored in quantity. While it is objection- ; able, it should cause no fear for the public health. It occurs at a time when the storage supply is being drawn upon, and is sometimes of very short duration, its disappearance being as , sudden as its coming, apparently due to some process of na- ture. Physicians and chemists agree that it has no deleterious •effect upon the human system. This condition therefore, should have no material bearing upon the question of the need ■of filtration. It is only periodical, is insignificant as affecting public health and its successful removal by filtration a very doubtful result. In our search after knowledge upon the practical results of filtration, it was learned that a supply of filtered water is not without its objections ; that filtration is not in all cases a suc- ■ cess, and that it is admitted by many Water Works Engineers - and Chemists, that the filtration of public water supplies is yet in its experimental stage and much yet to be learned to meet the conditions required to successfully make impure water ab- rSolutely pure and at the same time a palatable drinking water. Upon inquiry it was found that all or nearly all cities filtering ■water, were forced to do so because the character of the water • at their command was such as to be totally unfit at any and all times for domestic use ; that in many cases it was the only sup- ply obtainable, and filtration the only possible means to secure -wates fit to use. The cities located along the murky waters of the southern states ; those along the turbulent Missouri and Mississippi rivers and the polluted Ohio, and those depending Tipon the streams and rivers polluted with the drainage and .-^wage from the many manufacturing cities and towns of the New England State's, are forced to filtratron as- their only re- source. The situation of Reading is an entirely dnfferent one... ■ Looking to the filtration of our Antietam supply from a-", practical standpoint, we are convinced that the conditions per-- taining thereto are such that only mechanical filtration is feasi- ble and with this a pumping plant to maintain all the'advan-; tages which the reservoir now affords. The cost of such a!; plant will be not less than $6o,coo.oo, and, at least $5,000 an- nuall}' to operate it. The cost of the lands recently purchased^ has been $35,000.00, with $15,000.00 additional expended in- this manner and $2,500 annually for care of the lands; withj sanitairy inspection maintained over the few farms yet remain- ing on the water shed ; with proper and friendly efforts, ex- erted with the owners, nuisances would be abated, and all pollution would be prevented, and the water delivered to our. people in its normal condition, free from impiurities and always- palatable, and at less cost than filtration. It is admitted by some, of our expert chemists that the water from mountain streams,., free from dangerous contamination, is always more desirable.- than filtered water, and that the ownership or control of the.- water shed to prevent such contamination is far preferable ta any system of filtration. Another practical question to be considered was : If the An-^ tietam supply be filtered, why not the Bernliart? (The condi- tions are similar, though in not so marked a degree in the mat- ter of pollution.) And why not the Egelman and the Maiden- creek supplies? Should only one-third of the city be supplied, with filtered water and two-thirds not ? With the arrangement of our distribution system, either one of the different sources, at times must be used in the territory supplied by the other, which would mix the filtered water of the Antietam with the- unfiltered water from the other sources, thus for the time nul- lifying the filtration of the Antietam supply. The thought of a plan, for entire filtratian, suggests an ex- penditure of at least $250,000.00 ; increased fixed charges an- nually of $40,000.00, and increased water rents as a consequent' result. After a careful consideration of the entire subject, the con- clusion was reached, that the conditions affecting the Antietam: supply were such as to make any effort at filtration very doubt- ful of success, and that until the entire supply can be filtered,., any expense in this direction would be unjustifiable ; that the- ownership of such of the lands as most seriously pollute the- water, a systematic sanitary inspection maintained over all the- ^ater shed and the care of these lands under proper supervisiom 15 and direction, together with a proposed system of sedimentation and aeration, by the construction of, a series of small dams and weirs along the streams above the reservoir, will afford the cheapest and most permanent plan, and preserve the wat^r' in its natural state and bg free from any dangerous or objection- able contamination. Acting then on these conclusions, the properties hereinbe- fore mentioned have been purchased. The improvements will go on as fast as funds will allow, and no effort will be spared to eradicate the pollution of this supply. We wish to add that an important advantage in owning the land in question is that the many small springs scattfereS among the hills can be opened up and drained into the streams instead of the meadows and low-lands now soaking up the water, thereby increasing the flow of the streams, we believe, to the extent of a half a million gallons daily. i5 PRELIMINARY REPORT OF THE Board of Water Commissioners, RELATIVE TO THE Filtration of the Water Supply. Presented October ii, 1897. To the Honorable the Select and Common Councils of the City of Reading ■ Gentlemen : By resolution approved June i, 1896, your honorable bodies appointed a special Committee of six members of Councils "to act in conjunction with the Water Board to look into the mat- ter of filtration (of the water supply) and report to Councils as early as possible, giving probable cost of filtering that supply (Antietam), with a view of filtering our total supply if ad vis- able." Our Board received official notice of this action on July 7, 1896. The first joint meeting of the Committee and the Board was held on July 21, 1896. Meetings were subsequently held on February 23 and April 15, 1897. The special Committee expressed to the Board that it was the judgment of the members of Councils as well as many citizens, that the condition of our water supply was such as to require purification by some method of filtration, and, as the resolution states, Councils wanted some information as to the cost of filtration. During the discussion of the subject it was developed that the members of the special Committee had a very limited knowledge of the methods of filtering large supplies of water, and were not familiar with the several systems in use, other than that of the Pennsylvania Sanitary Sewerage Co. filtering sewage in this city. This being the case, the members of the Board who had given the subject considerable prior study, de- termined that if the filtration of our water supply was to be seriously considered by the special Committee and the Board, the Committee should be given the opportunity to in.spect the plants or systems in operation in other cities, to secure the information so necessary to an intelligent disposition of this Very important subject. 17 Your honorable bodies were then asked for authority for such inspection, which was granted by resolution approved June 30, 1897. Accordingly visits were made to Far Rockaway, L. I., Poughkeepsie, N. Y., and I^awrence, Mass., each operating a ■system constructed upon the prniciples of what is commonly ■designated " Natural Sand Filtration" (by some called Slow Sand Filtration), which consists of underdrained beds of sand built upon land, and the filtrate drained into pump wells or re- servoirs as required. The.se beds are usually designed to filter from two to four million gallons per acre of filtering area in 24 hours, which .slow rate has proved to effect the highest de- gree of purification. At Westerly, R. I., and I^ong Branch, N. J., the systems of the New York Filter Mfg. Co., the former operating their Open Gravity system and the latter their Pressure system of -Mechanical Filtration. At Atlantic Highlands, N. J., operating the Continental Fil- ter Co. Pressure system of Mechanical Filtration. At Elmira and Niagara Falls, N. Y., the Jewell system of 'Open Gravity Mechanical Filtration. At West Reading, Pa., operating the Warren .system of Mechanical Filtration. At Wilmington, Del., a specially designed system by the United States Filtering and Purifying Co. This is the only ■plant the company has in use, and might be termed Mechanical 'Upward Sand Filtration, using metallic iron as a coagulant, and has an aeration scheme as a special feature, and is designed to :filter somewhat less rapidly than the regular mechanical filters. At Reading, Pa. , the plant of the Pennsylvania Sanitary •Sewerage Company, filtering sewage. This system consists of ■elevated iron or steel structure, supporting beds of sand, the water .after passing through these beds dropping through air a ■distance of ten feet to a reservoir. This aeration feature dif- fers from that of any other system and is recommended by the •company as of special consideration. The company has no water filtration plant in use, but at the recommendation of your special Committee has placed drawings and .specifications be- fore this Board, showing the same principles applied to water filtration, except that the additional or lower bed used for sew- age is dispensed with. There are no mechanical devices used in its operation and its principles of construction are nearer that of " Natural Sand Filtration," although the rate claimed is much more rapid than usually adopted in the practice of ■"Natural Sand Filtration" con.struction. i8 In the foregoing we have given as concisely as possible alj the systems visited and about all the diiTerent kinds in use. All the systems termed "Mechanical" are so called because' the cleansing of the sand or filtering medium is done by mechanical devices operated by steam power. All of these use either alum, salts of iron or milk of lime as a coagulant, thus securing rapid sedimentation. This with the frequent washing^ of the filtering medium, made possible by the mechanical de- vices, secures a very rapid rate of filtration, and in some systems; satisfactory results as to purification. These systems are by- some called "Rapid Filtration." By reason of their ability to- filter rapidly, the area required for filtering a given quantity daily is very small as compared with that of the "Natural Sand Filtration Systems." In the "Natural Sand Filtration S3'stems" the rate of filtra- tion is kept low, and the water passed through thick beds of sand (in imitation of water passing through natural formations: of earth and sand), thus securing the highest degree of purifi- cation, at the expense of increased area. Neither aeration or coagulant is used to secure the purification of the water. The cost of these different systems might be recorded in this, communication, but as the statements would require a great deal of time and space they are dispensed with. The various conditions and problems affecting our supply, as compared with, that of other cities, prevents intelligent comparison with the cost of a system here. Such figures therefore would be unin- teresting and serve no purpose. We will reserve for a future: communication the question of cost for a system for our city. At this point we wish to say that this Board has not been oblivious to the needs of the city in the matter of the purifica- tion of its water supply. The various engineering problems, connected with the study of the several systems of water puri- fication have been continuously in view. Long before the appointment of your special Filtration Com- mittee the Board and its Engineer have had frequent consulta- tions upon the questions involved in the preparation of plans for the purification of our entire supply. The varied conditions, affecting our supply system, the several clas.ses of service to be cared for, the various results which may be secured from dif- ferent systems of filtration, and more particularly the important, bearing the efficiency of purification has upon the cost of con- struction and operation, make the subject one, which, if the- interests of the city are to be properly protected, cannot be hastily disposed of. By reference to the annual report of 19 1895-96 it will be found filtratiou was given serious consider- ation prior to the purchase of some of the lands in the drainage.- area of the Antietam supply. Since that time a careful study of the systems and experience of other cities has been made^ and the methods adopted by them to secure successful results- have been learned. All of which places the Board in a position to judge what can and should be done in the matter of the purification of Reading's water supply. What was learned on the recent visits of inspection by the Board and the special Committee, was but largely a repetition to the members of the Board. The knowledge secured was,, however, valuable to the Committee, and we are glad to say the keen interest shown by the members in the several systems, will aid them materially in the discussion of the subject when brought before your honorable bodies. The money expended', for this inspection was wisely given, for there can be no ques- tion that it will be of material benefit to the city. When in February, 1897, the Penna. Sanitation Co., at the- suggestion of your Committee, and without any request from, this Board, presented plans and specifications for filtering the- Maidencreek and Antietam supplies, the situation became such, as to make it evident to this Board, that the question of filtra- tion would have to be met in a business way, and proceeded, with in such a manner as to prevent the adoption of a system without a thorough knowledge of its efficiency, economy and. practicability as applied to our water works system. The con- clusion was reached that no sj'stem for the purification of the water supply should be put upon the city without knowing- what it would do, and in securing that knowledge the results, of efiiciency and practicability of the several systems must be passed upon by expert authority on filtration of public water- supplies, acting solely in the interests of the city, and not for- any particular system or company, and whatever system would, meet the standard of efiiciency and economy required by the- Board and its expert, would be recommended to your honor- able bodies. Having then assumed this position and having before us the plans of the system of the Penna. Sanitation Co. to be disposed of, we at once (February 15, 1897) referred the- said drawings to our Superintendent and Engineer, Mr. Nueb- ling, with instructions to prepare a report upon the whole sub- ject of filtration. At the same time it was agreed to appoint a consulting engineer of recognized ability to act with him. The- scope of action and authority given Mr. Nuebling, will be best explained in the following resolution passed by the Board at. the time, viz. : "Whereas, Councils have appointed a special Committee to confer with this Board relative to the filtration of the water supply of the