^^^iM^si^i^K 
 
 ^ 
 
€mmll Hmrmitg Jitatig 
 
 THE GIFT OF 
 
 ..^....S. ^ju^mJc. 9j. '0.{ytjuunA«x>'»X^u... 
 
 LiL4!^..i>s,.z,o. ^?.Mjja: 
 
 975S-I • 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924022884963 
 
' EEPRESENTAT 
 M Session. ' t" 1 No. 347. 
 
 61sT Congress, I HOUSE OF REPRESENTATIVES. I Document 
 
 5L^uLO^^ J)^ 
 
 PRELIMINARY INVESTIGATIONS AND SURVEYS FOR INCREASING 
 THE WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 LETTER 
 
 THE SECRETARY OF WAR, 
 
 TRANSMITTING, 
 
 WITH A LETTEB FROM THE CHIEF OF ENGINEEBS, BEPOUT ON 
 PRELrmiNARY INVESTIGATIONS AND SIJIIVEYS FOB, INCBEAS- 
 ING THE WATEB SUPPLY OF THE DISTRICT OF COLUMBIA, AS 
 CALLED FOB BY THE ACT OF CONGBESS APPBOVED MAY 26, 
 1908, MAKING APPBOPBIATIONS FOB THE DISTBICT OF COLUM- 
 BIA FOR THE FISCAL YEAB ENDING JUNE 30, 1909. 
 
 December 15, 1909. — Referred to the Committee on the District of Columbia and 
 ordered to be printed, with accompanying illustrations. 
 
 War Department, 
 Washington, Decemier IS, 1909. 
 Sir: I have the honor to transmit herewith a letter from the Chief 
 of Engineers, U. S. Army, dated October 1, 1909, together with copy 
 of report, with illustrations, from Maj. Jay J. Morrow, Corps of Engi- 
 neers, dated July 8, 1909, upon preliminary investigations and sur- 
 veys for increasing the water supply of the District of Columbia, 
 made by him in compUance with the provisions of the act of Congress 
 approved May 26, 1908, making appropriations for the District of 
 Columbia for the fiscal year endmg June 30, 1909. 
 Very respectfully, 
 
 J. M. Dickinson, 
 
 Secretary of War. 
 
 The Speaker of the House of Representatives. 
 
 War Department, 
 Office of the Chief of Engineers, 
 
 Washington, October 1, 1909. 
 Sir: I have the honor to submit herewith for transmission to 
 Congress a report of July 8, 1909, with illustrations, by Maj. Jay J. 
 Morrow, Corps of Engineers, on the preliminary investigations and 
 
2 WATER SUPPLY OF THE DISTEICT OF COLUMBIA. 
 
 surveys for increasing the water supply of the District of Columbia, 
 as called for by the act of Congress approved May 26, 1908, making 
 appropriations for the District of Columbia for the fiscal year ending 
 June 30, 1909. 
 
 I concur in the opinion of the District officer. 
 Very respectfully, 
 
 W. L. Marshall, 
 Chief of Engineers, U. S. Army. 
 The Secretary of War. 
 
 PRELIMINARY INVESTIGATIONS AND SURVEYS FOR INCREASING THE 
 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 United States Engineer Office, 
 
 Washington, D. C, July 8, 1909. 
 Sir: I have the honor to submit the following report, in com- 
 pliance with instructions contained in letter dated Office of the Chief 
 of Engineers, U. S. Army, Washington, June 5, 1908, assigning to 
 my predecessor in this office the duty of making the preliminary 
 investigations and surveys for increasing the water supply of the 
 District of Columbia, as called for by the act making appropriations 
 for the District of Columbia for the fiscal year ending June 30, 1909, 
 approved May 26, 1908: 
 
 DESCRIPTION OF PRESENT SYSTEM. 
 
 The present plant for supplying the city of Washington with water 
 consists of a diversion dam in the Potomac River at Great Falls, a 
 gravity conduit from Great Falls to Dalecarlia reservoir, a gravity 
 conduit from Dalecarlia reservoir to Georgetown reservoir, a pressure 
 conduit from Georgetown reservoir to McMillan Park reservoir, and a 
 pumping plant at McMillan Park reservoir, lifting the water through 
 a height of from 20 to 30 feet to the filter beds. It is then carried by 
 gravity mains to the District's pumping station, on Bryant street, 
 whence it is disposed of by the officials of the District of Columbia, 
 in part by pumping for higher services and in part by gravity ia 
 lower services. 
 
 The project for the first work originated with Capt. M. C. Meigs, 
 Corps of Engineers, U. S. Army, in 1852 and 1853, and the first 
 appropriation was made in the act approved March 3, 1853, appro- 
 priating $100,000. Under this project it was proposed to build a 
 dam at Great Falls, a receiving reservoir at Dalecarlia, and a dis- 
 tributing reservoir on the heights beyond Georgetown. The neces- 
 sary legal right was given to the Federal Government by act of 
 legislature of the State of Maryland, in May, 1853. The aqueduct, 
 according to this project, was so far completed in 1859 that the water 
 from the receiving reservoir was available and turned into the mains, 
 utilizing water only from Little Falls branch. From the distributing 
 reservoir one 30-inch and one 12-inch main were projected, being 
 considered ample for the supply for some years after the completion 
 of the aqueduct. As the demand increased a 36-inch mam was 
 added in 1873 and a 48-inch main in 1890. The conduit was com- 
 pleted and the water supplied from the Potomac in December, 1863 
 
WATEE SUPPLY OP THE DISTRICT OP COLUMBIA. 6 
 
 since which date the syst3m has been in successful operation. In 
 1886 the dam at-Great Falls, which had been a riprap dam, extending 
 across the Maryland channel only, was replaced by a masonry 
 structure which covered the entire channel, its top at a level of 148 
 feet above mean tide water. Finally, in 1896, this dam was raised 
 2^ feet. In the year 1901 there was completed a tunnel leading from 
 the Georgetown reservoir to McMillan Park reservoir, situated 
 between Howard University and Soldiers' Home, which reservoir 
 was also completed in 1901. During the years 1903 to 1905 the 
 filter beds at McMillan Park were constructed, which provide now 
 for filtering the entire water supply of the^ District. 
 
 The conduit, in a general way, consists of about 15 miles of conduit 
 and tunnel and flow through 3 reservoirs, 2 of which are provided 
 with by-conduits. The conduit itself is a 9-foot tube, for the most 
 part of brick masonry, but partly in its rock tunnel sections unlined, 
 the unlined sections aggregating 3,700 linear feet, which, of course, 
 could only be lined with extreme difficulty until a new conduit is built. 
 There are on the line of the tunnel 26 culverts, nearly all in made 
 ground, the failure of any one of which would break the conduit for 
 a short period. 
 
 CAPACITY OF THE CONDUIT. 
 
 In 1895-97 a very careful study of the flow in the aqueduct was 
 made by Capt. (now Lieut. Col.) David DuB. Gaillard, Corps of 
 Eagineers, U. S. Army, the details of which are reported in full in 
 the Annual Report of theChief of Engineers for 1897. Brieflystated, 
 he found the ultimate capacity of the system at the lowest stages of 
 the Potomac River to be about 76,500,000 gallons per day, witn the 
 water in the Georgetown reservoir standing at an elevation of 144. 
 At that time the Georgetown reservoir was the distributing reservoir — 
 a function now filled by McMillan Park reservoir — and Captain Gail- 
 lard's studies did not include any consideration of the maximum 
 quantity the aqueduct would deliver with the water in the George- 
 town reservoir at a lower level. A careful study has been made in 
 the progress of the investigations just completed as to the present 
 capacity of the aqueduct system, the details, of which are given in 
 the appendices. It suffices here to state that with levels established 
 approximately as follows : 
 
 Manhole No. 1 (near Great Falls intake) 150. 5 
 
 Dalecarlia reservoir 142. 6 
 
 Georgetown reservoir 140. 25 
 
 McMillan Park reservoir 137. 5 
 
 a supply of 87,500,000 gallons per day was obtained, which would 
 seem to justify the conclusion that the aqueduct system as it now 
 stands will discharge approximately 90,000,000 gallons per dav as a 
 maximum, but that this high rate of discharge can be obtained only 
 by the drawing down of the reservoirs to such an extent as to practi- 
 cally exhaust them as storage basins. The aqueduct can, however, 
 be operated with a mean discharge of 80,000,000 gallons per day or 
 more, leaving the reservoirs available for reserve supply. 
 
 SCOPE OF PRESENT INVESTIGATIONS. 
 
 In accordance with project approved by the Chief of Engineers; 
 U. S. Army, June 17, 1908, to Mr. F. F. Longley, assistant superin- 
 tendent of the filtration plant, was assigned the duty of m9,king the 
 
4 WATEB SUPPLY OF THE DISTEICT OF COLUMBIA. 
 
 necessary surveys, detailed studies, and plans, under the direction o f 
 this o lice, to carry into effect the law. Following these instructions, 
 Mr. Lon^ley submitted to this o.fice a report, which is attached hereto 
 in full with all appandicss and which covers in greater or less detail 
 the entire field, and gives studies into the advantages and disadvan- 
 tages of the following projects: 
 
 First. An additional supply from the Potomac Eiver, to be taken 
 from the river at Great Falls, to be carried by gravity through a new 
 conduit to McMillan Park reservoir, and then pumped to the present 
 filter beds, in all essentials a system differing only shghtly from the 
 present supply. 
 
 Second. An additional supply from the Potomac River ^ to be taken 
 from the river at Little Falls, to be delivered by pumping into Dale- 
 carlia reservoir and carried thence by gravity through a new conduit 
 to McMillan Park reservoir, and then pumped to the present filter 
 beds, the construction between Dalecarlia and McMillan Park reser- 
 voirs being the same as proposed in the first project. 
 
 Third. An additional supply from the Potomac River, to be taken 
 from the river at some point south of Aqueduct Bridge, to be delivered 
 by pumping to McMjUan Park reservoir, and then pumped to the 
 present filter beds. 
 
 Fourth. A supply from Rock Creek, to be taken from the creek at 
 a point just north of military road, and either delivered by conduit 
 to the filter beds (without necessity of pumping from McMillan Park 
 reservoir) or purified by new filter beds located in Rock Creek valley 
 at the dam site, and delivering into the city mains at the elevation of 
 the present flow line of the first high service. 
 
 Fifth. A supply from Seneca Creek, which flows into the Potomac 
 River about 8| miles above. Great Falls, to be impounded in the 
 creek at an elevation with sufficient head to allow for purification 
 in additional filters and delivery by gravity into the city mains at an 
 elevation at least as high as the present flow line of the first high 
 service. 
 