city, and said Committee having appeared before the Board on two occasions, and recommend that some plan for filtering water be adopted and put into service and ask that the Board present a report to Councils upon the subject ; and "Whereas, The Penna. Sanitation Co., of Philadelphia, is the owner of a patented system of filtTation, and has, of its own volition, presented for our consideration plans and specifications for filtering the Maidencreek water at the Hampden reservoir, -and the Antietam supply at the Antietam Lake ; and "Whereas, The subject of filtration having been luider consideration by this Board for the past two years, and being one of such magnitude, and serious concern as to cost as well as successful results, it is deemed advisable that sufBcient time be taken to allow the question careful study and research as well as experiment ; it is therefore "Resolved, That this Board proceed to the consideration of the entire subject of filtration of the water supply by a thorough investigation of all the systems of filtration, with a view to securing the best attainable method in accordance with the most advanced thought and experience upon the subject, and that, preliminary to making a report thereon, the plans and specifications of the system submitted by the aforesaid company be referred to the Superintendent and Engineer for examination ;and report to this Board ; that he is especially instructed to re- port upon its practicabilitj' and efficiency as applied to our water system, examine and compare same upon all points with ■other systems ; that in connection with the Board he investi- gate other systems, and ascertain the experience of other cities, make experiments, collect such data upon the subject as in his judgment will bear upon the successful and economical filtra- tion of such parts or the whole of the water supply ; that he is authorized to visit other cities, at the expense of the depart- ment, whenever expedient to secure necessary information, and that he report to this Board from time to time the course of his procedure ; and further "Resolved, That at the proper time a consulting engineer of recognized ability on filtration of water supplies be employd to advise and consult with the Superintendent and Engineer on this subject." Accordingly on the 6th of March, 1897, we engaged as con- sulting engineer Allen Hazen, C. E., of N. Y., an authority of oiational reputation on the filtration of public water supplies 21 and water works practice generally. One of the first steps- taken was to ascertain what results could be secured , chemi- cally and bacterially, and as well at what rate the water of the several sources could be safely filtered, and from these results, calculate the cost per million gallons to filter our supply. For this purpose our Engineer, Mr. Nuebliug, planned and erected two experimental filters, each 3 feet in diameter ; one built upon the principles of natural sand filtration, the other upon the principles given in the drawings and specifications submitted by the Penna. Sanitation Co. In the month of April, 1897, we placed in charge of these filters Mr. F. S. Hollis, a competent chemist and bacteriologist, formerly in the employ of the Massachusetts State Board of Health, and now acting- biologist of the Boston Water Department. For a period of two months daily examinations of the waters from the several sources were made, under all the varied conditions of our sup- ply during that time. All such tests were made as were deemed pertinent to an intelligent study of the question of filtering our water supply. Our Engineer, Mr. Nuebling, is also well versed on the sub- ject, and with the aid of such talent as above, we feel satisfied that we are in good hands, and that nothing will be left un- done to secure entirely satisfactory results. Particular instruction was given our Engineer to design and submit plans and costs of systems of natural sand filtration and of mechanical filtration, in order that their cost and efiiciency may be studied in comparison with the patented system of the- Penna. Sanitation Co. The work is necessarily slow, and having to be done in con- nection with the other improvements in the hands of Mr. Nuebling, has been delayed longer than was our intention. As the work is now well in hand, we expect within the next two months to receive the report of the Engineers with com- plete drawings, e.?ti mates, &c. It is not our purpose at this time to make any argument for or against filtration, but as soon as the report of the Engineers, is received, we will present to your honorable bodies the sub- ject in suchi shape as to afford you the opportunity for thorough inquiry, together with such recommendations as in our judg- ment will be for the best interests of the city. When all the data is submitted to you, we hope to place it in such light as to- prepare you for an early decision as to whether we shall have- our water supply filtered or not, and in what manner it should, be done. 22 We submit herewith an ordinance to make an appropriation •of $1,500, to pay the expense necessary for the experiments .^nd services of consulting engineer, and recommend its pas- rjsage. Respectfully submitted, M. Harbster, F. P. HelIvER, Geo. H. Felix, Frank A. Tyson. Board of Watet Commissioners. JReading, Pa., Sept. 27th, 1897. 23 ON THE OF THE CITY OF READING, PA, EMII, L. NUEBLING, Superintendent and Engineer. AIvI^EN HAZEN, Consulting Engineer. 'To the Honorable the Boat d of Water Commissioners of the City of Reading, Pennsylvania : Gentlemen : In accordance with your request, we beg to submit the fol- lowing report upon the filtration of the water supply of the City of Reading, and the feasibility and cost of accomplishing the same by various methods: SOURCES OF SUPPLY. The City of Reading is supplied with water from five sources. The first source consists of certain springs in the neighborhood -of the Hampden reservoir, which yield a limited amount of water of very satisfactory quality. This supply does not re- •Kjuire filtration. The second or Egelman supply is from an impounding reser- voir, holding 2,500,000 gallons, upon a water-shed of 0.54 of a square mile. There are several houses upon this water-shed. The geological formation of this drainage area consists of about -35 psr cent, of Potsdam sandstone and 65 per cent, of gneiss. About one half of the drainage area is woodland. The quality ■of the water is ordinarily good, excepting that algae growths :at times give rise to unpleasant tastes and odors, and after heavy rains the water becomes muddy. The Egelman reser- voir supplies the High Service District. The third source of supply is from the Bernhart reservoir, ■with a capacity of 42,000,000 gallons, supplied by a drainage area of 2.56 square miles. There are about thirty houses on this water-shed. The geological formation of this drainage ^rea consists of 80 per cent, of Potsdam sandstone and 20 per 24 cent, of gneiss, and fully 30 per cent, of it is woodland. The water is considerably harder than the Egelman, and suffers oc- casionally from mud and algae growths. The water receives, and is made much harder by the water of, several springs, which are much harder than the surface sources. This excess, of hardness is apparently due to contact with subterranean de- posits of limestone. The capacity of the pipe line from this, reservoir to the city is 2,500,000 gallons per day, and the ele- vation is such that it just serves for the supply of that part of the city lying below a contour of 270, and known as the I,ow Service District. At dry seasons more water is required in. this district than can be supplied from the Bernhart water-shed, and the deficiency is at present made up with water let dowa from the Intermediate Service ; but it is the intention, after certain changes in the pipe line shall have been made, to sup- ply this deficiency and the natural increase in consumption in. this district with water pumped from Maideucreek to the Bern- hart reservoir direct. The fourth source of supply is the Antietam reservoir, hold- ing 101,000,000 gallons, supplied with water from a drainage area of 5.44 square miles. There are about eighty houses upoa this water-shed. Your honorable Board has within the last few years purchased those places upon the water-shed which were nearest to the reservoir, and which from their locations probably contributed in greatest measure to its pollution. The geological formation of this drainage area consists of 25 per cent, of Potsdam sandstone and 75 per cent, of gneiss. The water from this water-shed is quite soft. The water in this, reservoir is occasionally muddy, and has algae growths, which make it quite unpleasant in taste aud odor. The conditions of the water in the Antietam reservoir will be considerably im- proved by works now under construction, namely, a series of small dams along the streams supplying the reservoir, which dams are to be so constructed as to aerate the water as it flows, over them. The capacity of the present pipe line from the reservoir to the city is 3,500,000 gallons per day, and the ele- vation is such as to allow its use without pumping in the Inter- mediate Service. The fifth source of supply is Maidencreek. Water is pumped from this creek at a point where its drainage area is 210 square miles. There is no storage reservoir on the creek. The geo- logical formation of this water-shed consists of 70 per cent, of slate, 21 per cent, of limestone and 9 per cent, of other forma- tions, the larger part of the limestone being on the water-shed of Willow creek, a tributary entering Maidencreek about half 25 a mile above the pumping station. There are several small villages and institutions upon the water-shed of Maidencreek, some of which are sewered, and the sewage is discharged un- treated into the creek or its tributaries.* The water of Maiden- creek is considerabl}^ harder than that from the Antietam and Egelman reservoirs. It is not subject to the disagreeable tastes and odors resulting from algae growths. It is muddy after heavy rains, and the discharge of sewage into the streams renders the use of this water in its raw state more or less in- jurious to health, and objectionable. REQUIREMENTS OF FII^TRATION. The problems presented are : (ist), the removal of the mud from all of the waters after heavy raius ; (2d), the removal of the tastes and odors resulting from algae growths from the wa- ters of the Antietam, Bernhart and Egelman reservoirs ; and (3), the removal of the bacteria which are or may be injurious to health, discharged into the Maidencreek by the above men- tioned sewers or other sources of pollution, and of similar germs from the reservoir waters, in case such germs are present or should be introduced from the inhabitauts living upon the respective drainage areas. The disease, which is believed to be most frequently caused by polluted water supplies, and the presence of which in a city is most characteristic of such pollution, is typhoid fever. Ty- phoid fever is caused in other ways than by polluted water supplies, but its continued and excessive presence in a large city, unless otherwise explained, is an indication of such pollu- tion, and in case of known pollution, can be taken as an ap- proximate index of the damage resulting from it, being far more reliable in this respect than other diseases also caused by puUuted water, but whose occurence is more dependent upon other causes not connected with the water supply. The following table shows the number of deaths from ty- phoid fever in Reading for the last twenty years, with the estimated population, and the death rates from this cause for each 10,000 living. These numbers have been compiled from the printed records of the City Board of Health by adding to- gether the deaths reported for each month of the year, and deducting cases of burials where the death did not occur in the *See Appendix B. 26 city. The figures thus differ from the 3-early totals in the pub- lished reports : DEATHS FROM TYPHOID FEVER, READING, PA., 1877 TO 1896. Deaths from Estimated Deaths from Typhoid fever for Year. Population. Typhoid fever, each 10,000 living. 1877 39.650 10 2.52 1878 40.780 13 3-19 1879 42,000 2 0.48 1880 43,278 U. S. Census... 10 2.31 1881 44.540 18 4.04 1882 45.900 16 3.49 1883 47.300 8 1.69 1884. 48,800 13 2.66 1885 50.330 14 2.78 1886 51.950 22 4.23 Average for ten years, 1877 to 1886, inclusive, 2.74 1887 53.550 19 3-55 1888 55.200 24 4.35 1889 56,880 27 4.75 1890 58,661 U. S. Census... 32 5.46 1891 60,400 29 4.80 1892 62,260 28 4.50 1893 64,200 26 4.05 1894 66,200 30 4.53 1895 68,300 28 4.10 1896 70,400 36 5. II Average for ten years, 1887 to 1896, inclusive, 4.52 These figures show that the tj'phoid fever death rate in Reading has been slowly increasing, and, although not yet ex- cessively high, is considerably higher than the rate in most American cities having thoroughly good water supplies, and it appears probable to us that the increase in the rate has been connected with the increased use of water from Maidencreek, which is more directly polluted than the other sources of supply. The algae growths in the reservoir supplies often give rise to very offensive tastes and odors, which are in themselves ob- jectionable, and which should be removed. Water charged with such growths may cause diarrhoea or other disturbances in the system of certain people, particularly those not accus- tomed to the water ; but as far as known, no serious disease is caused by the presence of these organisms in water. The algae are vegetable growths, complex in their nature, and is -EinmiMmmL mjvK 27 ••entirely diiferent from the bacteria whicTi cause typhoid fe- -ver, cholera and other zymotic diseases. A comprehensive Teport upon this subject was presented in the annual report of ~the Reading City Board of Health for the year 1880. METHODS OF FILTRATION CONSIDERED. We were particularly instructed by your honorable Board to ■ examine and report upon a method of purification submitted Jor consideration by the Pennsylvania Sanitation Company, of Philadelphia, Pennsylvania, and which was accompanied by ■proposals to your honorable Board for the erection of plants for the filtration of the Maidencreek and Antietam supplies, under -dates of March 2d and March 15th, 1897. We have also considered the method commonly known as Slow Sand Filtration, which is extensively used in purifying ■similar waters in Europe and in this countrj' ; and the method of rapid filtration with the use of coagulants, known as Mechani- •cal Filtration, which is extensively employed in the United States. EXPERIMENTS AND ANALYSES. In order to arrive at a better understanding of the principles ■involved in the method of the Pennsylvania Sanitation Com- pany, and also to determine the results which could be obtained from it in comparison with those obtained from simple filtra- tion, experimental filters were constructed, which were put in operation in April, 1897. T^^^ services of Mr. F. S. HoUis, Biologist, formerly and at present in the employ of the City of Boston, were secured to make microscopical and bacterial •examinations of the waters applied to, and the effluents from the filters, and also to examine the waters from the various :sources of supply of the city. Chemical analyses of the various waters were also made by Mr. H. W. Clark, of the Ivawrence Experiment Station of the Massachusetts State Board of Health. The results of these experiments and examinations are given in detail in an appendix to this report. The filters consisted, briefly, of iron cylinders 2.95 feet in diameter, with areas of 6.835 square feet each. One of these cylinders was provided with a grating- made in accordance with the plans of the Pennsylvania Sanitation Company, and upon this grating were placed eight inches of gravel of assorted sizes. Tanging from the coarsest gravel at the bottom to fine gravel at the top, and above this, two feet of filter sand having an effective size of o. 24 of a millimeter and a uniformity coefii- cient of 1.8, which sand is believed to be essentially the same 2'8' as that tised bj' the Pennsylvania Sanitation Company in its; sewage disposal plant at Reading. Above the sand a device- was arranged for distributing water over the surface according, to the design of the Pennsylvania Sanitation Co. and an appa- ratus constructed for delivering water to the same at exactly- the determined rate, which rate was in proportion to the quantity of water proposed to be filtered by the said Company in the above mentioned proposals. The effluent dropped from, this grating in a manner peculiar to the plans of the Pennsyl- vania Sanitation Company, and was collected in a galvanized iron tank ten feet below, aBd overflowed from the same through, a meter, which recorded the quantity of water filtered. The filter and appurtenances were in all respects constructed and operated as closely as possible in accordance with the designs- submitted to your honorable Board by the Pennsylvania Sani- tation Company. The filter was located in one of the build- ings of the Water Department ; but as special precaution.'