 Sixth." A supply from ground- water sources, to be obtained by 
 sinking deep wells to water-bearing strata. 
 
 Seventh. The suggested popular plan of a supply from the upper 
 branches of the Potomac River was also briefiy considered, the water 
 to be taken from the river at a point over 100 miles above the city 
 and delivered by a gravity conduit to the filter plant at McMillan 
 Park. 
 
 As the work of Mr. Longley was nearing completion this office 
 called into consultation Mr. Allen Hazen, consulting civil engineer, of 
 great experience in viork of this nature, to review the work done and 
 suggest such modifications as to him might seem proper. His final 
 report w as submitted on June 28 and is forwarded nerewith. 
 
 In addition to the projects studied by Mr. Longley, the project for 
 obtaining a supply from latuxent Kiver, Maryland, a valley a few 
 miles north of the District of Columbia, to be taken from the river 
 at a point near the crossing of the Columbia turnpike and delivered 
 by pipe line and conduit by gra\ity to the filter beds at McMillan 
 1 ark, w ithout passing through McMillan P ark reservoir, was suggested 
 in May by Mr. Hazen, and with the little time and funds at command 
 it was briefly examined. 
 
WATEB SUPPLY OF THE DISTEICT OF COLUMBIA. 
 
 In addition to a study of these projects, studies were also made of 
 the following methods for providing for increasing the capacity of the 
 present conduit: 
 
 First. By raising the crest of the dam at Great Falls, by either 
 permanent construction, to an additional height of 2^ feet, or by a 
 raising only at periods of minimum discharge to an additional height 
 of not to exceed^2 feet by the use of flashboards, designed to be carried 
 away at freshet stages. 
 
 Second. By the construction of a pumping station at Dalecarlia 
 reservoir, to pump the water not exceedmg 6 feet in height, thus 
 increasing the hydrauUc gradient with a resultant slight increase in 
 capacity of the present conduit. 
 
 Besides these studies looking toward an increase in the capacity of 
 present conduit, a study was made of accomplishing much the same 
 result by cutting off largely the waste of water by a complete and 
 thorough application of meters to all services. 
 
 A study was also made looking toward the providing of additional 
 storage for thfe water supply by the construction of a new reservoir at 
 some point on the old or proposed new conduit. T]jis study also 
 included consideration of the necessity of preliminary treatment of 
 the water by sedimentation or coagulation in both the present and 
 the proposed new works. 
 
 All these matters are presented in detail in the reports of Mr. Long- 
 ley and Mr. Hazen, hereto appended, with drawings. 
 
 The projects for increasing the water supply by any of the methods 
 studied excepting by a gravity supply from Great Falls and by a 
 gravity supply from the Patuxent River (the first and last of the pro- 
 jects previously enumerated) can be dismissed from consideration for 
 reasons fully detailed in Mr. Longley's report. Briefly given, the 
 second (pumping from Little Falls), the third (pumping from Wash- 
 ington Harbor), the fifth (gravity supply from Seneca Creek), and the 
 seventh (supply from upper branches of Potomac) are rejected as 
 showing considerably greater cost without compensating advantages. 
 The fourth (supply from Rock Creek) is rejected because of insuffi- 
 ciency of supply, prohibitive value of real estate, and the opposition 
 that would doubtless arise to the submergence of so large an area in 
 Rock Creek Park. The sixth (ground water supplies) is rejected 
 becaitse it is doubtful if a sufficient supply could be so obtained, and 
 practically certain that if it could be obtained, it would be uneconom- 
 ical as compa,red with the other plans. 
 
 PROJECTS FOR INCREASING SUPPLY. 
 
 Of the projects for increasing the supply but two remain for favorable 
 consideration — the first contemplating additional supply from Poto- 
 mac River to be taken from the river at Great Falls, to be carried by 
 gravity through the new conduit to McMillan Park reservoir and then 
 pumped to the present filter beds. The Potomac River at Great 
 Falls has a minimum flow approximately ten times as great as thie 
 present consumption of the water in the District, and the United 
 States has a right to a large enough portion of this flow to supply more 
 than twice the present consumption. The existing aqueduct dam 
 holds the water at a level suitable for the supply of this additional 
 
6 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 quantity. For the most part the right of way of the old aqueduct is 
 wide enough for the location of a new line within its limits. 
 
 The general route followed by, the old aqueduct is the only surface 
 route at the proper grade between Great Falls and Georgetown 
 reservoir. It would be entirely feasible to locate another aqueduct 
 parallel with the first to run almost its whole length. There can be 
 no question but that an alternative route should be selected if feasible, 
 in order to reduce to a minimum the chance of a single accident or hos- 
 tile military operation disabling both conduits. Careful -studies have 
 been made of all possible divergent routes between the same two ter- 
 mini . Between Dalecarlia and McMillan Park reservoirs a new line has 
 been found which could be recommended for economical reasons alone, 
 being shorter than the old route, and possessing the great advantage 
 of having all its tunnel sections at grade. But above Dalecarha 
 reservoir the separation of the two lines offers extreme difficulties. 
 Above Cabin Johns a location can be found that would be shorter, 
 but it would be considerably more expensive, being largely in tunnel. 
 From Cabin Johns Bridge south a separate line would have to be not 
 only wholly in tunnel but materially longer. 
 
 The proposed location of the parallel route is described as follows : 
 From Great Falls to a point near the Anglers' Club at the head of 
 the Conduit road, by a straight grade tunnel; thence on the south 
 side of the old aqueduct and parallel to it to the north end of Dalecarlia 
 reservoir, a cut-and-cover aqueduct, with one short tunnel; a by- 
 conduit skirting the south shore of the reservoir and crossing at the 
 narrowest point to the east end ; thence a tunnel almost due east to 
 Connecticut avenue extended; thence cut-and-cover aqueduct along 
 the side of the valleys of the branch just north of Linnean Hill and of 
 Piney Branch, to Fourteenth street; and thence a grade tunnel 
 mainly under Spring road and Warder street to the north end of 
 McMillan Park reservoir. The total length of this line, including the 
 by-conduit around Dalecarlia reservoir, is 78,120 feet, of which 30,440 
 feet is in grade tunnel. 
 
 Its estimated cost, including the land, intake and outlet gate- 
 houses, coagulating equipment for future installation, masonry 
 aqueduct bridge in the park across Rock Creek, and all engineering 
 expenses and contingencies, is placed at $3,879,383. 
 
 The alternative location, which might be designated as the tunnel 
 location, follows the same route as the parallel location from Cabin 
 Johns Bridge to McMillan Park reservoir, but above Cabin Johns 
 diverges from the first, running directly from Great Falls to a point 
 on the west side of the valley near Cabin Johns Bridge in grade 
 tunnel; thence in a short section of cut-and-cover aqueduct and a 
 siphon crossing to the east end of the valley, where it joins the line 
 just described. 
 
 The estimated cost for this project, including the land, intake and 
 gatehouses, coagulating plant for future installation, masonry 
 aqueduct bridge m the park across Rock Creek, and all engineering 
 expenses and contingencies, is placed at $4,333,686. 
 
 The details of these projects have been carefully worked out as 
 thoroughly as is believed to be necessary at the present time, and are 
 fully explained m the appendixes to this report, and detailed in 
 accompanying tracings. The principal structures are a combined 
 intake and gatehouse at Great Falls with a 48-inch drain leading to 
 
WATER SUPPLY OF THE DISTRICT OF COLUMBIA. / 
 
 a point below the falls, which will provide for drainage at the upper 
 end of not only the old but also the new aqueduct ; a construction to 
 provide for the crossing of the old aqueduct near the Anglers' Club; 
 the crossing of Cabin Johns valley by an inverted siphon some 70 
 feet below the hydrauhc grade line on a concrete masonry bridge 
 immediately over the stream; a gatehouse at the inlet of Dalecarha 
 reservoir; .another crossing of the old aqueduct in this gatehouse; a 
 48-inch drain from this gatehouse to the old diversion tunnel of 
 Little Falls branch; a by-conduit around Dalecarlia reservoir, cross- 
 ing the reservoir at the narrow neck near the south end, in a rein- 
 forced concrete section supported on piers and arches of concrete 
 masonry; an outlet gatehouse at Dalecarlia reservoir; a masonry 
 bridge across Rock Creek valley in the park, by an inverted siphon 
 some 35 feet below the hydrauhc grade line, the bridge to be of suitable 
 width for a highway, to provide a high-level crossing of the valley, 
 with suitable architectural treatment demanded by these surround- 
 ings; a 48-inch blow-off for drainage to be provided at this crossing; 
 and a gatehouse at the north end of McMillan Park reservoir, with a 
 small pumping equipment to provide for the final drainage of this 
 section, which can not be economically taken care of by gravity 
 drainage. 
 
 The principal structures on the tunnel location differ from those 
 just described by the eUmination of aU constructions between the gate- 
 house at Great Falls and the crossing of Cabin Johns valley; the 
 crossing of the valley would be very similar to that provided for the 
 other location, but the structure would be located immediately under 
 the present Cabin Johns Bridge, and at this point the old and the 
 new aqueducts would cross each other. 
 
 It should be noted that the first-described route would fit in well 
 with the additional storage hereinafter described, should such be 
 deemed advisable, at the Stubblefield site. The proposed Stubble- 
 field reservoir would lie immediately alongside both aqueducts, and 
 could be conveniently connected with both through gatehouses, thus 
 providing for storage of not only the new but also the old supply. 
 This site could not, however, be utilized for the tunnel location, as the 
 tunnel takes a more northerly route and is far removed from the Stub- 
 blefield reservoir site. The scheme, however, for storage in the 
 reservoir at Cabin Johns valley could easily be combined with the 
 tunnel location, but it possesses defects wmch have elsewhere been 
 described. These facts are mentioned as showing that the location 
 of the conduit paralleling the present conduit from the Anglers' Club 
 to Cabin Johns Bridge lends itself much more easily to the extension 
 of the reservoir system than does the tunnel location. The Stubble- 
 field reservoir could still be constructed, however, if deemed advis- 
 able, but storage in it would have to be restricted to water supplied 
 through the old conduit. 
 
 PATUXENT RIVER PROJECT. 
 
 The second project remaining for favorable consideration contem- 
 plates the utiUzation of the Patuxent River, Maryland, as a source of 
 supply. Consideration of this supply was suggested by Mr. Hazen 
 and was considered near the end of the investigations covered in Mr. 
 Longley's report. The project, however, has many features which 
 
8 WATEE SUPPLY OF THE DISTRICT OP COLUMBIA. 
 
 might demand its ultimate selection. Briefly outlined the project 
 would assume the following form : 
 
 A masonry dam 70 or 80 feet high and possibly 800 feet long, 
 founded upon rock, would be thrown across Patuxent Eiver at some 
 point of the gorge near the Columbia turnpike; this would form a 
 reservoir over the Patuxent Valley of a length measured along the 
 valley of about 9 miles, with a number of considerable branches on 
 tributary streams, and would provide for storage sufficient to pro- 
 vide not less than 100,000,000 gallons of water per day, even during 
 most unfavorable seasons. The elevation of the reservoir would he 
 sufficient to justify the use of a smaller conduit for the same capacity 
 than would be possible with the small head obtained at Great Falls, 
 and at the same time the water could be delivered to the filters at 
 McMillan Park by gravity, rendering unnecessary the pimipage 
 which would be required for a supply from Great Falls. A tunnel 
 approximately 3 miles long, through the ridge between the Patuxent 
 Valley and the valley of the Paint Branch of the Anacostia would 
 bring the water to a point from which it could be led to the filters by 
 a direct and easy route for a steel pipe conduit, it being probable that 
 a 6-foot conduit would suffice for more than doubling the present sup- 
 ply. The two great advantages of such a supply would be, first, the 
 great advantage of having the two means of supply entirely separated, 
 and, second, a saving of the necessity of pumping to reach the filter 
 beds. Information at hand is not sufficient to definitely determine 
 whether these advantages are sufficient to overbalance the known dis- 
 advantages, which are the construction of a dam, the purchase of 
 large tracts which would be flooded in the creation of the reservoir, 
 the diversion of sewage from some of the villages lying within the 
 drainage area and near to the reservoir, provisions necessary to be 
 taken to raise various roads and build bridges for them, the necessity 
 to first obtain legal authority of the State of Maryland to utilize the 
 water and condemn the necessary lands, and consideration of claims 
 for lost water rights which might arise from riparian owners below 
 the stream. It is, however, believed that the project is sufficiently 
 promising to justify a careful study, which can be undertaken before 
 the necessity arises for the increase in the supply. It is believed that 
 this study should include a rough topographic survey of the valley, a 
 study of the extent and value of the land necessary to be condemned, 
 and an estimate of the cost of all structures required, together with a 
 study of the comparison of the quality of the water to be obtained 
 with the quality of Potomac water, the degree of sedimentation the 
 reservoir may provide, -and the necessity for preliminary treatment 
 should this not be sufficient. It is estimated that these studies could 
 be made at a cost of 13,000, and that an appropriation therefor should 
 be made at a reasonably early date, as it should be borne in mind 
 that should this project be adopted for ultimate construction the 
 necessarv legal preliminaries would add possibly two years to its time 
 of completion over the time necessary to complete the work of pro- 
 viding a new conduit from Great Falls. - 
 
 INCREASING CAPACITY OF PRESENT CONDUIT. 
 
 Referring to the additional studies made it is believed that any 
 considerable raising of the crest of the dam at Great Falls is inadvis- 
 able. The two serious objections to this plan are the increase of 
 
WATER SUPPLY OP THE DISTRICT OF COLUMBIA. y 
 
 pressure on the interior of the present conduit, and the raising of the 
 water level above the dam, especially at flood stages. It is estimated 
 that the capacity of the conduit would be increased about .10 per cent 
 with an addition to its height of 2^ feet, and any greater increase is 
 believed to be ill-advised. To effect this permanent increase in the 
 height of the dam is believed to effect an increase in the supply in- 
 commensurate with the cost. The alternative plan, suggested by 
 Mr. Hazen, of providing flashboards to be installed at low stages, 
 raising the level of the pool above the dam by a height not to exceed 
 2 feet, would, however, be comparatively inexpiensive, would result 
 in no greater pressure on the conduit than it now frequently carries 
 at medium high stages, and would not operate to raise the upper 
 levels during flood stages, the flashboards being installed with a view 
 to beihg swept away at high stages, the levels in the pool at flood 
 stages being regulated by the permanent dam, as at present. It is 
 estimated that an increase in the capacity of the present conduit of 
 about 8,000,000 gallons per day could be effected ytith perfect safety 
 and at small expense by this method. A maintenance charge would, 
 however, be required, to supply an entirely new equipment of flash- 
 boards for each considerable freshet. 
 
 The study as to the insertion of a booster pumping station at 
 Dalecarlia has shown that the plan would have less etl'ect on increas- 
 ing the capacity of the present system than it would have on regulat- 
 ing the levels of the three existing reservoirs, all of which lie below the 
 site of the proposed pumping station. This is due to the fact that the 
 pumping station would have no effect on increasiag the capacity of 
 the conduit above the station except such slight effect as would result 
 from a drawing down of the level for a short distance back from the 
 . station, the effect of a steeper gradient beiug quickly nullified by the 
 decreased capacity of the portion of the aqueduct which would be 
 flowing only partly full. The principal value, however, of such a 
 station would be in regulating the levels of the reservoirs to provide 
 for the excessive consumption which is demanded on rare occasions, 
 and these advantages might at some future date justify the construc- 
 tion of this plant. Its cost as estimated by Mr. Longley, with the 
 necessary machinery, is $65,000, with an annual maintenance change 
 estimated at $3,000. Mr. Hazen, however, suggests a modification 
 in Mr. Longley's plan, which would considerably reduce the estimate 
 for first installation. 
 
 INSURANCE AGAINST ACCIDENT TO CONDUIT. 
 
 A study as to the necessity for providing additional supply from the 
 viewpoint of insurance against risk of failure of the present aqueduct 
 system shows the following conditions to be found: 
 
 In the unlined sections, which aggregate nearly 4,000 feet, seams 
 varying in thickness up to 4 inches, and some disintegration of the 
 surface, have been found, so that at every cleaning fragments of rock 
 are found, which are broken up and carried out of the tunnel. These 
 sections are inspected on every occasion when the conduit is drained, 
 and have all been personally inspected by me. Ihe worst points can 
 and will be gradually lined, the work being done with considerable 
 difficulty, due to the short and infrequent periods available. It is 
 not believed there is any danger of a fall of rock large enough to cause 
 
10 WATER SUPPLY OP THE DISTRICT OF COLUMBIA. 
 
 an interruption of the water supply, but what danger there is can not 
 be remedied until these unlined sections can be lined. 
 
 On the Ime of the aqueduct there are twenty-six culverts over 
 streams of greater or less size. Over each culvert the masonry tube 
 is built upon filled ground. Certain portions of the aqueduct, resting 
 on some of these artificial embankments, have become considerably 
 deformed in their fifty years of service, and it is doubtless due only 
 to the extreme foresight of Captain Meigs that these deformations 
 have not resulted in the failure of the aqueduct at several of these 
 points during its earlier years of service. It seems now, however, 
 probable that all such deformations occurred soon after the aqueduct 
 was built and that they have stood practically unchanged for many 
 years. It is certain that practically all of those that are now in evi- 
 dence were noted in the examination heretofore referred to, made by 
 Captain Gaillard twelve years ago. This office has now instituted the 
 practice of a careful and detailed examination of all critical sections 
 upon every occasion when the aqueduct is drained, which occurs on 
 an average of three or four times per year, and it is believed that no 
 serious threat exists at these points, especially as, the conduit being 
 under pressure of not above 2 or 3 feet head, any such washout could 
 be restored by a temporary flume on a timber trestle within the stor- 
 -age period provided for by the reservoirs, leaving the permanent con- 
 struction to be subsequently undertaken. 
 
 It is believed that the consideration of the stability of the struc- 
 tures of the old aqueduct system fails to show any defect or weak- 
 ness which threatens to interrupt the water supply of the District of 
 Columbia to a sufficient extent to warrant the expenditure of the 
 large amount necessary to provide a complete new aqueduct system 
 before such new works would be required by reason of the inadequacy 
 of the supply. 
 
 ADDITIONAL STORAGE. 
 
 The study as to the necessity for providing for additional storage 
 capacity is so intimately related to that for preliminary treatment 
 by sedimentation or coagulation that the two can best be handled 
 as a single study. The existing aqueduct system includes three res- 
 ervoirs (Dalecarlia, Georgetown, and McMillan Park) with a total 
 storage area of about 125 acres and a total storage capacity from 
 their bottoms to their flood lines of approximately 600,000,000 gal- 
 lons. As the reservoirs can not be filled to their flood lines when the 
 aqueduct is being used to anything approaching its full capacity, 
 and as they can not be drawn to their bottoms, only a part of tms 
 total, estimated at 300,000,000 gallons, can be utilized. This quan- 
 tity provides for a period of effectual storage for the present average 
 consumption of about five days. These reservoirs provide for a 
 natural purification by sedimentation of considerable value not only 
 in the removal of turbidity but in the attendant reduction of bacteria. 
 The degree of natural purification is, of course, reduced by any 
 increase in the rate of flow through the reservoirs ; but, more impor- 
 tant, the degree of natural purification will also be reduced as the 
 draft on the sj^stem increases, by the necessity of keeping the con- 
 duit open during periods of excessive turbidity in the Potomac 
 water at times when, with the present consumption, the conduit can 
 be kept safely closed. 
 
WATEB SUPPLY OF THE DISTEICT OF COLUMBIA. 11 
 
 In the two projects remaining for favorable consideration, viz, 
 a gravity suppfy from the Potomac at Great Falls to McMillan Park 
 reservoir and a gravity supply from Patuxent Kiver direct to the 
 filter beds, two sites have been considered, either of which could be^ 
 adapted for use as a reservoir of large capacity; both of these are* 
 on the line of the first-mentioned conduit. The necessity of a reser- 
 voir other than the one created in the Patuxent Valley would prob- 
 ably not be necessary in case that route were adopted, as the reser- 
 voir itself would provide large storage, with attendant piiiification, 
 and the use of the water from that supply would make it possible 
 to exclude the turbid water of the Potomac, ordinarily high in bac- 
 teria in periods of excessive turbidity, for longer lengths of time than 
 is possible under the present operation of the water-supply system. 
 The two reservoir sites investigated are, respectively, in Cabin Johns 
 Valley and the site known as the Stubblefield site, 9, comparatively 
 flat area of about 180 acres, Ijing between the old aqueduct and the 
 Chesapeake and Ohio Canal, about 4 miles below Great Falls. The 
 former of these two sites — the valley of Cabin Johns Creek — could 
 be converted into a reservoir by the construction of a dam immedi- 
 ately above Cabin Johns Bridge. Unless it were intended to utilize 
 the water of Cabin Johns Creek for the supply, a condition which 
 would become less and less desirable with the growth of the popula- 
 tion of the District of Columbia, expensive diversion works would be 
 required to conduct and to carry flood discharge of this stream around 
 the reservoirs. The size of a possible freshet flow of 4,000 cubic 
 feet per second seriously complicates the plans for its diversion and 
 the expense woidd be absolutely prohibitive. The project, how- 
 ever, has been given a further study and a detailed scheme worked 
 out, providing for the storage and subsequent diversion of ordinary 
 freshets — a scheme which at very infrequent intervals would fail of 
 total exclusion of the waters during high and prolonged floods, but 
 which brings the cost within figures which render the work practi- 
 cable. 
 