? were taken to protect it from dust and other impurities which might not occur in open air, and as. the windows were kept open, it is. not believed that its location affected in any substantial man- ner the results obtained. The second filter was provided with a water tight bottom, and the effluent was collected and taken through a device for indicating the rate of filtration, and afterwards through a meter which recorded the whole quantity of water filtered. Upon the bottom of the tank were placed six inches of assorted sizes. of gravel, and four feet of sand above. The sand and gravel were exactly like those used in the other filter. An excess of water was supplied to a small cup. on the side of the filter, the excess overflowing, and as much as was required passing through a small hole to the filter, maintaining a constant leveL The rate of filtration was regulated entirely at the outlet. This, filter was operated continuously, and except at times of scrap- ing, there was no chance for the admission of air to the filter- ing material, except the air carried in solution by the applied water. Water was applied to both filters- from the mains, and for about a month the water thus supplied was drawn from the Antietam reservoir. Afterward the mains were changed so- that the apphed water was from Maidencreek. The water from Maidencreek, however, was first pumped into the Hamp- den reservoir, and thence passed through the pipes to the filters, and the water as applied to the filters differed materially from the water pumped from the creek, owing to the sedimen- tation and other changes occurring in the reservoir and pipes. JA'SUPPLV PIPE _J_J, -sx-ps^mmmL tank OF=- ^unw Sand f lun^TinN. 29 The filter constructed on the plans of tlie Pennsylvania Sanitation Company was operated at a nominal rate of 12,- .500,000 gallons per acre daily ; but owing to imperfections in ithe measuring apparatus, the refusal of the filter to take the full quantity of water at times, and other causes, the average yield of the filter when in actual operation was 11,500,000 gal- lons per acre daily, and taking into account the times when it was out of use for the purpose of being cleaned, the yield was ^somewhat less. The sand filter was operated at a nominal rate at first of 3, 500,, 000 gallons and afterwards at 4,ooo,,noo gallons per acre daily, but the actual rate was slightly greater, averag- ing 4.03 million gallons per acre daily for the time it was in ^actual operation The water supplied to these filters contained at times some of the organisms giving rise to the disagreeable tastes and odors, and itself had such tastes and odors. The efQuents from both filters were substantially free from both the organisms and the odors, and were thus satifactory in this respect.* The water applied to the filters was not at any time seriously •discolored by mud, and the filters were not thus subject to as severe tests in this respect as would occur in practice. The numbers of bacteria also were not excessive, and further, the period covered by the experiments was not long enough to allow the best bacterial results to be secured. It seldom happens that a new filter gets into condition to give its highest bacterial efficiency in a shorter period than two or three months, while the whole period covered by the daily bacterial examinations •of these filters was less than two months from the time when they were first put into operation. The results in this respect ^re not as conclusive as could be desired. A special experiment was carried out on June ist, when a culture of a special kind of bacteria was mixed with the water applied to each of the filters. This kind of bacteria, called Bacillus Prodigeosus, is red in colrar, and can be readily dis- tinguished and counted ,; and as it has been found that it is removed by filtration in substantially the same proportion and under the same circumstances as ordinary sewage bacteria, and as it does not occur in the water or efluent except as applied ;and as actually passing through the filter, it forms a ready tneans of showing the bacterial ef&ciency of filters. At the time of the test both filters had been running several weeks .-and were in relatively good condition. The number of germs «of this species added amounted to 14,000 per cubic centimeter *See Appendix "A."' 30 of appKed water for a period of nine hours. In the effluentt from the sand filter a single colony of Bacillus Prodigeosus was. found in one sample only. In the effluent from the filter con- structed on the plans of the Pennsylvania Sanitation Company,, colonies were found in all but one of eleven samples takeni within twenty- four hours after its application, the aggregate- number of colonies found being 58. DESCRIPTION OF PROJECTS AND ESTIMATES OF COST. We have prepared outline plans and estimates of cost of fil- tration plants for filtering the different supplies by the above mentioned systems. In some respects these projects have not. been very fully worked up, and if adopted would probably re- quire to be modified somewhat before being put in final form. For the present we have only aimed to assure ourselves of the practicability of the various schemes, and to secure sufficient, data to allow reasonably close estimates of the cost to be made. Sand Filters : Maidencreek. — Two projects have been con- sidered. One provides for the construction of filters upon ground somewhat higher than the pumping station, to which, water would be pumped and afterwards returned to the. present pumping engines, which would pump it as at present. The project provides for 144,000 square feet of effective filter- ing area, and for open filters, with filter sand four feet in depth,, and for devices for regulating the rate of flow, etc., necessary to give the best results. A clear water reservoir is also pro- vided to allow the pumps to be operated steadily and without, trouble. The cost of filters by this project, including additional, pumping machinery, land damages, connections, and all inci- dental expenses connected therewith, is estimated at $140,500. The alternate scheme provides for placing the filters at so- low an elevation that they can be flooded from the creek, by gravity, and protected • against floods by high embank- ments. This project avoids the second pumping, and no ad- ditions would be required to your present pump house. The: estimated cost of construction is, however, materially greater, and including connections, land damages, and all incidental' expenses, amounts to $186,000. This total is less exact than: could be desired, owing to uncertainty as to the amount of rock excavation to be encountered, and further and more detailed' examinations by surveys and borings should be made of the: proposed site if this project is seriously considered. This pro- ject has the advantage of taking the water from Maidencreek: above Willow Creek, where the water is softer and otherwise.- of better quality. 31 Sand Filters : Aniietam. — This project provides for the construction of three filters having a combined effective filtering area of 45,738 square feet. The filters are placed in such a position that they can be flooded by gravity from the present reservoir, but it will be necessary to draw the effluent out at a lower elevation than the present hydraulic grade of the line. This would naturally result in decreased water pres- sure in the city. To compensate for this loss of head, provi- sion is made for a new 24-inch pipe line from the filters to Nineteenth street and Perkiomen Avenue. The present pipe line for a part of the distance is only 16 inches in diameter and is badly tuberculated. The friction of the water and the loss of head in the proposed line will be very much less than the friction in the present line, the difference being fully equal to the head lost at the filters, so that the water will be delivered in the city at fully as great a pressure as at present when the pipe is carrying 3,500,000 gallons per day. The estimated cost of these filters, including a new pipe line, land damages, and all incidental expenses connected therewith, is $50,500. Sand Filters : Betyihart. — The project for the filtration of the Bernhart water is very similar to that proposed for the Antietam. Three filters are contemplated with a combined effective filtering area of 32,670 square feet. The filters would be located below the reservoir and water would flow to them from it. To maintain the pressure in the water mains, an ad- ditional line of 20-inch pipe is provided, which will take the water from the outlet of the filters and deliver it to the city at somewhat greater pressure than will the present smaller lines taking water from the reservoir direct, and when delivering at a rate of 2,500,000 gallons of water daily. The estimated cost of filters by this project, including the new pipe line, land damages, connections and all expenses incidental thereto, is $44,900. An alternate project has been suggested for the filtration of this supply, which would involve pumping the water to filters at a greater elevation, from which it would flew through the present pipe lines to the city. This project would give a somewhat better pressure and the first cost would be a little less, but the operating expenses of an additional pumping station would very much more than off- set the saving in first cost, and this project has not been further considered. Sand Filters : Egelman. — This plan provides for the con- struction of two filters with a combined effective filtering area 32 of 4,360 square feet, immediatel}' below the reservoir, the out- let of the filters being connected with the present supply main. An automatic device would control the height of water on the filters, drawing water from the reservoir as required. This plant would be extremely simple in its construction and opera- tion. As the present watchman could take all necessary care of the filters, the only expense for the operation would be re- placing a small quantity of sand each year. The cost of the filters, including land damages, etc., is estimated at $6,900. Pennsylvania Sanitation Company's Project : Maiden- creek. — The basis of this estimate is a proposal submitted by the Pennsylvania Sanitation Company to your honorable Board, under date of March 2, 1897. The project provides for the location of filters over the Hampden reservoir on an elevated steel structure. The water would be pumped to these filters, and would fall from them into the reservoir. As the above mentioned proposition provides for the construction of a plant smaller than we deem advisable for this supply, we have in- creased the number of sections of filters from six to ten and have increased the price in the same ratio. The ten beds have an effective filtering area of 35,1 12 square feet. At pre.sent the Maidencreek water is pumped to the Hampden reservoir, but in the plans adopted for the developments of the city supply, it is the intention to pump a part of this water directly into the distribution system without first going to the reservoir. We have therefore included in our estimate such additional piping as will be necessary with the increased supply to carry the whole of the Maidencreek water to the Hampden reservoir. We have not taken up the question as to whether or not there will be damages to the reservoir by the location of such filters over it, nor the question as to whether or not suitable provi- sions have been made for foundations, nor whether the struc- tures proposed were in all respects strong and substantial enough for the purposes intended. The cost of the filters, as given in the above mentioned proposal, amounts to $3.42 per square foot of effective filtering area, or $120,000 for the area provided, and with the additional piping required, and all other expenses incidental thereto, is estimated at $189,500. Pennsylvania Sanitation Company's Project : Antie- tam Supply. — The basis of this estimate is a proposal made by the Pennsylvania Sanitation Company, under date of March 15, 1S97. The filters are to be located upon an elevated steel structure immediately below the Antietam dam, and to have an effective filtering area of 14,045 square feet. As the efflu- ent would be delivered at a n;uch lower level than the present 33 (hydraulic grade in the pipe, it would be necessary in order to maintain the same pressure to lay additional pipe to decrease the friction. In the project for sand iilters this reduction of friction is effected in the lower part of the line where the fall is rapid and the present pipe line smaller. In this case this is impracticable, as the pipe line below the filters for a long dis- tance has a very flat grade, and increasing the size of the pipe at the lower end of the line would not increase the delivery or pressure, which would be limited by the possible discharge through the flat section of pipe under the head available at its upper end. The only way to secure the necessary discharge and pressure, is to lay an additional line and thus still further reduce the friction in the flat part of the line. The cost of the filters as stated in the above mentioned proposal, is $49,000, or $3.49 per square foot, and with the additional pipe and con- nections, land damages, and all incidental expenses, is esti- mated at $79,200. Pennsylvania Sanitation Company's Method : Bern- Jiart and Egelman. — No estimates have been made