 The study of the Stubblefield site shows that it would be possible 
 to construct at that point an artificial reservoir, not complicated with 
 any questions of floods from feeding streams, with a maximum avail- 
 able storage capacity of about 425,000,000 gallons, and a total ca- 
 pacity of 850,000,000 gallons, at an estimated total cost of $1,380,588. 
 As this cost figiu*e is considerably less than the figure of cost of the 
 Cabin John site the details of the latter project are not presented 
 with this report, but are filed in this office for futiu'e reference. The 
 details of the Stubblefield reservoir site are, however, submitted with 
 this report. 
 
 Additional storage besides the apparent advantage of providing 
 a reserve available for supply in the event of an interruption of the 
 flow in the aqueduct above the point of storage, and the equally 
 apparent advantage of providing a desirable flexibility to the water 
 system for an occasional peak in the draft, due to conditions of use 
 above the average capacity of the supply, possesses a very important 
 advantage, hot so apparent to the layman, of improving the quality 
 of the water physically and bacteriologically, both by increasing the 
 period of sedimentation and lengthening the periods during which 
 undesirable water may be excluded from the conduit. It might 
 well be that a system designed to attain these first two objects should 
 
12 WATER SUPPLY OF THE DISTEICT OF COLUMBIA. 
 
 fail in the third, and in such a system it might be discovered that 
 another means could be found to attain the third desired result, and 
 at a cost which would commend it over the project to provide for 
 more storage. This is believed to be actually the case in the Wash- 
 "ington supply. 
 
 NECESSITY OF IMPROVING CHARACTER OF RAW WATER. 
 
 The operation of the filtration plant was started in October, 1905, 
 and has provided a most satisfactory effluent, so much so that it 
 may be assumed that any additional supply that may be provided 
 will have to be filtered. The (question naturally arises as to whether 
 the present plant will be sufficient for a large increase in the supply, 
 The filtration plant was designed for a rate of filtration of 3,000,000 
 gallons per acre per day. 
 
 At the time of its design it was in accordance with the most 
 approved practice. Experience since that time, however, has shown 
 that the plant can under favorable conditions of the raw water be 
 operated at a rate higher than 3,000,000 gallons per acre per day, 
 without any deterioration in the quality of the effluent, the increase 
 in the rate oeing abundantly siifficient to provide for a much greater 
 consumption than the District of Columbia will demand for a great 
 many years to come. There are, however, at the present time some 
 conditions which will require improvement in order to obtain such 
 results. In order to attain any such high rates of filtration it would 
 be necessary to supply the raw water to the filters in at least as fa- 
 vorable a condition as it is now applied under ordinarily favorable 
 conditions ; in other words, the water must be either excluded at the 
 inlet to the conduit during periods of excessive turbidity, or, at such 
 periods, a preliminary treatment must be applied. The exclusion 
 of the water is practically impossible without a large increase in the 
 storage. Experiments conducted by this office in recent years with 
 preliminary treatment by coagulation have shown this to be the 
 only method that will give unfailing results in this character of the 
 raw water during periods of such tiurbidity. This process possesses 
 great flexibilty and could be applied to whatever extent might be 
 indicated by the results of filtration operations. 
 
 Besides providing for the filtration of all water demanded by the 
 growth of the city to even twice its present consumption of water, 
 without any increase in the size of the filter beds, such a plant would 
 simplify enormously the present operation of the filters. The tur- 
 bidity of the water at the present time furnished, to the filter beds fre- 
 quently causes great reductions in the capacity of the plant, requir- 
 ing efforts lasting many days to furnish the quality of water de- 
 manded, and frequently without success. Coagulation will give with 
 certainty the objects desired by effecting the sudden and immediate 
 precipitation of a large percentage of bacteria in any quantity in 
 which they are likely to enter the system. It could, moreover, do 
 this with more effectiveness and more dependency than any system 
 of reservoirs feasible to construct as a part of the system. It would 
 be necessary to coagulate the water only during periods of high tur- 
 bidity and under conditions and during periods of greatly increased 
 consumption. 
 
WATEB SUPPLY OP THE DISTRICT OF COLUMBIA. IS 
 
 Even at the increased capacities required a coagulating plant could 
 be installed and the Georgetown reservoir remodsled in connection 
 therewith, at a cost of $130,000, with a maintenance and operating 
 cost, depending largely upon the frequency and duration of the periods 
 of excessive turbidity, which would probably not exceed $15,000 
 per year. Consideration of these facts, together with the fact that 
 the coagulating plant already recommend d by this office has re^* 
 ceived the unqualified approval of the Board of Consulting Engineers 
 and the indorsement of tne medical society, justifies the assumption 
 that coagulation may be considired an assured adjunct to the aque- 
 duct system and filtration plant in the future, and that after its 
 installation, which should be promptly provided for, there is no argu- 
 ment which demands the construction of an additional reservoir 
 in connection with either the present or an aqueduct system provid- 
 ing for a large increase in the supply. 
 
 WASTE PREVENTION. 
 
 A study looking to the reverse side of the question, viz, the propo- 
 sition to reduce largely the waste of the water, is carefully gone into 
 in the appendixes of this report, and justifies the statement that it 
 is possibly the most important part of the report. The present 
 maximum capacity -of the Washingtfon Aqueduct is believed to he 
 about 90,000,000 gallons per day. The present average consumption 
 demanded per day is slightly less than 60,000,000 gallons. Of this, 
 amount a large part is undoubtedly wasted, the amount wasted 
 being, however, considerably reduced within the past three years. 
 
 The abnormal consumption of water which occurred during the 
 years 1905 and 1906 is responsible for the introduction of the system 
 of metering private services, which was inaugurated three years ago. 
 The work has naw, however, been nearly halted by the lack of funds. 
 There are now, out of about 60,000 service connections, a little less, 
 than 13,000 meters in service, or about 21 per cent of all. The 
 funds are not more than sufficient to keep up with the new services, 
 so that there will probably continue to be in the neighborhood of 
 about 47,000 unmetered services in the District unless money is 
 provided for their installation. The cost of purchasing and install- 
 mg meters has been about $16.50 each, at which rate it would cost 
 about $775,000 to equip every private service with a meter. 
 
 In addition to this good start in the work of metering services, 
 the water department of the District of Columbia, about three 
 years ago commenced the work of its pitometer division, which 
 has been pursued continuously since that time in a survey of the 
 waste in the distribution system, and has produced excellent, even 
 remarkable, results. The work of the pitometer surveys has cov- 
 ered possibly two-thirds of the area of the District of Columbia, and 
 it is probable that the work over the other third will cut off as large 
 a waste in proportion to the area, and probably a little larger, after 
 which time the investigations are expected to be continued as long 
 as they show efficiency in the suppression of waste. The amount or 
 leakage detected and cut off by this service and by such meters as 
 have oeen installed in the two and one-half years of operation is 
 equal to nearly 12,000,000 gallons per day. For this worK it would 
 be impossible to give too much credit to tne efforts of Mr. W. A. Mac- 
 
1^ WATER SUPPLY OF THE DISTKICT OF COLUMBIA. 
 
 Farland, superintendent of the water department of the District 
 of Columbia. The combined result of these measures for suppression 
 of waste is shown markedly in the decrease in daily per capita con- 
 sumption since 1906 and in the fact that the average daily con- 
 sumption now is less than it has been at any time within the past 
 five years, despite the increase in house services followihg the growth 
 of the city. 
 
 A careful consideration of the data contained in this study, all 
 of which is detailed at length in the appendixes of this report, would 
 seem to indicate that a continuation, of the work of the pitometer 
 survey of the District water department, and the completion at an 
 early date of the work of installation of meters on house services, 
 would result in a cutting off of the unnecessary waste of water which 
 was occurring three years ago in the District of Columbia by not 
 less than 24,000,000 gallons per day. If this saving could have 
 been effected instantly, at the time of the beginning of measures of 
 waste preventions, the average daily consumption at that time would 
 have been approximately 135 gallons per day per capita, which is 
 the figure assumed as a reasonable average consumption per capita 
 for all necessary purposes, including a liberal allowance for waste 
 which can not reasonably be prevented. This same figure is deduced 
 independently by Mr. Longley iij his report, and seems to be entirely 
 reasonable and one which universal metering may hope to attain. 
 It is a curious fact that the figure deduced oy IVir. Longley, of 90 
 gallons per capita per day, to which is added a 50 per cent allowance 
 for unpreventable waste, is the exact total figure used by the original 
 designers of the Washington Aqueduct, and was placed at that 
 figure by them in excess of any allowance that had previously been 
 made for any public water supply. 
 
 The use of meters on house services does not imply a desire upon 
 the part of the officials to restrict even the lavish u'se of the water, 
 but merely to provide a means for placing upon the consumer the 
 burden of any extravagant or useless waste that may be found upon 
 his individual premises. The meters are intended more for the 
 detection of leaks in mains or service pipes on the premises, and are 
 intended to place upon the consumer the pecuniary responsibility for 
 any flagrant waste from faucets, hose, etc., or of any waste caused by 
 a deliberate neglect of any defective plumbing in the house. 
 
 In the studies made as to the consumption of water in the District 
 of Columbia it is necessary to carefully consider not only the mean 
 daily demand, but the question of peak loads, or excess in the rate of 
 consumption which may occur in periods of prolonged extreme cold 
 weather, and on occasions in the summer time due to periods of pro- 
 longed heat and drought. This maximum draft to meet such peaks 
 in cold weather has been as high as 45 per cent, but it is certain that 
 the means recommended for a restriction of waste should tend to 
 reduce this, the most serious, overdraft. There remains, however, to 
 be considered the summer peak, which represents an extra use of 
 water for cooling streets and watering lawns, which doubtless will 
 continue at more or less frequent intervals even after the installation 
 of the measures recommended. It is believed that the figure sug- 
 gested by Mr. Hazen, viz, 30 per cent above the annual average rate, 
 IS ample to cover these summer peaks, and that this figure is more 
 than ample to cover the overdraft which may occur during the win- 
 
WATEB SUPPLY OP THE DISTRICT OP COLUMBIA. 15 
 
 ter peaks, and that the figure assumed by Mr. Hazen, viz, an annual 
 average consumption of 65,000,000 gallons per day, can be safely 
 taken as the figure at which the average annual consumption per day 
 of the present system may be placed without any additions to the 
 system, except the universal installation of meters, the continuation 
 of the pitometer surveys, and the addition of the coagulating plant 
 for preliminary treatment of the water supply; and that this mean 
 average consumption provides for sufficient expansion in the system 
 to meet all unusual demands. 
 
 This figure, on the assumption of a reduction of the daily per capita 
 consumption to the figure of 135 gallons, would place the date at 
 which the present conduit will be insufficient to meet demands at the 
 date when the population of the District of Columbia shall have 
 reached 480,000. A careful study of Mr. Longley's data shows that 
 there appears no reason to believe that the mean per capita consump- 
 tion can not be reduced to this figure. 
 
 It can not be too insistently urged, however, that the probability 
 of having to provide only this quantity of water is contingent solely. 
 upon the reduction of waste, which can not be accomplished except 
 by the extension of the meter system to cover all services. It is sug- 
 gested that an appropriation to this end to provide the water depart- 
 ment with these funds should be made in the next District appro- 
 priation act, to provide for approximately one-third of the necessary 
 work in the fiscal year 1911, to be followed in each of the two suc- 
 ceeding years by an approximately equal appropriation, with the 
 provision that funds are to be repaid from the revenues of the water 
 department in equal annual installments covering a sufficiently long 
 period until completed. Should this be done, it is the belief of thi& 
 office that the necessity for providing for an increase in the volume of 
 water supplied to the District could be safely postponed for at least 
 twenty years. 
 
 The question of metering the supplies used in the buildings and 
 grounds owned by the United States has also been studied. The 
 Federal Government is a large consumer of water in the District of 
 Columbia, and it has been difficult, especially on the part of the Dis- 
 trict officials, to convince the federal officials having charge of the 
 various buildings and grounds that all extravagant waste could be 
 cut off. Not only have large leaks been frequently discovered, but 
 the use of the filtered water for operating condensing plants, even 
 where a plant had been installed which would enable Potomac water 
 pumped direct from the river at small expense to be used instead, has 
 been frequently protested against, but without avaU in all cases. 
 The argument which is frequently made is that the United States 
 furnishes water free to the District of Columbia, and the District of 
 Columbia should not object to any quantity of waste on the part of 
 the federal officials. It is believed that the necessity for metering of 
 all supplies to government buildings and grounds is equal to the 
 necessity of metering supplies for private consumers of the same 
 quantity of water; these meters should, however, be installed at the 
 expense of the United States, and preferably under the direction of 
 the officer in charge of the Washington Aqueduct, who should also be 
 charged with the duty of reading the meters and talcing such steps as 
 might be necessary to restrict waste. It is believed that this treat- 
 ment of the question would remove all chance of friction between the 
 
16 WATER SUPPLY OF THE DISTKICT OP COLUMBIA. 
 
 District officials and the federal officials, and would result in a con- 
 siderable saving in water and consequent lengthening ot the penod 
 at the end of which the supply will have to be increased. This office 
 has made estimates as to the cost of this installation, at $50,000, and 
 an item to eflfect this installation should be included in the District 
 appropriation act for the fiscal year 1911. 
 
 ft remains then to endeavor to forecast a date at which the pres- 
 ent water supply will be insufficient. Mr. Longley has made studies 
 as to the growth to be expected in the population of the District 
 of Columbia in the coming years, has entered on this study with 
 great care, and seems to have considered all factors which might 
 entsr. According to his studies it is deduced that the population of 
 the District of Columbia will reach the figure of 480,000 at a date not 
 far removed from the year 1930. It seems safe to accept this esti- 
 mate, although it may be that the rate of growth of the city will 
 increase to some extent in the coming years, as there seem to be indi- 
 cations justifying such an assumption. The date fixed is, however, 
 but about twenty years from date, and as there remain two important 
 factors of safety which at small expense and at short notice can be 
 made to increase to a considerable extent the capacity of the present 
 worl« it is believed that this office is safe in reporting that, with 
 these adjuncts, there will be no necessity for the completion of the 
 construction of a new conduit until about the date mentioned. The 
 two expedients referred to are those heretofore outlined, viz, the use 
 of temporary flashboards on the dam at Great Falls and the con- 
 struction of a booster pumping station at Dalecarlia reservoir. It is 
 believed that the first-mentioned expedient can be made to add to the 
 capacity of the aqueduct about 8,000,000 gallons per day, and that 
 the booster station at Dalecarlia, while not adding greatly to the 
 capacity during periods of normal consumption, would operate in 
 times of peak loads to provide for an increased flow in the aqueduct 
 below Dalecarlia reservoir, making it safe to reduce the allowance ot 
 30 per cent, suggested by Mr. Hazen, and previously referred to, as a 
 reserve supply to meet the excess demand during these periods of 
 overdraft. It is believed that these expedients would provide for the 
 safe management of the aqueduct to meet the demands that would 
 be made upon it by a population of about 530,000, which figure of 
 population, it is bslieved, can be safely predicted will not be reached 
 by the District of Columbia in less than twenty years' time. At the 
 risk of repetition it should be again stated that these figures depend 
 for their justification upon the early installation of the measures 
 herein recommended for the reduction of waste and for the prelimi- 
 nary treatment of the supply. 
 
 It should be stated here that upon the installation of all meters, 
 as recommended, and the completion of the wast6 survey, this 
 office will become possessed of data which will enable a more reliable 
 figure for the future per capita consumption to be deduced than 
 the office is now justified in assuming. 
 
 CONCLUSIONS. 
 
 As a result of these investigations the following conclusions are 
 arrived at: 
 
 First. The efforts of the water department of the District of Co- 
 lumbia of the last three years, looking to a suppression of waste, 
 
WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 17 
 
 have produced such remarkable results as to imperatively demand 
 their continuance and extension hj rapidly covering all services, 
 both private and pubUc, in the District of Columbia by meters. 
 
 Second. As an alternative to providing for the construction of an 
 additional storage reservoir, which will become more necessary as 
 the demand for water increases and to provide for the more effi- 
 cient working of the present filtration plant during periods of exces- 
 sive turbidity in Potomac water, the construction of a coagulating 
 plant and remodeling of Georgetown reservoir for use in connection 
 therewith is imperatively demanded as a necessary adjunct to the 
 present system, and should be at once begun. 
 
 Third. Contingent upon the adoption of these measures, there 
 can be no doubt that the consumption and waste of water in the' 
 District of Columbia in the next few years will be far enough belo\% 
 the safe working capacity of the existing aqueduct to make it safe 
 to postpone new construction until fuller and more definite results 
 are at hand to predict accurately the date at which such construc- 
 tion must be completed. 
 
 Fourth. There appears to be no justification for an increase in 
 the supply through fear of failure in any of the structures of the 
 existing system. 
 
 Fifth. When the necessity arises for an increase in the supply,; 
 there appear to be three projects, any of which would give a satis- 
 factory supply at least as great as the present supply and at reason- 
 able cost. From Great Falls an aqueduct with ail necessary ap- 
 purtenant structures could be constructed, with such slight varia- 
 tions from plans which have been worked out in considerable detail 
 as might be deemed necessary by advances in the art of providing; 
 
 Eublic water supplies, at a cost of $3,900,000, and the work could' 
 e done in from four to five years' time, of which at least a year 
 would be consumed in the necessary work of acquisition of title to 
 lands necessary. On another route from Great Falls, withdrawn 
 from the line of the existing aqueduct as far as topography will' 
 permit, an aqueduct with all appurtenant structures could be con- 
 .structed at a cost of $4,350,000, and in about the same length of 
 time. From the Patuxent Valley an aqueduct with all necessary 
 appurtenant structures could be constructed in from four to six. 
 years' time. Sufficient data as to cost of this development is not^ 
 however, at hand, and a known additional obstacle appears to its 
 adoption, as the power to enter the State of Maryland for this pur- 
 pose must first be obtained from that State. The project, how- 
 ever, offers known advantages which would justify its adoption as 
 against the Great Falls project, even at an increased cost of nearly 
 $1,000,000. These advantages justify a careful study at an early 
 date, for which the sum of $3,000 should be appropriated. Unti^ 
 the completion of this investigation it is unnecessary to inquire 
 further into the merits of the three routes from which selection will 
 have to be made. 
 
 Sixth. Contingent upon the construction of the coagulating plant 
 recommended, no additional storage capacity need be provided, 
 even when the additional conduit is constructed, nor will any addi- 
 tion be required to the present filtration plant beyond shght modi- 
 fications in the arrangement of ducts, even with a demand for 
 twice the present maximum consumption. 
 
 H. Doc. 347, 61-2—2 
 
18 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 Seventh. For reasons outlined in the report of Mr. Hazen, the 
 officials of the District of Columbia should make arrangement to 
 cover the present filtered-water reservoirs at Reno and Brightwood 
 at as early a date as will be permitted by funds at their disposal. 
 
 It is a matter of no small gratification to this office to be able to 
 make this report, which is made possible only by the excellent re- 
 sults which have been obtained by the introduction of such meters 
 as have already been installed and the thorough and efficient work 
 in the suppression of waste in the distributing system. Especially 
 is this so because the great consumption of water of from three to 
 five years ago justified the officials having charge of the water sup- 
 
 Ely and of the management of the affairs of the District of Colum- 
 ia in the assumption that additional works would be urgently 
 fieeded, possibly within the period required for good construction 
 work, and the investigations entered upqn under this approjjriation 
 were approached by all in the belief that the result of uie investi- 
 gations would show the necessity for early construction. It is 
 further gratifying because it can be demonstrated, as suggested 
 in Mr. Hazen's report, that the installation of all the measures rec- 
 ommended for immediate adoption can be effected at only a frartion 
 of the cost of interest charges on the outlay necessary for new works, 
 which interest charge will be a direct saving for so much time as the 
 construction of the additienal works can be postponed. 
 