of the cost 'of filtering these supplies by this method, as the difficulties 'existing at Antietam exist in even greater measure in connec- tion with tlicse supplies. The cost of the Pennsylvania Sani- tation Company's project for the Antietam supply is consider- ably greater than that of either sand or mechanical filtration, and, as will be developed below, there is no reason for prefer- ring it. Since it is obvious that the conditions at Bernhart and Egelman would not be in any respect more favorable for the installation of this system, we have not thought it necessary to make estimates for it. Mechanical Filters : Maidencreek. — This project provides for the construction of thirty-two 12-foot mechanical filters, or their equivalent in filters in other diameters, having a combined effective filtering area of 3,680 sq. ft. Water would be pumped from the present pump well to the filters, and from the filters would drain to a pure water reservoir, from which it would^ be pumped by the present pumps and through the present mains. A suitable house is required for covering the filters, and new pumps, boilers, &c. , will be required. The total cost of filters, including connections, pumps, house, land damages, and all expenses incidental thereto, is estimated at $128,000. Mechanical Filters: Afiiietam. — "^h& location of me- chanical filters for the Antietam supply would depend upon the type of filters employed. Mechanical filters can be constructed to be operated with as little loss of head as sand fiUers, and if such a type of mechanical filter should be employed, filters 34 could be located at the sites selected for sand filters, and the- cost of a mechanical filtration plant, having a combined effec- tive filtering area of 1,380 square feet, with power for stirrings pumps for wash-water, changes in piping and connections^ house, small clear water well, land damages and all incidentat expenses, is estimated at $41,700. If, on the other hand, mechanical filters should be used, re- quiring a greater loss of head, such as most of the filters now- in use require, a much greater change in the piping system, would be required, and the cost would be considerably in- creased. Mechanical Filters : Bernhart. — The project provides, for nine 12-foot filters, having a combined effective filtering area of 1,030 square feet, located below the reservoir. In case filters should be used having only a slight loss of head, the- efiluent would not require to be pumped, but the pipe line- would require to be increased in size in tlie same way as men- tioned in connection with the sand filter plant for Bernhart. A. small boiler and pump would be required for pumping wash-- water and power for stirring, also a small pure water reservoir. The total cost, including land damages, is estimated at $44,000. If mechanical filters of the ordinary type, requiring a greater loss of head, should be employed, and the pipe line should be left as at present, and a pumping plant installed to pump the- efHuent into the mains, the estimated cost, including land damages and all incidental expenses, is $35,500, but the operat- ing expenses in this case would be considerably greater. Mechanical Filters : Egelman. — This project include.s, two weclianical filters with a combined effective filtering area of 226 square feet, placed below the reservoir and discharging- their effluent directly into the main, with a .suitable house, a small boiler, engine, and pump for wash-water and stirring.. The total cost of this plant is estimated at $8,300. OPERATING EXPENSES. In estimating the operating expenses, it has been assumed that the quantity of water from Maidencreek for some years ta come will average 1,000 million gallons per annum ; that 1,050. million gallons will be used from the Antietam reservoir ; 65a million gallons from the Bernhart reservoir, and 75 million gal- lons from the Egelman reservoir. The care of sand filters has been estimated at $2.50 per million gallons of water filtered, which seems an ample allowance in view of the comparative clearness of your waters and the long periods which will occur between .scrapings, tending to low operating expenses. 35 The care of the Pennsylvania Sanitation Company's filters;- has been estimated at the same figure, it having been found by calculations from our experiments that the area of filter surface- to be cleaned in the course of a year would be slightly greater for the Pennsylvania Sanitation Company's apparatus than for sand filters, the cost of cleaning per unit of area being sub- stantially the same. The rate of filtration was three times as ■ great with the Pennsylvania Sanitation Company's filter, but the filter required to be scraped three to four times as often. No allowance has been made for painting and other care of the- steel. In mechanical filters one-half a grain of alum to the gallon has been allowed for the filtration of the reservoir supplies, and one grain to the gallon for the Maidencreek supply. Estimates, have been made in each case of the additional cost of pumping where required, and of the wages of the men required for the- operation of the mechanical filters, and of the wash-water re- quired, which has been estimated at ten dollars per million gal- lons. In the case of the Maidencreek supply this may be taken as representing the actual cost of the water at the works pumped and filtered, but is probably less than the actual costs - for the reservoir supplies. The operating expenses estimated in this v^'ay are included below in the tabular summary of es- timates. SUMMARY OF ESTIMATES. Below is given a summary of the areas and capacities of the- various proposed filters mentioned above, and their estimated, costs, and the charges for interest and sinking funds, reckoned thereon at five per cent. , and also the operating expenses when, filtering the annual quantities of water mentioned, and the- sum of the operating expenses and five per cent, of the first; costs, which sums taken together represent approximately- what the annual costs would be to the city for the filtration of: the water by the various plans. 36 Maidencreek. Sand filters by gravity JMaidencreek. — Sand filters by -Sand pumping. JVntietam.-Sand filters .. Bernhart filters . . Egelman. -Sand filters Maidencreek. Penna. San Co.'s method.. Antietam. — Penna. San Co. 's method. Maidencreek. — Mechan. filters „Antietam. — Me- chan. filters... Bernhart. — Me- chan. filters, pumping Bernhart. — Me- chan. filters, gravity Egelman. — Me- chan. filters... OJ V 144,000 144,000 45,738 32,670 4,360 35,112 14,045 3,680 1,380 1,030 1,030 226 9.0 9.0 3-5 2.5 0.5 9.0 3-5 9.0 3-5 2.5 2.5 0.5 c t- 5 1,000 1,000 1,050 650 75 1,000 1,050 1,000 1,050 650 650 75 h q u ? to s & ^ M 14-. CJ 3 1-1 a 1 U a e a 10 186,000 $9,300 $2,500 140,500 7,025 3,750 50,500 2,525 2,625 44,900 2,245 1,625 6,900 345 225 189,500 9,475 2,720 79,200 3,960 2,625 128,000 6,400 4,784 41,700 2,085 4,750 35,500 1,775 4,156 44,000 2,200 3,656 8,300 4>5 1.317 u u ap. O in $11,800 10,775 5,150 3,870 570 12,195 6,585 11,184 6,835 5,931 5,856 1,732 PRINCIPLES OF PURIFICATION INVOLVED IN THE DIFFERENT METHODS. Sand Filiation. — The saud filters in the above mentioned ••estimates are the saud filters commonly employed in Europe ■and in this country, provided with most approved regulating appliances. The general form and construction of these filters is shown by the accompanying drawing. They consist of open basins with water-tight bottoms surrounded by masonry walls. Underdrains are placed at the bottom, surrounded by gravel ..and over the gravel four feet of filter sand are placed. The "water is taken over the surface of the sand and filters down- 37 ward through the sand, is collected by gravel and underdrains; and passes out purified. The area provided is such that at times of greatest consumption the rate of filtration of the reser- voir supplies will never exceed 5,000,000 gallons per acre daily with one filter out of use, and for the Maidencreek supply the highest rate will be 3,000,000 gallons per acre daily with re- serve for cleaning. The processes resulting in the purification of the water in pa.ssing the sand are very complex and need not be discussed in detail here. They consist in part of the .straining out ot the suspended matters and in part of the oxidation of sus- pended or dissolved organic matters, this oxidation being carried out by the oxygen dissolved in the water applied. It is thus important that the water applied to the filters should be- aerated, and devices are included in the estimates for aerating the water as it enters the filters. The system of sand filtration can be depended upon at all times to remove the muddiness in the waters, and the organic matters producing tastes and odors- and the odors themselves, and should there be any objection- able bacteria in the applied water, the filters can be depended, upon to remove substantially all of them. Pennsylvania Sajiitation Company s Process. — The filters of the Pennsylvania Sanitation Company are constructed under certain patents. The peculiarity of this system consists in the use of elevated filters supported on steel structures. The sand is supported by a thin layer of gravel which in turn rests upon an open grating, and there is no system of uuderdrainage, but the effluent from every part of the filter drops at once from the grating into a receptacle or reservoir below. The depth of sand employed is two feet, or only half as great a depth as is used in sand filtration. The prices in the proposals submitted to your honorable Board amount to about $3.40 per square foot of filtering area. We have taken these bids as a basis of estimates and have not taken up the question as to whether the prices were reasonable ones for the construction of the apparatus in question, as the apparatus is covered by patents, and the Pennsylvania Sanita- tion Company can undoubtedly make such prices as it sees fit on that part of the apparatus covered by patents and not open to competition. The cost is about four times as much per square foot as the cost of sand filters, and in order to bring the system as a whole- into the same range of cost as other systems of filtration, or for reasons not known to us, the Pennsylvania Sanitation Company recommend that it be used at a rate of filtration of 12,500,000 gallons per acre daily. 38 The rate of filtration which can be safely employed in sand filtration is the outcome of the experience of sixty years, and of hundreds of cities. The rates of filtration mentioned in con- nection with sand filtration are considered perfectly safe. It is well known that under favorable conditions rates of filtration can be employed considerably greater than those considered in our estimates for sand filters. There is, however, an element of danger in the use of higher rates, and the proportion of bac- teria which passes a filter, increases very rapidly as the rate in- creases. In our experiments, and particularly in the experi- ments with bacillus prodigeosus, the proportion of that kind of bacteria which passed the filter constructed on the plans of the Pennsylvania Sanitation Company, was much larger than the proportion passing the sand filter ; and we believe that one of the most important elements in this difference was the higher rate, and also the less depth of sand, and we do not consider the depth of sand sufficient, and the rate of filtration low enough in the plant of the Pennsylvania Sanitation Company, to insure the removal of objectionable bacteria, should such be present in the applied water. It has been suggested that the aeration of the filtering ma- terial and of the effluent in this system play an important part in the purification. The effluent is certainly most thoroughly aerated as it falls from the filter, and if it contained objection- able odors from dissolved gases, this aeration would tend to evaporate and remove the same. The aeration will not, how- ever, remove any turbidity or muddiness resulting from too high a rate of filtration or too shallow a sand bed, nor will it remove any objectionable bacteria which may have passed the filter, for similar reasons. The alleged removal or destruction of bacteria by electrification due to the impact of the falling drops of water on the surface of the water below has not made itself apparent in our experiments, and we have not been able to discover the slightest reason for supposing that any bacterial purification is effected in this waj'. This system of filtration is also lacking in one respect most essential to securing the best results by any system of filtra- tion, namely, the regulation. The bottom being open at every point, the effluent is free to pass away as fast as it can get through the sand, and the only possible regulation consists in applying the water at a constant rate and distributing it over the whole area of the sand. A perfect distribution of water over a large filtering surface, that surface being open enough to carry off all the water it receives, -has not yet been achieved. 