 This office has. therefore to make the following recommendations: 
 
 1. That the sum of $50,000 be requested of Congress in the District appropriation 
 act to provide for the expenses of the fiscal year 1911 for the installation of meters 
 on all federal supplies, the sum to be appropriated entirely from United States funds, 
 to be disbursed under this office, the legislation to provide for the reading of the 
 meters also by this office. 
 
 2. That the appropriations previously asked for by this office of $130,000 for the 
 construction of a coagulating plant and for remodeling the Georgetown reservoir, 
 for use in connection therewith, be included in the District appropriation act for the 
 fiscal year 1911, the sum to be payable one-half from the revenues of the District of 
 Columbia and one-half from United Statei funds. 
 
 3. That there be included in the same appropriation act the sum of $250,000, and 
 in the appropriation acts for two ensuing years an equal sum, to be increased or re- 
 duced in the third year by such sum as the experience of the first two years may indi- 
 cate, in order to provide funds for the extension of the meter system to all private 
 Bervices in the District of Columbia, these appropriations to be made payable from 
 United States funds, with provision for their entire repayment to the United States, 
 with interest charges, over a period of about twenty years' time, as has been done in 
 cases of like nature in the past. 
 
 4. The inclusion in an early District appropriation act of an appropriation of $3,000 
 to provide for complete investigation of the project heretofore briefly described for 
 increasing the water supply of the District of Columbia from the Patuxent River. 
 
 In closing it is desired to place upon record an appreciation of 
 the work of Mr. Longley upon this project. His thorough and pains- 
 taking report, with all appendixes, will bear close examination and 
 jepay careful study on the part of all engineers interested in the 
 iSaportant questions related to supply and purification of water for 
 large municipalities. 
 
 Very respectfully, Jay J. Morrow, 
 
 Major, Corps of Engineers. 
 The Chief of Engineers, U. S. Army. 
 
 (Through the Division Engineer, Eastern Division.) 
 
WATER SUPPLY OF THE DISTRICT OP COLUMBIA. 19 
 
 KEPOET OF ASSISTANT SUPERINTENDENT FRANCIS F. LONGLEY. 
 
 Washington, D. C, May 1, 1909. 
 
 Sib: The act making appropriations to provide for the expenses of the government 
 of the District of Columbia for the fiscal year ending June 30, 1909, contained the 
 following: 
 
 "For preliminary investigations and surveys for increasing the water supply, ten 
 thousand dollars." 
 
 This was made upon the repeated recommendation of the oflScers in charge of the 
 Washington Aqueduct. The following project was approved for the expenditure of 
 the funds provided by this appropriation: 
 
 "The above sum of ten thousand dollars, or so much thereof as may be necessary, 
 is to be applied to a study of the available sources of additional water supply and the 
 means of bringing this water to the city; to making surveys, plans and estimates, 
 and to such other engineering and ofiSce expenses as may be necessary for completing 
 the work; to defraying transportation or other expenses of employees properly charge- 
 able to the work, and to the securing of a review and expert opinion upon the com- 
 pleted project." 
 
 In accordance with the verbal instructions of the officer then in charge of the Wash- 
 ington Aqueduct, I took charge of this work on July 1, 1908, and I now have the 
 honor of submitting the accompanying report, which sets forth in detail the procedure 
 of the work, the general conditions affecting it, brief consideration of all the various 
 possibilities for additional supply which have been investigated and rejected, a 
 more complete consideration of those possibilities which seemed favorable, together 
 with preliminary designs and estimates of cost thereof, and lastly, the conclusions 
 reached and recommendations made upon the entire .investigation. 
 
 The view of the expert employed for a review and opinion upon this report will 
 shoirtly be presented." 
 
 THE present water SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 The water supply of the District of Columbia is obtained from the Potomac River. 
 The existing worKs were constructed largely before the civU war. They consist in a 
 general way cf about 15 miles of conduit and tunnel and three reservoirs. The intake 
 IS just above the Great Falls of the Potomac on the Maryland side of the river. The 
 conduit follows generally parallel to the river and to the Chesapeake and Ohio Canal 
 discharging into the first or Dalecarlia reservoir located just above the District line 
 - and about 9 miles from Great Falls; thence through another 2-mile section of con- 
 duit to the second, or Georgetown reservoir; and thence through about 4 miles of 
 tunnel under the city to McMillan Park reservoir, adjoining which the filtration plant 
 is located. 
 
 The capacity of this system has been the subject ola careful study which is presented 
 as an appendix to this report (Appendix D). The point in which we are interested 
 in considering the present water supply resources of the District is the limiting quan- 
 tity that the system will discharge continuously from the river at Great Falls to the 
 Dumping station at the filtration plant. This quantity is 90,000,000 gallons per day. 
 he city can not, however, count upon drawing this quantity day after day throughout 
 the year. If the draft on the system — that is, the quantity of consumption and 
 waste — were absolutely uniform, it could be done, but this is not so. For well un- 
 derstood reasons, the daily draft in almost every city is subject to wide fluctuations. 
 The maximum variation that has been recorded in this city of the .consumption for 
 a day from the mean daily consumption for the year was about 45 per cent. The 
 necessity for maintaining a reserve capacity of this magnitude fixes the "safe annual 
 average" that can be drawn without danger of too serious depletion of the reservoirs 
 under abnormal conditions at about 62,000,000 gallons per day. 
 
 The studies elsewhere described, that were madf some twelve years ago upon the 
 flow in the aqueduct, indicated a discharge, in round numbers, of 75,000,000 gallons 
 per day with the reservoirs at levels which had been shown by experience to give 
 adequate storage and fair pressures in the city. This quantity has commonly been 
 stated as the "nominal capacity" of the system. It would represent the "safe annual 
 average " if, by the introduction of measures referred to later, the 45 per cent maxi- 
 mum excess draft could be reduced to about 17 per cent. 
 
 " See Mr. Hazen's review and opinion accompanying. 
 
20 
 
 WATER SUPPLY OF THE DISTBICT OF COLUMBIA. 
 
 With the water at the levels required for the flow of 62,000,000 gallons per day 
 
 through the conduit, the capacities of the three reservoirs above the lowest pomt to 
 
 which they can be drawn are as follows: 
 
 *' Gallons. 
 
 Dalecarlia reservoir 141, 000, 000 
 
 Georgetown reservoir :}^^> 000, 000 
 
 McMillan Park reservoir , 162,000,000 
 
 Xotal 436, 000, 000 
 
 This quantity may not all be considered as available storage because the rate at 
 which it would be drawn would diminish rapidly as the water levels approached their 
 lowest points. From 300,000,000 to 350,000,000 gallons could be drawn from the 
 reservou:B under the conditions assumed above, if the supply from the river were 
 shut off. This reserve is used fieely for the purpose of excluding muddy and bacteriar 
 laden water during freshets, thereby improving the quality of the supply. Its impor- 
 tant functions from the point of view of the capacity of the system are two, namely, 
 (1) to maintain a continuous supply to the city for a time in the event of an accident 
 which might interrupt the flow through the conduit, and (2) to supply the excess if 
 at a time of abnormal consumption the demand should exceed the limiting capacity 
 •of the system. 
 
 OBJECTS OP INCREASING THE WATER SUPPLY OP THE DISTRICT OP COLUMBIA. 
 
 There are in general two objects to be attained in increasing the water supply of 
 the District of Columbia. They are (1) to provide an additional supply to meet the 
 increasing demands of the city, and (2) to provide positive insurance against the 
 risk of a disastrous water famine in the event of an accident to the present system at 
 some of its weak points. 
 
 ADDITIONAL SUPPLY TO MEET INCREASING DEMANDS. 
 
 The mean annual consumption of water in the District for a number of years past 
 has been : 
 
 [Prom Annual Report ot tlie Chief of Engineers for 1906, p. 2092, and subsequent reports.] 
 
 Fiscal year. 
 
 Total con- 
 sumption 
 per day. 
 
 Per cap- 
 ita, daily 
 consump- 
 tion. 
 
 Fiscal year. 
 
 Total con- 
 sumption 
 per day. 
 
 Per cap- 
 ita, daily 
 consump- 
 tion. 
 
 1890 
 
 Gallons. 
 
 36,600,000 
 
 38,600,000 
 
 41,200,000 
 
 46,700,000 
 
 49,200,000 
 
 47,200,000 
 
 44,100,000 
 
 46,300,000 
 
 47,300,000 
 
 60,100,000 
 
 Gallons. 
 149 
 158 
 165 
 183 
 189 
 177 
 162 
 163 
 167 
 173 
 
 1900 
 
 Gallom. 
 48,600,000 
 50,700,000 
 56,000,000 
 . 68,300,000 
 63,000,000 
 70,600,000 
 67,700,000 
 66,900,000 
 64,900,000 
 
 Gallont. 
 164 
 
 1891 
 
 1901 
 
 169 
 
 1892 
 
 1902 
 
 183 
 
 1893 
 
 1903 
 
 187 
 
 1894 
 
 1904 
 
 199 
 
 1896 
 
 1906 
 
 218 
 
 1896 
 
 1906 .- 
 
 207 
 
 1897 
 
 1907 
 
 203 
 
 1898 
 
 1908 . ... 
 
 191 
 
 1899 
 
 
 
 
 
 The safe annual average consumption which could be depended upon without dan- 
 ger of shortage at times of abnormal draft is about 62,000,000 gallons per day. The 
 mean annual consumption tas for a number of years been approaching this safe limit 
 at an alarming rate. This has been evident to the several officers who have been in 
 charge of the Washington Aqueduct from year to year. As early as 1896 attention 
 was called to the necessity for providing an additional supply in the then near future. 
 (Report of the Chief of Engineers for 1896, p. 3966.) Since that year this has been 
 repeatedly urged, and accompanied by recommendations for an appropriation for the 
 preliminary steps leading up to a new supply. 
 
 In the year 1904, the consumption for the first time exceeded the safe annual aver- 
 age and in February, 1905, the capacity of the aqueduct was stretched to its limit and 
 . the reserve storage drawn alarmingly low. The consumption during the three years 
 since 1905 has not been quite as high as that year, but it has been above the safe 
 annual average. The need for a remedy for these conditions is one of the factors that 
 urged the consideration of means for increasing the water supply. 
 
 Insurance against shortage of water due to accident to present system. — The question of 
 insuring the District against risk of disaster due to the failure of some part of the pres- 
 
WATEE SUPPLY OF THE DISTBICT OP COLUMBIA. 21 
 
 ent aqueduct system was forcibly set forth by Ool. Smith S. Leach in his report on 
 the Washington Aqueduct for 1905, an extract of -which is here given: 
 
 "Attention is invited to the danger of an interruption of the supply by an accident 
 to the conduit. This structure has stood for nearly half a century, and without inters 
 ruption in its use other than to empty it periodically for cleaning. This fact is no 
 guarantee that it will not be interrupted in the futiu-e. 
 