39 The plans shown by the Pennsylvania Sanitation Company «how the water applied at a large number of points, but there is no attempt made to distribute the applied water to every point in the filters. When the sand is cleaned, water is ap- plied to it at various points and runs through the sand at these points. With a head equal to the depth of sand, as it prac- tically is in this case, before the water starts to accumulate on the surface, and counting the water column in the sand itself, which is equal to the height of the sand, water will flow through sand of the coarseness used in these filters at about five times the nite recommended for the operation of the plant. That is to say, if the nominal quantity of water is applied to a given section, the sand being clean, it will run through about one-fifth of that section at about five times the nominal rate, the other four-fifths of the section remaining out of use. As the sand in that part of the filtering area first brought into use becomes somewhat clogged by the suspended matters in the applied water, the rate of filtration in this section is re- duced, the water on the surface extends farther, and other parts of the filtering area are brought into action which filter at finst at the same rate, that is, at five times the nominal rate. This clogging and moving of the center of action gradually goes on until the water extends over the whole surface of the filter. When the water has covered the whole surface it gra- dually increases in depth and the head is thus increased, over- coming the increased friction. This goes on until the water is -a foot deep, and the head amounts to one and a half times the depth of sand. The filter must then be allowed to drain off -and be cleaned, otherwise the applied water would overflow. It is obvious that this procedure is not calculated to give a high bacterial efficiency, and that this system of filtration can- not be depended upon to remove objectionable bacteria, should such be found in the raw water, and for this reason it is not suitable for the Maideucreek supply. The filtration removes nearly all of the organisms causing disagreeable tastes and odors and the odors themselves, and would thus, if applied, remove one of the most serious objections to your reservoir supplies. An inspection of the table of estimates, however, -shows that this system is more expensive than either sand or mechanical filtration for the Antietani supply, and we have reason to think that it would also be more expensive for the Bernhart and Egelman supplies. As between sand filtration and the Pennsylvania Sanitation Company's filtration, the latter substitutes a smaller area of much more expensive filtering surface, and is open to the dis- 40 advantages of requiring a much higher rate of filtration, which is objectionable ; to that of using a less depth of sand, which is also objectionable ; to the entire absence of adequate regula- tion, which is very objectionable ; and to numerous structural defects which we deem it unnecessary to take up at the present time in detail. Its advantage consists in the very thorough aeration of the efBuent, but this aeration is often unnecessary, and if at any time required, can be much more cheaplj' pro- vided in another way. Mechanical Filters. — The subject of mechanical filtration is in several respects in un.satisfactory shape. The essential feat- ure of mechanical filtration consists in the use of a rate of fil- tration forty or fifty times as great as is employed in sand filtra- tion, and in the use of sulphate of alumina or alum. The function of alum is to coagulate and draw together the very fine suspended matters of the water and allow them to be more readily removed, and so to allow rates of filtration to be em- plo3'ed which would otherwise be quite impossible. The suc- cessful purification of water with very high rates of filtration is absolutely dependent upon the use of alum or other coagulants. The Hyatt patent covering the use of alum applied as it would necessarily be applied with your conditions does not expire until 1901, and prior to that time there is a question as to your right to use mechanical filtration unless the filters are secured from the owners of the above mentioned patent. Our esti- mates for mechanical filters are for plants constructed in accord- ance with the principles which we believe best adapted to secure good results by this system of purification. Whether or not you would be able to secure the use of such filters prior to the expiration of the above mentioned patent, is a question which we are unable to determine, but which must be settled before this system of filtration is adopted. Mechanical filtration for the reservoir supplies is open to the objection that small pumping or power plants would be re- quired for pumping wash-water and stirring the filtering mate- rial, necessitating keeping up steam and the cost of constant attendance. These features are avoided by sand filters, which can be made largely automatic in their action and which re- quire no steam or power plants for their operation. With the reservoir supplies the difference in operating expenses is an im- portant reason for not adopting mechanical filters. For the Maidencreek supply this objection is less important, as all of the machinery could be concentrated in the immediate neigh- borhood of your present pumping station, and the present staff , with such increase as might be necessary, would attend to its 41 operation. In this case mechanical filters, all question of patent rights aside, would be $12,500 cheaper than sand filters on a corresponding site and with corresponding pumping machinery. The cost of operating expenses is about $1,000 greater, owing to the alum required. Taking into account the interest on the cost of construction and the operating expenses, the difft-'rence in cost between sand and mechanical filtration is not important > and we prefer the one which we consider the most certain to yield at all times a thoroughly purified effluent, and the one which avoids all possible discussion or litigation in regard to patent rights, namely, Sand Filtration. CONCLUSIONS. We find that the waters of the difTerent supplies are subject to various sources of pollution, and are, or may become, at some time, injurious to health. All of them are sometimes muddy, and three of them are subject to disagreeable tastes and odors. We find that these unhealthy and disagreeable qualities can be removed by suitable filtration, and we consider it for the best interests of the City of Reading that they should be so re- moved. We find that the system of filtration of the Pennsyl- vania Sanitation Company is very expensive per unit of filter- ing area as compared with sand filtration, and that it has no corresponding advantages, while it is open to very serious and fundamental objection, and in the case of the Maidencreek supply, at least, is unsuitable for the work to be done. We find that mechanical filtration properly carried out with filters of the best type will not be very much cheaper than sand filtration, while the operating expenses will be considerably greater. We do not consider that it will give in any respect better results than those obtained from sand filtration. It is not even certain that as good results can be obtained from it, while there may be difficulty in getting filters of the best type while avoiding litigation in regard to patent rights. We respectfully recommend the construction of sand 'filters as being the best adapted to the requirements of the City of Reading. Respectfully submitted, EmIL 1,. NUEBLING, Superintendent atid Engineer of Water Works. Ali Pan. Dicttomaceae : Asterionella. 336 16 6 13 2.5 4 IS 3-5 0.5 2 1-5 I 0.2 0-5 TVIelosira 0-5 Chlorophyceae : Protococcus Desmiaeae r Staurastrum Infusoria : Chloromonas TVEonas 2 Cercomonas Total 398 108 vegetable strong Oder of asterion'a. 300 3 58 faintly vegetable. 138 15 86 faintly vegetable. cent's some iron rust and veget'e fibre. 149 7i very faintly vegetable. contains much clay. 1320 Odor Remarks Bacteria per C. C 45 EIOI.OGICAI. EXAMINATIONS, &.C.— Continued. Date op Coli 1897- Number of Sample. 59 60 61 62 Applied Water. Effluent Sand Tank. Effluent Sanit'n Co.'s Bed. Effluent Sanit'n Co.'s Pan. JDiatomaceae : Asterionella 17.S 1.5 22.5 1 5-5 272.0 8 6 12 2 4 20 24 1-5 12 I 2 I 20 4 0.5 I Navicula 2 I Cymbella Cyclotella Chlorophyceae : Pandorina Cyanophycea : Microcystis Fungi : Mould Infusoria : Chloromonas Trachelomonas. . . . Cercomonas I Rotifera : Rotifer Anuraea Total 396 146 vegetable. contains considerable clay. 676 590 730 14 40 faintly vegetable and earthy. contains considerable clay and rust 127 160 130 27 60 very slightly vegetable. contains much clay. 166 180 179 5 Amorplious 56 ■Odor sliefhtlv veg- Remarks etable and unpleasant. Bacteria per C. C, May II 225 Bacteria per C. C, May 12 290 320 Bacteria per C. C, May 13 53 BIOI.OGICAI, EXAMINATIONS, 8cC.— Continued. Date OF CoLivECTiON, May 17, 1897. Number of Sample. 63 64 65 66 Applied Water. Effluent Sand Tank. Effluent Sanit'nCo.'s Bed. I I 10 I Effluent Sanit'nCo.'s Pan. Diatomaceae : 60.5 34-5 J5 0-5 1 I 24 0-5 2 44 48 I I 4 Fragilaria Navicula Cyclotella I 0.5 Chlorophyceae : Pandorina Fungi : Mould-. Infusoria : Monas Cercomonas Chloromonas Cryptomonas Rotifera : Rotifer 20 Total 231 134 slightly vegetable. contains clay and iron rust. ■980 995 198 154 6 36 slightly earthy. 162 160 95 40 13 58 faintly vegetable and earthy. 215 134 120 42 22 Amorphous 48 Odor faintly Remarks vegetable. Bacteria per C. C, May 14 385 Bacteria per C. C, 270 Bacteria per C. C, May 16 Bacteria per C. C, May 17 210 105 54 BIOLOGICAL EXAMINATIONS, &l^.— Continued. Datk of Collection, May 21, 1897. JMumber of Sample. JDiatomaceae : Asterionella Synedra Cyclotella Navicula Chlorophyceae : Scenedesmus Protococcus Cyanophyceae : Oscillaria Fungi : Mould Infusoria : Chloromonas Mallomonas Miscel/aneous : Pine pollen Total Amorphous Odor Remarks Bacteria per C. C, May 18. Bacteria per C. C, May 19 Bacteria per C. C, May 20 Bacteria per C. C, May 21 75 Applied Water. 13 57 5 I 5.3 12 105 134 none. contains some clay. 203 232 132 142 76 Effluent Sand Tank. 0.5 0.5 2 38 none. contains some clay. 42 45 32 40 77 Effluent Sanit'nCo.'s Bed. 78 Effluent Sanit'n Co.'s Pan. I 0-5 10 I 8 19 very faintly vegetable. 38 170 62 42 .46 slightly vegetable. 86 320 150 200 55 BIOIvOGICAL EXAMINATIONS, SlC— Continued. Date of Collection, May 24, 1897. Number of Sample. 79 80 81 82 Applied Water. Effluent Sand Tank. Effluent Sanit'nCo.'s Bed. Effluent Sanit'n Cc.'s Pan. Diatmnaceae : Synedra Asterionella 25.5 2.5 12 6 6 [ I 4 I 2 2 0.2 3 0-5 Chlorophyceae : Pandorina Protococcus Desmideae : Staurastrum Fungi : Mould 10 Infusoria : Chloromonas IMonas Total 53 102 faintly earthy. very clear. 2CX) 140 ' 93 5 34 none. contains a little clay. 15 25 26 5 52 none. 14 lost. 27 13 Amor'Dhous 46 Odor none. contains a Bacteria per C. C, May 22 little clay. 150 Bacteria per C. C, TVTav '^% 233 Bacteria per C. C, May 24 275 56 BIOIvOGICAI, EXAMINATIONS, BiC.— Continued. Date of Collection, May 27, 1897. Number of Sample. 88 89 90 9t Applied Water. EfHuent Sand Tank. Effluent Sanit'n Co.'s Bed. Effluent Sanit'n Co.'s Pan. Diatomaceae : Synedra 39 2 1-5 6 6 0.75 8 1-25 0.5 I Asterionella Navicula Chlorophyceae : Protococcus Raphidium Cyanophyceae : Oscillaria Miscellaneous : Grain pollen 6 Total 54 124 slightly vegetable and earthy. 185 175 lost. 9 40 faintly earthy. contains some clay and rust. 18 17 20 2 60 faintly vege- table and earthy. 15 10 10 7 52 faintly earthy. Amorphous Odor Bacteria per C. C, May 25 186 Bacteria per C. C, 193 205 Bacteria per C. C, May 27 57 BIOLOGICAIv EXAMINATIONS, ScC— Concluded. Date of ColIvECTion, May 31, 1897. Number of Sample. Diatomaceae : Asterionella. . . . Synedra Cyclotella. Navicula Melosira Cymbella Chlorophyceae : Protococcus.... Fungi : Mould Spores Rotifera . Rotifer Miscellaneotis : Worm Total.... Amorphous . Odor Remarks. Bacteria per C. C, May 28 Bacteria per C. C, May 29 Bacteria per C. C, . Ma}^ 30 Bacteria per C. C, May 31 Bacteria per C. C, June I Bacteria per C. C, Jun e 2 92 Applied Water. 16 15-5 2 2-5 6 I 24 73 130 slightly vegetable and earthy. 170 185 184 193 194 93 Effluent Sand Tank. 1.5 I 2 36 none. 19 i5 18 iS 12 94 Effluent Sanit'n Co.'s Bed. 14 64 none. contains considerable clay and amorphous matter. 41 24 13 12 16 53 95 Effluent Sanit'n Co.' Pan. 1-5 7 54 faintly earthy. contains considerable rust. no sample, 43 67 165 217 160 58 SHOWING AVEKA-GB NUIvlBER OK BACTKRIA IN APPLIED WATER AND EFFLUENTS ; ALSO PERCENTAGE REMOVED. TIME. 1897. SAND FII,TER. SANITATION CO.'S FILTER. AVRRAGE I NUMBER OFJ BACTERIA, I AVERAGE NUMBER OF BACTERIA. tu > =-» 2 AVERAGE NUMBER OF BACTERIA. 3 C3 "ftrt 4^ ODOR OF THE ANTIETAM WATER COULD BE REMOVED BY FILTRATION. During the latter part of August, 1897, the water from Au- tietam Lake became very offensive to the taste and smell. The water was then turned onto the experimental filters. At no time during the test could the disagreeable odors or tastes be detected in the effluents of the filters^, thus showing that these objectionable tastes and odors can be removed by filtration. In order to ascertain the cause of these tastes and odors, sam- ples taken from the surface of Antietam Lake, from the inlet to filters, and from the effluents of filters, were sent to Mr. F. S. Hollis, Bacteriologist, for investigation. The results of his examinations appear in the following table. In his remarks, on the results, he says : ' 'The Anabaena is. the important organism in the lake and applied water, and is the one which is causing the trouble, by imparting its charac- teristic odor. It is generally a surface growth, which probably explains the smaller amount present in the applied water,, which is, I believe, drawn off at some depth below the surface. The larger amount of amorphous matter seems to indicate that some may also have died in the pipe. The organisms are as. usual mainly removed in the efSuents, and but little odor re- mains." ANAB/€NA (CYANOPHYCE/E) MAGNIFIED 485 DIAMETERS 63 Microscopical Examination of Antietam Water. By F. S. HOLLIS. Date of Collection, August 30, 1897. Date of Examination, September i, 1897. Diatoinaceae : Asterionella Cyclotella Navicula Gomphonema.. Synedra Tabellaria Chlorophyceae : Scenedesmus.... Gonium Pediastrum Pandorina Protococcus Destnideae : Staurastrum .... Cyanophyceae : Anabaena Clathrocystii.... Microcystis Infusoria : Cercomonas Ceratium Chloromonas.... Codonella Glenodinium.... Monas Peridinium Trachelomonas. Rotifera : Rotifer Surface Antietam Lake. 34 64 9 15 2085 120 36 6 144 3 IS 21 31a 21 Applied Water. (Antietam.) 12 42 3 87 8 80 24 18 14 639 60 6 36 72 360 27 72 Effluent Sand Tank. Effluent Sanit'n Co.'s Pan. Total.... Amorphous . Color Turbidity.... Sediment Odor 2941 375 .27 distinct. 1608 1005 distinct. considerable considerable Vegetable. Ch'r cteris ic odor of ana- baena dist'ct Vegetable. Odor of anabaena. 14 34 •15 very slight. none. faintly earth3'. 24 68 .20 slight. slight rust. slightly earthy and sweetish. 64 APPENDIX "B/' EXAMINATIONS OF WATER SUPPLIES. OF WATER SUPPLIES. By H. W. Clark. In Parts Per ioo.ooo. Samples collected, 1897. April 24. April 24. April 24. April 24. April 24. April 24. Analysis made April 28. April 28. April 28. April 28. April 28. April 28. Kumber of corre- sponding Biologic'l Sample. 2 3 I 6 Antie- tani East Stream. 7 8 Sources. Maiden- creek above Willow Creek. Willow Creek at Davies' Mill Dam. 67 none. slight. 0.10 none. I5-.20 11.30 0.18 0.0038 0.0068 0.1410 ' 0.0006 0.0 100 Maiden- creek at Pump'g Station. Antie- tam West Stream. Antie- tam near Over- flow. Temperatnre[Fahr] Turbidity very slight. very slight. 0.15 very slight. 9.10 6.10 0.14 0.0024 0.0090 0.0980 o.ooo^ 0.0600 65 none. slight. 0.15 very slight. 930 6.90 0.14 00046 0.0082 0.0980 0.0006 0.0500 65 none. very slight. 0.19 very slight. 9.40 350 0.12 0.0028 0.0128 0.0570 0.0000 0.1500 63 none. slight. 0.21 very slight. 9.40 2.60 0.1 1 O.OOIO O.OIIO 0.0500 0.0002 O.I 70c 64 none. ver}' Color slight. 0.21 Odor none. Total solids (in 7.20 Hardne.ss 2.50 Chlorine 0. II Free Ammonia Albuminoid Am'ia Nitrogen as nitrates Nitrogen as nitrites Oxygen consumed.. 0.0056 0.0150 0.0460 0.0002 0.2400 65 OF WATER SUPPLIES.— Continued. By H. W. Clark. In Parts Per 100,000. Samples collected, 1897. April 24. April 24. June 2. June 2. June 2. June 2. Analysis made April 28. April 28. June 4. June 4. June 4. June 4. Number of corre- sponding Biologic'l Sample. 4 5 96 97 98 99 Sources. Bern- hart Influent Bern- hart Reser- voir near overflow 61 very slight. very slight. 0.17 none. 8.80 510 0.1200 0.0056 0.0146 0.0330 0.0002 0.1300 M'Kent- ley's Spring. Bern- hart Reser- voir at Pier. Egel- man Influent Egel- man Reser- voir near overflow Temperature[Fahr] Turbidity 66 very slight, slight. 0.20 none. 8.90 2.90 0.1 100 0.0040 0.0138 0.0100 0.0000 0.1700 none. verv .slight. 0.00 none. 13.00 8.90 0.1400 0.0008 0.0016 0.0560 0.0000 0.0100 slight. decided. 0.25 none. 950 4.10 O.IOOO 0.0040 0.0156 O.OIIO 0.0002 0.2200 slight. decided. o.iS none. 9.00 2.00 0.1900 0.0002 0.0078 0.0300 0.0000 0.1500 slight. Sediment decided. Color 0.26 Odor none. Total solids (in 100 000) 8.20 Hardness. . * 1.80 0.1800 Free Ammonia...... Albuminoid Am'ia Nitrogen as nitrates Nitrogen as nitrites Oxygen consumed.. 0.0026 0.0108 0.0250 0.0000 0.2000 66 BIOLOGICSL EXHMINHTIONS OF WATER SUPPLIES. By F. S. HoIvLIS. Maidertcreek. Standard Units Per C. Date of c ollectio n, 1897. Number of sample Source. Diatomaceae : Navicula Synedra Cyclotella Melosira Gomphonema.. Pinnularia Tabellaria Cymbella Asterionella Chlorophyceae : Spirogyra Pandorina Desmideae : Cosmarium Staurastrum Closterium 1' yanophyceae : OsciUaria Infusoria : Monas Codonella Chloromonas.... Trachelomonas, Rotifera : Conochilus Anuraea Rotifer Miscellaneous: : Ve geta ble fibre, ToFalTTTT" Amorphous Odor Remarks Bacteria per C. C, Apr 24 'A pr 24 2 Maid'n cr'k ab. Willow creek. II 05 0.5 12 42 194 vege- table and earthy, con- tains consi- derable clay. 286 Willow Cr'k at Da vies' Dam. 5 18 23 294 earthy and un- pleas- ant, cont'ns fcwl'rg pieces vegmat sm clay 365 Apr 24 Maid'n creekat P'mp'g Station 10 15 0.4 10 2.4 47 158 veg'ble and aro- matic 292 Maid'n creekat P'ra'ng Station M'yii M'y ii Maid'n c'kinl't H'pden res'rv'r 19-5 21 I _4o 93 264 veg'ble and earthy contn's much clay t'k soon aft hv rain 1174 7 27-5 4 60 io5 348 slight- ly vege- table con- tains much clay. 330 __54_. Hamp- den Reser- voir. Maid'n c'kinl't H'pden res'rv'r I 17 5-5 ATyiy 67 23 21 I 9 3 40 12 2 2 8 9-6 80 235 104 vege- table and earthy 238 107 530 slight- ly vege- table sample t'kenat lower end of ov'rfl'w 860 At upper end of overflow 826 67 Biological Examinations of Water Supplies. — Co7itinued. Antietani. Date of Collection Apr 24 Number of sample Source. East East Stream Stream 32 7 70 12 Diatomaceae : Melo.sira Navicula Synedra Pinnularia Cyclotella Asteriotiella ... Gomphonema Surirella Cymbella Chlorophyceae : Scenedesmus . Protococcus ... Pandorina Closterium .... Spirogyra Conferva Raphidium. ... Desmideae : Staurastrum. . Cyanophyceae : Microcystis.... Infusoria . Chloromonas. Monas Codonella Ceratium Cercomonas. . . Rhisopoda . Arcella Actinophrys . . Roiifera : Anuraea Conochilus ..., Rotifer Miscellaneous : Crustacean ... Vegetable fibre.. Total 12, May 6 Apr 24 47 380 T2 18 _4o_ I458 May 6 Apr 24 West Stream 3 25 10 I 46 West Stream 4 5 12 2 1.5 2 160 0.5 259 May 6 M'y ig'M'yig Lake near over- flow. 1-5 12 29 257-5 4 42 56 Lake near over- flow. 2.5 15 I II 50 238 36 12 6 36 40 243 529 416 73 Lake at Gate House. Bot torn of Lake- at Gate- House. 2.5 I 157 2.5 16 260 1-5 I 2 20 6 48 112 80 20 74 12 5 19-5 6.S 17 40 729 :ii5 68 Biological Examinations of Water Supplies. — Continued. Antietam. — Concluded. Date of Collection Number of sample Source. Apr 24 East Stream May 6 47 East Stream Apr 24 West Stream May 6 46 West Stream Apr 24 Lake near over- flow. May 6 M'y igM'y 19 I Lake near over- flow. 73 Lake at Gate' Hou.se. 74 Bottom of Lake at Gate House. Amorphous Odor Remarks. .., Bacteria per C. C. Bacteria per C. C. 15 feet below surface Bacteria per C. C. 34 feet below surface 450 earthy and slight- ly vege- table. 700 340 earthy. con- tains consi- derable clay. 850 very faintly vege ta- ble. confer- va and Uloth'x grow'g on bed of str'ni 720 580 vegeta- ble and earth y. cont'ns much clay & amor- phous matter. 2925 268 slight- ly vege- table and ear thy. con- tains consid- erable clay. 62 268 vege- table. con- tains consid- erable clay. 580 660 vege- table. 16 312 445 700 vege- table and ea rthy. 445 69 Biological Examinations of Water Supplies. — Continued. Bernhart. Date of collection Apr 24 M'yii 56 M'yl7 Apr 24 M'yi7 M'yii M'yi7 June 2 Number of sample 4 68 5 70 57 69 97 Source. B'rnh't influ'nt B'rnh't influ'nt B'rnh't influ'nt B'rnh't resne'r over- flow B'rnhl resne'r over- flow B'rnh't res'rv'r at Pier. B'rnh't res'rv'r at Pier. B'rnh't res'rv'r at Pier. Diatomaceae : Cymbella 13 13 8.5 60 2 I I 2 10 14 80 1-5 2.5 10 2.5 50 100.5 2 II.5 1-5 44 26 3 144 18 0.4 12 20 2 123 I 10 13 I 16 45 52 4 1-5 154 24 2 263 2 34.5 136 14 2.5 I 10 I 115 20 0.5 68 16 52 44 4 4 2 88 48 16 Synedra Navicula Melosira 4 Cyclotella Piniiularia Asterionella Gomphonema... Chlorophyceae : Stigeoclonium.. Pandorina Protococcus Scenedesmus.... Raphidium Pediastrum Cyanophyceae : Osoillaria 7 57 4 28 Infusoria : Chloromoiias.... Codonella Cercomonas 90 Cryptomonas . . . Trachelomonas Rhizopoda : Actinophrys Rotifera : Rotifer 78 72 10 Anuraea Conochilus Miscellaneotis : Vegetable fibre Total 99 7S0 faintly vegand earthy contii'.s many plant stalk.s n.so 106 670 vege- table and earthy cont'ns much fragm'l and am orpho's 700 66 550 veRe- table and earthy cont'ns conis'bl'' clay M amor'h ous mat 896 383 190 vege- table 1806 446 460 vegand slightly aro- matic cont'ns much clay 295 465 480 vegand slightly aro- matic 478 368 vege- table cont'ns much clay ^50 361 160 Odor slightly Remarks veg'ble and earthy Bacteria per C. C. 70 Biologic Ai, Examinations of Water Supplies. — Continued. Egelman and McKentley s Spring {Bernharf). Date of collection Apr 27 M'yig June 2 Apr 27 M'yi9 June 2 M'yil June 2 Number of sample 21 71 98 22 72 99 58 96 Source. Kgel- man In- fluent. Egel- xnan In- fluent. iigel- man In- fluent. ijgelm resne'r over- flow. Egelm resne'T over- flow. Egelm resne'r over- flow. ..ic- Kent- ley's Spiiner. Mc- Kent- ley's Spring. Diatomaceae : 3 8 2 10 7 4 I I 0.2 I 20 17 6 7 1.5 I 4 6 10 0.25 2 2 6 3 182 36 12 24 20 24 11 II 3 38 10 8.5 5 2.5 4 8 0.5 221 2 20 2 Pinnularia I Cyclotella Cymbella Surirella Tabellaria Gomphonema... Chlorophyceae : Spirogyra Protococcus Desmideae : Staurastrum Cyanophyceae : Oscillaria Infusoria : Monas Cercomonas Peridinium Dinobryon Rotifera : Rotifer Miscellaneous : Vegetable fibre Total 23 364 faintly vege- table. 13s 34 500 none. 355 53 520 distinct ly vege- table. 287 176 faintly vege- table. 264 97 256 none. 135 249 242 slightly v'g'tble and earthy. 22 28 none, 57 23 74 none. -Amorphous Odor Remarks Bacteria per C. C. sample cont'ns a little rust. 71 BioLOGicAi< Examinations of Water Supplies. — Concluded. Sacony Creek at Kutztown, etc. Date of collection M'y26 Number of sample Source. M'y26M'y26 83 84 Sacoiiy Creek abKutz town St NorSch Sewer Outlet. oacony C'k 150 feet bel Kutztn StNSch Sewer Outlet. 85 M'y26 86 JDiatomaceae : Synedra 131.5 Navicula ; 43.5 Cymbella 3 Cyclotella 6.5 Melosira 10 Gomphonema... 2 Pinnvilaria 6 Asterionella... Pleurosigma.. Surirella ■Chlorophyceae :■ Prijtococcus... Pandorina Raphidium . . . Spirogyra Scenedesmus. Desmideae : Staurastrum.. Cyanophyceae : Oscillaria Infusoria : Trachelomtmas Monas 0.5 Chloromonas.... 2 Ceratium Peridinium.... Cercomonas... Rhizopoda : AcHnophrys... Rotifera : Rotifer Anuraea Miscellaneous : Vegetable fibre Crustacean frag Total 449 28 no 32.S 4 6.5 j6 6 Sacouy Cr'kioo feet bel Slaugh ter House. Sacony Creek 50 feet below tann'ry 135-5 43 3 7-5 40 18 6 420 40 433 684 0.2 I 117 57 I 13-5 10 6 240 0.2 2 40 M'y26 M'yii 87 Saconv Creek second bridge below tann'ry 137 51 4 5-5 44 24 53 Hamp- den Spring Apr 29 Apr 28 28 23 Penn Street Reser- voirs. 160 3 16 0.2 I 487 '446 75.5 3-5 188 1 120 6 12 4 48 60 Distrib System Low Service Court St n'r 7th. 240 5 7-5 5-25 I 60 26 0.2 4 16 190 40 575 l6n" 72 BioLOGiCAi, Examinations of Water Supplies. — Concluded. Sacony Creek at Kutztown, etc. — Concluded. Date of collection M'y26 M'y26 M'y26 M'y26 M'y26 M'yil Apr 29 Apr2S Number of sample 83 84 85 Sacony Cr'kioo feet bel Slaugh ter House. 86 87 53 28 23 Source. Sacony Creek abKutz town St NorSch Sewer Outlet. Sacony Cr'kiso feetbel Kutztn StNSch Sewer Outlet. Sacony Creek 50 feet below tann'ry Sacony Creek second bridge below tann'ry Hamp- den Spring Penn Street Reser- voirs. Distrib System Low Service Courtst n'r 7th. Amorphous 268 vege- table and earthy. cont'ns much d't'ch'd algae growth mainly spir'g'a 750 230 vege- table and earthy. 2,123,90c bact'ri' perc. c. in sew- age frm Sewer Outlet 13.740 276 vege- table and earthy. ,5,526 290 slightly vege.' and earthy. 3,105 264 vege- table and earthy. 2,900 18 none. cont'ns a little clay. 5 170 d'cid'ly veg'ble and oily. Sample cont'ns much fragm'l mater'l 77 180 Odor vegetbl and slightl> oily. Sample Bacteria per C. C. taken f faucet nr dead end in water main. 347 Bacteria per C. C. May ii, 1897, in Hampden Drift water, 122. 73 oiPiisTionsrs — OF— ERASTUS G. SMITH, Professor of Chemistry or Beloit College, Beloit, Wisconsin. "W". P. MASON, Prof, of Chemistry, Rensselaer Poly- technic Institute, Troy, Ne-w York. RUDOLPH HERRING, Consulting Engineer, News^- York. [Copy.J Thomas M. Thompson, Director Board of Public Works, Philadelphia, Pa.: Dear Sir : In reply to your inquiry under date of July i5tli, concerning: the purification of waters by filtration, I beg leave to submit, the following as an expression of opinion formed from study- and research upon the general problem, and especially frora. personal examination of filtration plants as they have been in- stalled and .sticcessfully operated on a large scale in the larger- European cities. The problem of the economical and efficient filtration of wa- ters on the large scale for domestic use is one of the most seri-- ous presented to any municipal body, and any suitable complete- discussion would far out-reach the limits of this brief com- munication. It is now about fifty years since the first attempts were made- at I/Ondon to purify the waters for that metropolis, by filtra- tion. The aim was then merely to deliver a bright, clear wa- ter. It was my fortune to see recently this first filter bed still. in operation, the same, having been used continuously, and with practically no expense for repairs, since its first construc- tion — a striking illustration of the desirability of thorough, work in all engineering enterprises. Equally striking is the. fact that the type of filter thus first built has been retained, from the finst, and those most recently constructed in Eondon,. and now under process of construction there and on the Conti- nent, are of the same form and plan. Such differences as da^ exist are merely those of detail, as e. g., the slope of the wall^ 74 character of the sand used, under drainage, etc. The system 'lias spread to other portions of England and the Continent un- til most of the important large European cities have adopted this form of filter for the satisfactory and efficient purification • of their water supplies. Paris, Antwerp, Berlin, Hamburg, were among the mo.st important places visited this summer. In the process of slow sand filtration all filter beds are con- structed in exactly the same manner. A ba-sin with strong re- taining walls, covering an area from about 0.5 to 1.6 acres, the sides and bottom of the ba.siu being coated with heavy cement to prevent both the escape of the iiltered waters and the admix- ture with the filtered waters of natural "land-waters." Such 'basins are usually about 9 feet in depth. Upon the cement floor is placed a series of canals made from tile or brick laid loosely together for the purpose of collecting the filtered waters into a central pipe, discharging into the main pump well. Upon these are placed succe.ssively a layer of coarse broken stone, then a layer of coarse gravel, then a layer of fine gravel and fi.nally a layer of sand, usually from 24 to 36 inches in thickness. The specifications for two new beds, now under process of construction at Paris, distribute these relative beds ■ as follows : Fine sand to inches Coarse sand 10 " Fine gravel 4 " Coarse gravel 6 " Bricks and broken stone 8 " These beds at Paris are somewhat thinner than usually con- structed in England, but the proportions between the layers are >quite the same, and careful daily analyses and examinations have proven them to be quite as efficient. Upon the surface thus prepared the water quietly flows, and .-gradually settles down through the sand. When the process is properly conducted, efHuent waters are bright, clear, are grealtly improved chemically, and careful examination shows them to have lost all, even the lowest and most minute forms "Of life. The filtering process continues until the surface of the bed becomes coated with a layer from the sediment and flotaut matters contained in the waters, which finally completely clogs =and prevents further passage of the water. The beds are then • allowed to drain, this superficial layer removed, the beds •cleaned and smoothed over, and the waters again admitted. This process is so simple it can be understood by anyone. 'There is no expensive and complicated machinery to install and 75 keep in repair, uo chemicals are essential to the process, or Tised, and yet it requires the utmost care in supervision. It is . a distinctly scientific process, and its efBciency and success de- pends upon a most careful scientific supervision. The more ■careful this supervision, the better results obtained. No system of filtration will operate itself, and this process least of ■ all. The services of an expert chemist are necessary. Per- haps the most interesting plant I have seen is at Berlin, at the works at Muggelsee. Here are 22 filters with a combined area ■of 12.7 acres, delivering a maximum of twenty-three million . gallons of water per diem. Each filter is separate, covered with a vaulted roof. Analyses of the water are made daily. No water is admitted to the mains until it has reached the de- . gree of purity demanded by the German health regulations, and if at an J' time the number of bacteria exceed the limits allowed the bed is immediately cut off, cleaned and not again put into -service until the standard is reached. Generally speaking, with river waters in their normal condition, filter beds need scraping and cleaning once every three to six weeks. A multi- plication of beds secures greater uniformity in quality of water ■discharged, so that thus serious fluctuations and variations in the supply are avoided. The system of slow sand filtration has ■proved itself satisfactory on the large scale, on that there is but one voice. Difficulties have been encountered, but as they have arisen the}'' have been overcome, until there is but one • opinion among those operating and, therefore, most conversant with such plants, viz. : That for treating waters on the large . scale this is the only method which has proved sati.sfactory., It is not the purpose of this communication to explain fully the theory and action of these beds. What you ask is simply an expression of my opinion based on such studies as I have been able to make, and observation on foreign works in actual operation. But yet, it is of the utmost importance for us to in- ' quire, what is the efficiency of these beds? And what assur- : ance can we have that a similar system will grant the needed purification of the waters and guarantee protection to the ■ consumers? I will not attempt to re- array the figures proving beyond all question the protection offered to a community by the use of such a system of filtration of its water supply. Modern science recognizes that system of filtration only as ■ effective which removes the minute micro-organisms from a water. A water may be clear, the organic matters may be re- ■ duced to a minimum ; but unless, also, the process shows a -practical removal of micro-organisms (or more popularly the removal of the "bacteria," or "microbes," or "germs"), the 76 system fails. Did space allow, I would take pleasure in quot- ing to you the data issued by foreign Health Boards and exam- iners bearing on this most important point ; all of which agree in this, that a water however