 "It includes six tunnels aggregating 3,700 linear feet in seamy rock, and unlined. 
 They can not be lined until'a new conduit is built. At every cleaning fragments of 
 rock are found to have been dislodged from the roof. There is an appreciable risk of 
 more or larger pieces falling in, sufficient to block the tunnel for a considerable time. 
 
 "There are 26 culverts, nearly all in made groimd, the failure of any one of which 
 would break the conduit. They are of massive construction, but equally strong ones 
 have been washed out, and these may be. 
 
 "The Dalecailia reservoir is formed by an earthem dam across a valley. In this 
 dam is the by-conduit. If this dam should break, the by-conduit would be destroyed 
 and the supply of the District completely cut off. _ 
 
 "The statement' is justifiable that notwithstanding its fifty years of uninterrupted/ 
 use, the Washington Aqueduct is more vulnerable than the average of similar struc- 
 tures and that there is risk of an accident to it which would cut off the supply for a 
 period long enough to cause a water famine, which would involve a cancellation of 
 all fire insurance and a partial depopulation of the city, besides other and more calami- 
 tous consequences. This risk can be underwritten by the construction of another 
 conduit, and in no other way. " 
 
 NECESSITY FOR INCREASING THE SUPPLY. 
 
 The necessity for increasing the water supply of the District of Columbia, from a 
 consideration of the objects outlined above, depends in each case upon certain condi- 
 tions and assumptions that must be presented and considered in detail. 
 
 Prom the consideration of increasing consumption. — The consumption of water in 
 liie District has been for several years in excess of the safe annual average. If the 
 tendency were now toward a still further increase in total consumption there would 
 be absolutely no question as to the necessity for providing an additional supply with 
 the least possible delay. Whereas three years ago the rapid and unretarded rate of 
 increase of the consumption would have made such a recommendation imperative; 
 conditions of water consumption have arisen during those three years which introduce 
 a question of reasonable doubt as to the urgent necessity for increasing the supply at 
 the present time to meet increasing consumption. 
 
 Alarmed by the rapid increase in total consumption, as already pointed out, which 
 seems largely chargeable to useless waste of water, the officials in charge of the water 
 department of the District began in 1905 or 1906 an active campaign against waste. 
 The details of this work are given as far as the data are available in Appendix A. As- 
 there stated, the waste of water has been so reduced that the per capita daily consump- 
 tion has been lowered from 207 gallons in 1906 to 191 gallons in 1908, with a correspond- 
 ing reduction in the total mean daily consumption from 67,700,000 to 64,900,000 gal- 
 lons per day. Those in charge of the work of waste prevention believe that thesp 
 measures will still further reduce the per capita daily consumption, and to a degree; 
 sufficient to cause a material reduction in the total amount of water required. If 
 their judgment in this question is not in error and they should effect a considerable 
 reduction in the total consumption, the argument of rapidly increasing consumption, 
 which seemed so unmistakably potent three years ago would lose its force and would 
 not at the present time justify the expenditure of the large amounts necessary for 
 providing an additional supply. 
 
 The question of the use and waste of water in the District of Columbia and the 
 reduction in consumption through the efforts aimed at the suppression of waste are 
 considered at length elsewhere in this report (Appendix A). The conclusions thereon 
 may be summed up briefly as follows: 
 
 (1) A per capita quantity of 135 gallons of water per day should be ample for all 
 the legitimate needs of the District and for a generous allowance for unpreventable- 
 waste. 
 
 (2) A consideration of the results so far effected by the investigations for the detec- 
 tion and suppression of waste lead to the certain conclusion that the per capita con- 
 sumption can be considerably reduced and should be reduced to 135 gallons per day, 
 provided the water department be given the necessary funds for this work. 
 
 _ If this reduction can be effected without unreasonable delay, the total consump-i 
 tion will be greatly reduced and should not exceed the safe limit of the existing aque- 
 duct system, according to the present rate of increase in population of the District,. 
 
22 WATER SUPPLY OF THE DISTRICT OE COLUMBIA. 
 
 until Bome time after 1930._ Obviously, then, there ie no necessity at the present 
 time, from the point of view of increasing demand for water, for the construction of 
 new works, provided the District water department is given full support in the pre- 
 vention of waste. 
 
 • From the consideration of insurance. — The necessity of providing an additional sup- 
 ply of water at the present time from the consideration of insurance against risk of fail- 
 ure of the present aqueduct depends chiefly upon the stability of the structures there- 
 of, especially at the weak points. These weaknesses have been investiKited In all 
 the detail possible. Personal trips of inspection have been made through the aque- 
 duct by the officers in charge, accompanied bjr the assistants in local charge and those 
 engaged upon this investigation. From a critical examination of the weak points the 
 following statements may be made which are concurred in by those who were engaged 
 in that examination: - ' 
 
 (1) The present aqueduct system includes about 4,000 feet of unlined rock tunnel. 
 The rock itself is durable, but the whole mass is intersected by numerous seams of 
 comparatively soft material. The seams vary in thickness up to 3 or 4 inches or more, 
 and in many places the material in them has disintegrated to a very considerable 
 depth. At every cleaning fragments of rock are found to have been dislodged from 
 the roof or sides of the section. These are not of great size, and are broken up and car- 
 ried out of the tunnel. It is not believed that there is any danger of a fall of rock 
 great enough to cause an actual interruption of the water supply. A remedy for this 
 condition would lie in the complete lining of all the tunnel sections, especially where 
 the soft seams are most in evidence. This is practically impossible, since the tunnels 
 can not be drained for a period. longer than about three days. No real dangeris antic- 
 ipated from this source before new works will be required by reason of increasing 
 consumption. 
 
 (2) On the line of the present aqueduct there are twenty-six culverts, which cany 
 the drainage of considerable watersheds on the north side of the line. Over each cul-. 
 vert a fill of earth was made and the aqueduct built upon this fill. There is no record 
 of any of these embankments having been endangered by freshets, but this is of course 
 within the range of possibilities. This condition is guarded against by careful inspec- 
 tion and proper cleaning of the culverts, and there is believed to be no real danger 
 due to the flow of storm water through tUem under the aqueduct. 
 
 (3) Certain portions of the aqueduct resting upon some of the artificial embankments 
 described above have become badly deformed in their fifty years of service, due pre- 
 sumably to the settlement of the embankments and consequent change in direction 
 and intensity of earth pressures about them. The most noticeable deformation is 
 over culvert No. 20, where the aqueduct has settled about 5 inches. It has flat- 
 tened at this point so that its horizontal diameter has become from 8 to 12 inches 
 greater than its vertical diameter. It seems probable that such deformations as this 
 occurred soon after the aqueduct was built and that they have stood for many years 
 in their present condition. The deformation at culvert No. 20 was noted, but not 
 measured, as far back as 1896. If it be true that the deformation occurred when the 
 masonry was green, and that the section is not undergoing an^ further change of shape, 
 then there is probably no weakness inherent in the deformation, as the strength of the 
 structure is so largely dependent here as elsewhere upon the support of the surround- 
 ing earth. A continuing series of observations is to be instituted upon the shape of 
 the aqueduct section over culvert No. 20. If there is no change in shape, the struc- 
 ture is quite safe from danger due to this defect. 
 
 (4) Attention has been called to the condition of the brickwork of the aqueduct. 
 The mortar is inferior to the first-class Portland cement mortar that would be used 
 to-day in such work. This is most noticeable in the invert, in which the brick have 
 been loosened in many places by the occasional passage of men for inspection and 
 repair. In the entire length of the aqueduct, however, there is no place where a fall 
 of bjick has occurred, and in the total absence of any unusual forces acting upon the 
 brickwork, and with systematic inspection each time the water is drained off, there 
 is no reason for apprehending danger from this source. 
 
 (5) In the tunnel between Georgetown and McMillan Park reservoirs there is a por- 
 tion just east of Foundry Branch having a section vertically elongated in order to bring 
 the invert down to a grade which permits proper drainage. The brick side walls 
 lining this section are straight and have no arch action to support them. As a result 
 they have separated from the wall in two places and bulged toward the interior of the 
 tunnel. The maximum movement has been about 5 inches and the length of wall 
 affected about 15 feet in each place. These two existing defects have been held 
 against further movement by the use of adjustable struts made of heavy 4-inch pipe 
 No further danger of this kind threatens, though it is realized that bulging might occur 
 at otier points of this elongated section. .Even the complete collapse of a section of 
 this hmng, however, would not be a senous accident, though it could not be repaired 
 
WATEB SUPPLY OF THE DISTRICT OF COLUMBIA. 23 
 
 until after a new aqueduct is built, for it would not interrupt the flow of water through 
 the tunnel. 
 
 (6) The earthen dam forming Dalecarlia reservoir has been mentioned as one of the 
 vulnerable points of the system. It is true that the failure of the dam would destroj/ 
 the by-conduit and cut off the supply. But the dam has stood half a century, and, 
 unlike a brick-masonry structure, the lapse of time tends to strengthen rather than 
 weaken it. The reservoir is not subject to freshets. In view of these considerations, 
 this can not be seriously regarded as a point of weakness. 
 
 Stated briefly, the consideration of the stability of the structures of the old aqueduct 
 system fails to show any defect or weakness which threatens to interrupt the water 
 supply of the District of Columbia. It is confidently believed that no real danger of 
 this sort need be apprehended in the near future which would warrant the expen- 
 diture of the large amount necessary to provide a complete new aqueduct system 
 before such new works will be required anyway by the increasing consumption in 
 the District. 
 
 POSSIBLE SOURCES OP ADDITIONAL SUPPLY. 
 
 While it is an unexpected and most important conclusion that no new works are 
 needed at present for the extension of the water supply, and while it will be wel- 
 comed by Congress, still it was the intention of that body in appropriating money 
 for this investigation to secure a broad view of the problem of additional supply for 
 the District of Columbia, and, as far as possible, detailed presentation of the most, 
 favorable possibilities. The surveys originally intended have therefore been carried 
 out, aU possible sources given consideration, and general designs and estimates pre- 
 pared for those which seemed to present advantages that would give them any chance 
 for serious consideration in the future. 
 
 The possible sources of additional supply of water for the District of Columbia are 
 given below. Remarks are appended to indicate approximately the quantity that 
 might be developed from such a course, as compared with the quantity the city will 
 need during, say, the next fifty years. 
 