dirty and wholesome is rendered practically sterile and suitable for domestic use by the process, of slow sand filtration intelligently operated. I will, however, make one such quotation, and that from the last Report of the results of analyses of the Metropolitan Water- Supply undertaken during May, i8gj, by Sir E. Ftankland, K. C. B., F. R. S., on behalf of the Local Government Board. These analyses are made by Sir E. Frankland,in behalf of the- Local Government Board for the Citj' of Loudon, and reports- made monthly. There is probabl}' no higher authority in the world than Sir E. Frankland, and due weight should be givea to his official utterances. Regarding the Thames River, he .saj's : "The Thames at. Hampton was, on May 5th, turbid and pale yellow in color ; it; was of fairly good chemical, but indifferent bacterial quality. The water supplied by five companies drawing from this source- was, in every case, efficiently filtered before delivery, but three out of fourteen samples collected infringed the standard of 100- microbes per c. c. The bacterial improvement effected by the various companies is expressed by the following percentage numbers: Chelsea, 99.88; West Middlesex, 99.62; South- wark, No. 4 filter well, 99.91 ; No. 5 filter well, 98.37 ; Grand Junction, general filter well at Hampton, 98.64 ; general filter well at Kew, 97.58 ; .south filter well at Kew, 99.52 ; and. Lambeth, 99.59." Regarding the New River : ' ' The New River cut, before- entering the storage and .subsidence reservoirs of the New- River Company, was, on May 6th, turbid and very pale yellow- in color. It was of excellent chemical, but not good bacterial quality. After efficient filtration, it was delivered to consum- ers in most excellent chemical and bacterial condition, surpass- ing in respect of organic purity the majority of deep well! waters. The bacterial improvement is represented by the fol- lowing percentage numbers : General filter well, 98.81 ; No. 2- filter well, 99.41 ; and No. 4 filter well, 99.21." Regarding the water from the Lea : ' ' The water taken f rom- the Lea at Angel Road, by the East Loudon Company, was. turbid and very pale yellow. It was, in respect to chemical quality, exactly equal to that of the Thames at Hampton, but 77 •was much inferior in bacterial quality. After efficient filtra- tion, it was delivered to consumers in a high degree both of •organic and bacterial purity. The bacterial improvement ef- fected by this company amounted to the following percentage numbers : No. i, Essex filter wells, 99.15 ; No. 2 Essex well, ■99-75 ; and Middlesex well, 99.56. As is well known to you, I,ondon draws its supply mostly from the above rivers, which is delivered to the city through -seven companies. The average daily supply to the City of lyondon during May of this year was 247,155,978 United States .gallons, an average consumption per capita of 43.4 United •States gallons. The relative proportions from various sources for this vast supply were as follows : From the Thames 56.20 per cent. " " L,ea 27.02 " " springs and wells 16.65 " " ponds 13 It may be added that the water from ponds is not used for 'drinking and domestic purposes ; the water from springs and wells is from the Kent deep chalk wells, a hard but otherwise very satisfactory water, so that eighty-three per cent., or more than 205,000,000 gallons of water, was taken daily from the rivers, and satisfactorily filtered, before delivery to the City of I,onJon, during the month of May. As corroborating above report I quote from the report of the •chemists of the company, of Sir William Crookes and Pro- fessor James Dewar, concerning their examinations of the wa- ters ^delivered by the companies during May, 1897 '■ "The results of our bacteriological examinations of 258 : samples are recorded in the following table ; we have also ex- amined sixty-nine other samples, taken from special points -either at the different filter beds or at stand-pipes : Thames water, unfiltered (mean of 26 samples) 2,937 microbes per c. c. "Thames water, from the clear water wells of five Thames derived supplies (mean of 28 samples).. 40 microbes. Thames water, from the clear water wells of five Thames derived suppHes (mean of 128 samples) highest 233 microbes. "Thames water, from the clear water wells of five Thames derived supplies (mean of 128 samples) lowest 10 microbes. New River, unfiltered (mean of 26 samples) 791 microbes. .ISTew River, filtered (mean of 26 samples) 44 microbes. .River lyea, unfiltered (mean of 26 samples) 388 microbes. 78 River Lea, from the clear water well of East Lon- don Water Co. (mean of 24 samples) 68* microbes. . *Two additional samples abnormal. The water supplied to the City (not environs) of Paris is. brought from different springs, but the suppl)' is inadequate during the summer months. Large filters have been con- structed at St. Maur to filter the water of the Marne, with view to introduction into the city. These have been in opera- tion for more than a year now, under the direction of M. Cran- moisan acting under the Prefect of the Seine. Results have not yet been put in print, but records of the action of the fil- ters, kindly shown me, are quite as favorable as those given above from London. This plant interested me greatly, as the- beds are thinner than those in general use. Built after plans, of beds when the Anderson method of preliminary treatment of" waters was in u.se, it was designed originally to add this system. later ; but so satisfactory have been the results with simple fil- tration this year, and so much more economical the operation, M. Cranmoisan informs me, that all idea of adding the Anderson system to the plant has been abandoned. At Antwerp, simple filtration without any pteliminary treat- ment of the waters with iron has given fo? about eighteen months- now most satisfactory results. At Berlin the number of bacteria, after filtration, is never allowed to reach the limit of 100 per cubic centimeter, fixed by the Government as standard. At Hamburg, though using the more dirty Elbe water, the same degree of efiiciency is maintained, as shown by the re- markably concordant results of investigation at the laboratories • of both the Water Works Administration and the Health Board. It is difficult to read such results obtained at these great water filtration plants, under both private and public control, and under such exceedingly diverse conditions, without becom- ing convinced that the system of slow sand filtration has an abundant weight of argument in its favor. It has passed be- yond the stage of experiment. The unbiased and unprejudiced, records secured concerning the action of the filter beds in puri- fication of waters, attest the efficiency of the method, and are the best guarantee we can possibly have for the success of a plant constructed after the same principles and conducted in the same intelligent manner. One of the most important questions in connection with the filtration of a public water supply, is that relating to the eflect- of same on the public health. Sanitary science has shown in. recent years that some forms of disease are undoubtedly water- borne, notably cholera, typhoid fever, with great probability- 79 that other forms are transmitted in the same manner. So im- portant, however, is typhoid fever in this connection, that it is . not infrequently taken as the measure of the purity or inipurity- of a supply, and the death rate from typhoid fever may faiiljr be taken as an index of the efficiency of filtration methods and other protective methods adopted. If we should take the space to tabulate the results from leading cities, at home and abroad, the most surprising fact developed would be that the best American supplies are higher in the death rate from typhoid fever than any European supply protected by filtration. This . seems an extravagant statement, but one I believe practically borne out by actual statistics. The Engineering News for May 21, 1896, gives a table compiled by the eminent authority, John W. Hill, C. E. , in which statistics for five years for 66 cities . are quoted. According to this tabulation, accepted as authpr- atative, the city of Brooklyn stands at the head of all American large cities in its low typhoid fever death rate, viz.: 19 per 100,000, but yet it ranks 16 in the list quoted. New York stands 17; Davenport, Iowa, 18; New Orleans, ig; Milwaukee, 28; Boston, 30; Detroit, 31; Dayton, 32; Buffalo, 35; Providence,. 36; Covington, 37; San Francisco, 38; Minneapolis, 40; Balti-- more, 41; Newark, 42; St. Louis, 43; Newport, Ky., 44; Phila- delphia, 45. I can quote the table only in part, merely showing^- the relation of Philadelphia to foreign well filtered .supplies. TyphopeT Frver Death Rates in Americakt and For- eign Cities, 1890-1894. Inclusive. City. Source op Water Supply. 1890 1891 1892 1893 1894 3 6 9 12 4 8 4 6 5 2 5 5 3 5 8 9 15 10 12 8 15 9 10 4 6 19 II 15 16 9 16 15 11 16 15 19 18 13 14 15 28 23 34 18 6 30 40 28 25 29 64 64 40 4T 32 1. The Hague... 2. Rotterdam... 4. Dresden 8. Berlin 9. Breslan 10. Amsterdam.. 13. London 14. Edinburgh... 21. Hamburg 25. Paris 45. Philadelphia Filtered from sand dunes... Filtered from River Maas.. Filter gallery by River Elbe Filtered from Lake Tegel and Muggel Filtered from River Oder. . . Filtered from Haarlam dunes Kent wells, 17 per cent.; filtered from Thames and Lea, 83 per cent Filtered from reservoir in Pentland Hills From River Elbe, filtered since May, 1893 Rivers Seine, Marne, Vanne Ourque, Canal Art, wells and springs Schuylkill and Delaware I Rivers 4-9- 5-2 8.0. >3-»- 14.6. 15S- 21. S- 264;. 48.2-t 8o Since the introduction of a s^'Stem of filtration into Hamburg in 1893, the annual rate has fallen to correspond to other plants ■quoted above where filters were in operation during the whole period of five years cited. In the period of five years the aver- ■'.age for Philadelphia is 48.2. In a previous communication I •cited the rate for Philadelphia covering a period of 22 years. The annual reports of the Department of Health for Philadel- phia show that from 1870 to 1891, there was an average death rate from typhoid fever in the city of 61.9 per 100,000 popula- tion, the highest being 92.2 in 1876 and the lowest 38.2 in 1879. Equally striking in this connection is the fact that the 'di-sease does not show its ravages for a period and then disap- pear, but is constant through all the years — the deaths repre- :senting never less than 1.94 per cent, of the whole number of ■deaths in any year or more than 4.03 per cent. Such data as these (and I could extend them indefinitely did space permit) demonstrate beyond question that where the en- tire water supply is subjected to slow sand filtration, there typhoid fever is practically stamped out. Were this country ■subject to cholera epidemics I would repeat the argument for "the protective influence of sand filtration against this dreaded ■scourge. The terrible experience of Hamburg in i892-'93 ■proved the water-born nature of this disease. Hamburg -with its population of 640,000, had 8,605 deaths from cholera. Altona, separated from Hamburg only by an imaginary line lalong one street, with a population of 143,000, had but 328 ■deaths. In the words of the eminent savant Dr. Koch, ' 'cholera in Hamburg went right up to the boundary of Altona and ■stopped." Both cities took their water supply from the Elbe, l)Ut with this difference, the Altona supply was most carefully ■filtered before distribution, the Hamburg supply was delivered Taw. The one city, protected by filtration of its water supply -was immune to disease, the other, without such protection, ex- ■posed to its dreadful ravages. Profitting by this experience, most strenuous efforts were immediately put forth by the City of Hamburg, resulting in the installation of a system of slow sand filtration of the Elbe waters, one of the most admirably arranged I have ever seen, and which operated in a thorough, •careful manner, furnishes a supply of the most acceptable •character. Concerning the future of simple slow sand filtration abroad, 1 could have but one opinion. It has proved so satisfactory sthat its extension by cities now using it and its adoption by others is only a question of time. At St. Maur two additional large beds are under process of construction for filtering water Si from the Marne for the City of Paris. At Berlin, I chanced to meet the Engineer for the service at St. Petersburg, where fil- tration, already somewhat in use, is to be enlarged and per- fected after the most modern improvements in the process. At lyondon four additional large beds are in process of construction for the Grand Junction Company. Immense reservoirs for im- pounding Thames waters for a population of London estimated to reach in 1941 more than twelve millions, and to require an amount of water upwards of 500 millions of gallons daily, are now under contract and one approaching completion. Opinion from authorities endorsing the system can be quoted. The Royal Commission says regarding the London supply : CONCLUSION AS TO QUANTITY AND QUALITY. ' ' We are strongly of opinion that the water as supplied to the consumer in London is of a very high standard of excellence and of purity, and that it is suitable in quality for all house- hold purposes. We are well aware that a certain prejudice ex- ists against the use of drinking water from the Thames and the Lea, because these rivers are liable to pollution, however per- fect the subsequent purification, either by natural or artifical means, may be ; but having regard to the experience of London during the last 30 years, and to the evidence given to us on the subject, we do not believe that any danger exists of the spread of disease by the use of this water, provided that there is an adequate storage, and that the water is efficiently filtered before delivery to the consumers." Such eminent authorities as Dr. Frankland, Dr. Ray Lan- caster, Mr. Crookes, Dr. Odling, Dr. Percy Frankland have all testified to the wholesome character of the water derived from the basins of the Thames and the Lea, and by a comparison of the health statistics and death rate of other cities and large towns have proved that the quality of the present water supply to the Metropolis is entirely satisfactory, and that this ex- cellence is likely to be maintained, and even, if possible, im- proved under careful and more extended supervision. Reports are not yet in print from the filters at St. Maur for Paris, but opinions given by M. Cranmoisan already referred to, in connection with laboratory notes examined, are of the satis- factory nature of the filters there established. Herr Anklamm, Engineer in charge of the works for the City of Berlin, in view of experience with the filters of that city, and from data from more than sixty thousand analyses of the waters made at the filter laboratories in operation of the plants, expresses a most un- qualified endorsement of the system. Herr Ad. Kemna, man- S2 I* ager of the works at Antwerp, expresses the same opinion. The management of the Hamburg plant the same. Your communication asks for any information regarding mechanical filtration, and " if you have found in your experi- ence the use of alum injurious to the human system." Regard- ing the use of alum, would say, the whole efi&ciency of mechan- ical filtration is based on use of it or similar salts. When rightly and intelligently used, no alum should be found in the filtered water. I have repeatedly advised its use at different points, but such use of alum must be very carefully supervised as I am of the opinion that alum in such filtered water is de- cidedly harmful, and hence objectionable. As to mechanical filtration, I deem it inadvisable to enter into extended discussion at this time. Mechanical filters are in use in this country at something over ICO points. All of these plants are smaller plants. I think the largest is at Denver, of capacity not to exceed 25,- 000,000 gallons per diem, and not operated to that extent. The capacity of some of the best known larger plants in opera- tion is as follows : 5,000,000 at Quincy, 111. 6,000,000 at Davenport, Iowa. 7,000,000 at Atlanta, Ga. 9,000,000 at Chattanooga, Tenn. 10,000,000 at Wilkesbarre, Pa. (Manual Am. Water Wks., 1897.) It seems to me very clear, therefore, that the introduction of mechanical filters on any such gigantic scale, ' 'capable of filter- ing not less than 300,000,000 gallons of water daily," as you state in your letter, would still remain so much of an experi- ment, and insure so many uncertainties, as not to justify the attempt in face of certain results obtained by other methods already proven successful, on scale commensurate with the de- mands of Philadelphia. Mechanical filters are cheaper to in- stall, but much more expensive to operate. As to experience with mechanical filters abroad, would say, that so far as my investigation went, I found only the Ander- son method of revolving purifiers, using metallic iron, in use by private companies supplying the environs of Paris. At the only point where water is being filtered for the City of Paris itself, viz. : St. Maur, it was found as already stated that the process of slow sand filtration gave equally good results, and at less cost, and hence the management could see no reason why any mechanical system should be recommended. At Antwerp, the Anderson system has not been used since a year ago last March, on account of the added expense, and re- 83 . suits of the analyses are equally satisfactory. At all the other cities visited slow sand filtration is exclusively used. Briefly, in conclusion, to your inquiry of the 15th of July, I would say, I have no hesitation in speaking clearly and firmly for the process of slow sand filtration for Philadelphia, Pa. It is the only system in use on the large scale abroad, and the re- peated testimony by all who have given the problem careful study, is for the beneficent results of the same. I do not coh- sider there can be any reasonable doubt but that if you intro- duce this system for the purification of your water supply under suitable supervision and control, the results obtained in other cities will be repeated, and the citizens of Philadelphia will be assured of a supply at once most satisfactory wholesome and safe to use. Respectfully submitted, (Signed) Erastus G. Smith, Professor Chemistry Beloit College. Beloit, Wis., September loth, 1897. (Copy.) RENSSELAER POLYTECHNIC INSTITUTE. Troy, N. Y., September 10, 1897. Mr. Thomas M. Thompson, Diredot of Public Works, Philadelphia, Pa.: Permit me to reply to your inquiry of recent date, as follows? For a filtering plant of the size you propose I have always leaned towards the form known as the English Filter Bed System, for the reason that such system has been in operation in many cities of Europe during a great number of years, and we are thoroughly acquainted with just what it will do and what its efficiency is. We know that such a system, if properly managed, will remove 99 or more per cent, of all bacteria pres- ent in the water, and we also know from voluminous statistics that cannot be contradicted, that such a removal of bacteria has resulted in a most marked lowering of the typhoid death rate. It is scarcely necessary for me to give you here a de- tailed description of the English plant or statistics with refer- ence to its efficiencj-, cost and mode of operation ; allow me to refer you for all such information to the second edition of my ^^X% on "Water Supply " ; suffice it Ijere to say tb^t Ewropf 84 is abundantly satisfied with its sand filter beds, and their use is increasing, not diminishing. I make this latter statement because my attention has been incidentally called to a para- graph in the public press, which said that the London filters had been practically condemned ; the writer of such statement was in decided error. I have personally seen and carefully examined the London filters while working, upon more than one occasion, and I am in constant receipt of the latest returns showing their high efficiency. With reference to mechanical filtration, with the use of alum, allow me to say, that I have seen very excellent results ob- tained by such plants, and I have had charge of extended ex- periments with them, which experiments were instituted in order to show just what such plants would do ; the results were very satisfactory, and whenever such a plant was prop- erly run, I could detect no trace of alum in the filtered water. I am in the habit of recommending the English filter bed or the mechanical filter plant according to the size of the town to be supplied and the local conditions to be met. For instance, I have recently advocated mechanical filtration for Troy, N. Y., and English filter beds for Albany. It seems to me that for small cities and towns the mechanical plant is more economical and efficient, but when called upon to supply a city requiring 300,000,000 gallons of water per day, we are confronted by very different conditions. No mechanical plant, even approaching so large a size, has ever been constructed. You contemplate erecting the largest filtering plant in the world, larger than all of the great plants supplying the City of London put together. As already stated, no plant of the mechanical type has ever been constructed that at all approaches the vast size that your Philadelphia establish- ment would have to assume. The number of filters of the mechanical form necessary for j'our work, together with their housings, would unquestionable be of imposing proportions. While I look favorably upon mechanical filtration for small towns, where the plant with its attendant steam plant and its 85 experimental field in question. One additional thought : As I understand the situation, the entire plant is to be turned over to the city at the end of a stated period. The life of an Eng- lish filter plant should be indefinite like that of a reservoir or other similar structure, and its routine management should al- ways keep it in perfect condition. As to what the life of a mechanical plant may be, there are no facts whereon to base an opinion, but it is self-evident that the valves, the tubs holding the sand, and the machinery used for washing, must be renewed after no great lapse of years. In other words, the plant must be replaced after an interval more or less long. Yours respectfully, (Signed) William P. Mason. (Copy.) Office of Rudolph Herring, New Yore:, Sept. 14, 1897. Thomas M. Thompson, Esq., Director Department of Pubcic Works, Philadelphia, Pa. Dear Sir : Your favor of August 25th was duly received, asking m}^ views and opinions on the subject of filtering the Philadelphia water supply. The pressure of other business has delayed my responding to the questions contained in 3'our letter, and now I find that further lack of time prevents my answering them as fully as I had at first intended. I trust, however, that the reply may be sufficient for your present purpcse. The questions and my answers to them are as follows : First ■ What would be, in j'our opinion, the best system of filtration for the City of Philadelphia to adopt ? The Delaware and Schuylkill rivers are the two available sources of supply which are now being considered. The waters of both rivers are at times very turbid and at other times clear. Both rivers receive the sewage of the towns situated on their course, the Schuylkill water being the more highly polluted. The Schuylkill river also contains at times much coal 4ust vyasj;§4 ^97711 f ronj tjie coal regiot};?, 86 With these conditions and in order to keep the necessary fil- ter areas and the expense of filter cleaning as small as possible, it is economical to have either a sufficiency of subsiding basins, in which the turbid water can clarify itself before it is passed through the filters, or to have sufficient reservoir capacity for the filtered water, so as to allow pumping from the river to be suspended while the water runs excessively turbid. The former method is usually the better, as it is also the more common one. Filtered water should be consumed as quickly as possible after leaving the filters ; if it is to be subse- quently exposed in open reservoirs to the sun's rays and to the dust of the air, some of the benefits which are expected to be derived from filtration will be lost unless the reservoirs are covered. To answer your question regarding the best system of filtra- tion for Philadelphia, with some degree of definiteness under the conditions just stated, I am obliged to distinguish only be- tween two systems of filtration, namely the rapid and slow systems. These terms are used because the former contem- plates purifying water at a rate of 300 to 400 cubic feet, and the latter only from 5 to 10 cubic feet per day, per square feet of filtered surface. The former system represented by a variety of designs and embodied under the general term of ' ' mechanical filter, ' ' re- quires that a solution of alum or a simular coagulant be added to the water, in order to rapidly form a film which retains the fine suspended matter and the bacteria. The system of slow filtration, represented by large areas of sand beds and a variety of details, does not necessarily require such material to be added, and depends mainly upon the oxygen contained in the water, to oxidize the least stable particles of the organic mat- ter, and upon a natural film of organic matter, which soon coats the top layers of the filters, to retain the bacteria. After cleaning the filters in both systems, it is necessary to allow the water to run to waste until the respective films have formed again. As mechanical filters are of neces.sity washed more fre- quently, perhaps once a day, while the slow filters can run from one to two months, the less of water for cleaning the rapid filters is greater. Regarding the efficiency to remove bacteria, we have still but few data. The results of the Louisville experiments with mechanical filters, which will soon be published, I believe, are expected to throw considerable light on this subject. Mean- time, we are safe in stating that when in operation their prac- tical efficiency in this fespegt will not ^3^9?e4 \h^ ^fj^iency qf §7 *I*herefore, -with some uncertainties still on the side of the mechanical filters, with a bacterial efficiency no greater, and perhaps less than the slow filters, and with the larger quantity of water that must be used for washing, and must run to waste until the effective film has formed again, I consider that the slow filtration method requiring large beds of sand, the cost being the same, is the preferable one, with our present knowl- edge, for the waters under consideration. Preliminary settling basins are, no doubt, essential for the Schuylkill River water, and possibly also for the Delaware River water, which question may only be answered by experiment, and certainly only after thorough examination. Second : Have you information as to the kind and character of filtration adopted in Europe, and of what success? I have visited many of the European filter plants, and am familiar with their character and success. Briefly stated, they are all of the type which I term " slow filtration," and no city of any size in Europe has as yet used mechanical filters, al- though the city of Moscow is at present considering the expe- diency of adopting them. The success of the European filtration works, such as those of Ivondon, Liverpool, Hamburg, Altona, Berlin, Amsterdam. Rotterdam, Zurich and others, some of which purify badly pol- luted water, is undoubted, although it is admitted that not only the construction, but mainly the operation, must be conducted with considerable intelligence and care in order to obtain the desired results. Third : What .system of filtration is the best to remove bac- teria and disease germs? Awaiting the publication of the results of experiments made on this important subject in lyouisville, it is practicable now to state only that we have ample evidence from Europe that slow filtration removes almost all bacteria, including disease germs, from the water, and cannot say that mechanical filters, where they have been operated and examined with equal care, have done any better, and generally not as well. Fourth : What system of filtration is the most permanent, and in the end proves most satisfactory ? So far as durability is concerned, the slow filters are the more permanent, because they are built of masonry, while the me- chanical filters are built of iron or wood, and require more machinery than the others in operation. As to which system will in the end prove most satisfactory, depends upon the character of the water, the degree of its poUu- tion, and the care and intelligence with which the filters are designed and operated. Assuming the latter condition to be the same in both cases, then I can say that the more highly- polluted the water, the greater is the satisfaction to be derived from slow filtration, because the reduced number of cleanings lessens the chances of careless operation, and the opportunities for the passage and escape of disease germs. Water that is not polluted with sewage can, in my opinion, be filtered with equal satisfaction by mechanical filters. Fifth : Have you any data as to the cost of mechanical and sand filters ? I am in possession of such data as they have been available to the profession. Reducing the cost to a convenient unit, namely for filtering 1,000,000 gallons of water, these data show that, exclusive of land, and under favorable conditions, a large plant of open sand filters, operated at the rate of 2,000,000 gals, per acre, per day, costs from $20,000 to $30,000 per 1,000,000 gallons. The cost of operating a large plant, exclusive of in- terest, etc., ranges from $4.00 to $6.00 per 1,000,000 gallons. The cost of mechanical filters for a plant comparable with the above should be between $15,000 to $25,000 per 1,000,000 gallons, and the cost of operating, exclusive of interest, etc. , also from $4.00 to $6.00 per 1,000,000 gallons. The cost, however, depends so much on local conditions that no general statement can be made which has more than general interest. Under average conditions we may say that the cost of slow and rapid filters is nearly the same. There are special conditions, however, where one or the other system may be cheaper. Only an investigation of each particular case will give the basis for a comparative estimate of cost. Very truly yours, (Signed) Rudolph Herring. j^";*