 Source. • 
 
 Comparative quantity available. 
 
 Potomac River at Great Falls 
 
 Unlimited. 
 
 
 Do. 
 
 Roek Creek 
 
 Relatively small quantity. 
 
 
 
 Groundwater . 
 
 Indeterminate. 
 
 Potomac at the upper branches .•• 
 
 Unlimited. 
 
 
 
 Brief consideration has been given also to the possibility of increasing the supply 
 by raising the dam at Great Falls, by the construction of a low-lift pumping station 
 at Dalecarlia reservoir, and by a separate imfiltered supply for industrial purposes 
 and fire protection. 
 
 CONSIDERATIONS AFFECTING THE CAPACITY TO BE PROVIDED FOR IN A NEW SYSTEM. 
 
 From the standpoint of design, the capacity to be provided for in an additional 
 supply of water for a city such as this will depend upon one of the following consid- 
 erations: 
 
 (1) If a source be selected which will yield a comparatively limited quantity, 
 provision may properly be made for the full capacity of that source. 
 
 (2) If a source be selected which will yield a quantity far in excess of the city'8 
 needs, provision may be made for an estimated consumption for an arbitrarily assumed 
 period of years into the future. 
 
 (3) Or, from a large source like the one just mentioned, the quantity may be deter- 
 mined by some special consideration, such as a legal decision, duplication of existing 
 camcity for some justifiable reason, etc. 
 
 jFrmn the Potomac at Great Falls. — The minimum flow of the Potomac at Great 
 Fails is approximately ten times the present consumption of water in the District ol 
 Columbia. The capacity of a new system from that point can not therefore be made 
 to depend upon the total quantity of water available. There are two factors, how- 
 ever, that have a bearing upon the point important enough to justify their use in 
 determining the capacity of,a gravity system from Great Falls. . 
 
 A corporation known as the Great Falls Power Company holds title to part of the 
 lands adjoining Great Falls, and, as far as has been determined by the courts, the' 
 title seems to include a portion of the water rights. The government dam at Great 
 Falls was raised in 1895-96 from elevation 148 to elevation 150.5, which resulted in 
 submerging some of the lands belonging to the above-named company. Following 
 upon a suit for damages in the United States Court of Claims,. the Great Falls Power 
 
24 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 dompany, for certain considerations named, on May 10, 1902, signed a quitclaim 
 deed^ an abstract of which follows : 
 
 "There is also hereby conveyed and released to the United States, as against any 
 Ownership by said company of land or water rights at or near the said Great Falb of 
 the Potomac, whether above or below the said government dam, the right on the 
 part of the said United States to take and divert from the said Potomac River from 
 above said government dam for the water supply of the city of Washington an amount 
 of water equal to 153,000,000 of United States gallons in twenty-four hours, which 
 amount is now estimated to be double the capacity of the present aqueduct, and all 
 claims for damages present or prospective on the part of this grantor or its successors 
 or assigns on account of the taking of water to that amount are hereby forever re- 
 leased." 
 
 There is then no apparent legal restriction to prevent the United States providing 
 works to take an additional quantity equal to the normal capacity of the present 
 system, which may be taken as 75,000,0000 gallons per day. 
 
 The present aqueduct system is half a century old. It is very much in need of 
 tepair at many points. These repairs can not be made under existing conditions, 
 -wmch require the system to be in practically continuous service. For this system 
 to fulfill its proper functions for a long period into the future it must be thoroughly 
 repaired. The completion of another project equal in capacity to the present one 
 Would enable this to be done properly and economically before the needs of the city 
 had grown to a point where they demanded the continuous service of both thq old 
 and the new systems. The increased cost and the depreciation of a project for 75,000,- 
 000 gallons per day, over one for say one-half that quantity, will be relatively small, 
 especially in view of the appropriation of most favorable sites and locations for the 
 first construction. 
 
 Governed by these factors the Great Falls gravity projects have been -worked out 
 on the basis of a nominal capacity of 75,000,000 and a maximum capacity of 90,000,000 
 gallons per day, which is equal to the capacity of the existing system. 
 
 Prom, the Potomac at Little Palls and at the tidal basin. — In the consideration of the 
 projects for an additional supply from Little Falls ajid from the Potomac at the tidal 
 basin, the normal capacity ultimately to be provided for has been assumed as 75,000,000 
 gallons per day, or equal to that of the Great Falls gravity project. 
 
 Prom the upper branches of the Potomac. — Because of the small consideration given 
 to a project for development of an additional supply from this source, no assumption 
 was made as to the quantity to be developed. 
 
 From Seneca Creek. — In order to facilitate the direct comparison of costs, in the brief 
 . consideration given this project, the nominal capacity was assumed at 75,000,000 gal- 
 lons per day, equal to the Great Falls gravity project. 
 
 From Rock Creek. — The total quantity that might be developed by impounding 
 from Rock Creek, according to the data at hand, and after providing tor maintaining a 
 constant and sufficient flow in the stream below the dam, is about 28,000,000 gallons 
 per day. The capacity for this project has therefore been assumed at that figure. 
 
 From ground water. — Because of the slight consideration given this source, no 
 assumption has been made as to the quantity to be developed. 
 
 JKBATMBNT OP ADDITION.^L SUPPLY. 
 
 The active satisfaction that the people of the District of Columbia have felt over the 
 Condition of the water supply since the filtration plant was started in October, 1905, 
 clearly justifies the assumption that any additional supply of water from the Potomac 
 River inust be purified. The question that naturally arises is whether new filters will 
 be required for this purpose or whether the capacity of the present plant will be suffi- 
 cient for the increased supply. 
 
 The filtration plant used for the present supply was designed for a rate of filtration 
 of 3,000,000 gallons per acre per day. This was in accordance with the most approved 
 practice at that time in Europe and in the United States; and any efforts to design 
 the plant for rates higher than that would no doubt have met with vigorous opposition 
 as being radical and unwarranted. 
 
 With the filtering area varying inversely with the rate adopted, and consequently 
 the cost of construction varying nearly in that ratio, the consideration of the allowable 
 rate of filtration for the large projects developed in the last few years has become of 
 great importance. This has resulted in the recommendation and adoption of rates 
 very much higher than formerly. For the new additional water supply of the city of 
 New York, the consulting engineers, Messrs. Allen Hazen and Geo W Fuller have 
 plpproved the preliminary steps leading to a design which will involve rates of at least 
 6,000,000 gallons per acre per day for the entire flow of the new Catskill Aqueduct. 
 Thia rate was most prominently mentioned in their communications, but they ex- 
 pressed their belief in the possibility of rates even somewhat higher. 
 
WATEB SUPPLY OF THE DISTKICT OF COLUMBIA. 
 
 25 
 
 Studies have been made and information obtained at the filtration i)lant in this city 
 during the three years it has been in operation which indicate conclusively that it can 
 be operated at a rate higher than 3,000,000 gallons per acre per day without any dete- 
 rioration of the quality of the effluent. Filters Nos. 1, 2, 3, 4, and 5 were run at rates 
 approximating 4,500,000 gallons per acre per day from December 14, 1906, to July 1, 
 1907. The remainder of the plant, twenty-four filters, was run during that period at 
 a 3,000,000 gallon rate. The monthly averages of bacterial counts and turbidity for 
 these two groups are given in the table below; 
 
 Date. 
 
 Filters 1 to 5, Inclu- 
 sive, rate 4,600,100 
 gallons per acre 
 per day. 
 
 Bacteria 
 per cubic 
 centime- 
 ter. 
 
 Turbid- 
 ity. 
 
 rilters6to29, incluv 
 sive, rate 3,000,000 
 gallons per acre 
 per day. 
 
 Bacteria 
 per cubic 
 centime- 
 ter. 
 
 Turbid- 
 ity. 
 
 December 14-31, 1906. 
 
 January, 1907 
 
 February, 1907 
 
 March, 1907 
 
 April, 1907 
 
 May, 1907 
 
 June, 1907 
 
 Average 
 
 42 
 
 43 
 
 Except for the different rates at which they were operated, the conditions influenc- 
 ing the two groups were identical. As far as bacterialefficiencies and turbidity results 
 are concerned, the filters operated at a rate of 3,000,000 gallons per acre per day do not 
 show any advantage over those operated at the rate 50 per cent higher. 
 
 During the month of September, 1908, certain experiments were started at the 
 ■ filtration plant with an equipment of small experimental filters for the jx^ress pur- 
 pose of studying the effects of various rates of filtration upon the efficiencies of the 
 process. It is intended to carry on this work for several years, in order to cover all 
 possible conditions of temperature and of bacterial and turbidity contents of the 
 Potomac water. General conclusions as to the laws governing the efficiency of filtra- 
 tion can not be drawn from the data so far obtained, but they do serve clearly to indi- 
 cate that within a certain limited range the rate of filtration at the Washington plant 
 may be increased without any deterioration in the quality of the effluent. 
 
 In the following table is given a summary of the results of operation of three of these 
 experimental filters, by runs: 
 
 Summary of results of experimental rate studies for rates (^ approximately 3,000,000, 
 6,000,000, and 10,000,000 gallons per acre per day. 
 
 
 Days. 
 
 Rate. 
 
 Turbidities. 
 
 Total 
 quantity 
 per acre. 
 
 Bacteria. 
 
 Per cent 
 
 Date. 
 
 Applied. 
 
 Filtered. 
 
 Applied. 
 
 Filtered. 
 
 bacteria 
 removed. 
 
 JFilter No. i, rurminal rale, 
 8,000,000 gallonaperday. 
 
 Sept. 9 to Oct. 21, 1908.... 
 Oct. 23,1908, to Feb. 3, 
 1909. 
 
 42 
 104 
 
 15 
 36 
 
 76 
 
 11 
 24 
 46 
 
 92 
 
 3.16 
 
 3.39 
 3.20 
 
 6.24 
 6.86 
 
 6.71 
 7.30 
 
 8.6 
 10.2 
 10.7 
 
 10.5 
 
 15 
 
 12 
 33 
 
 14 
 IS 
 
 10 
 30 
 
 15 
 15 
 10 
 
 20 
 
 
 
 
 5 
 
 
 
 
 
 
 4 
 
 
 
 
 
 3 
 
 132.6 
 352.5 
 
 88 
 
 3,325 
 
 837 
 
 28 
 
 30 
 
 92 
 
 68.2 
 ' 99 1 
 
 Feb. 10, 1909, to date 
 
 89.0 
 
 litter No. S, nominal rate, 
 8,000,000 gallons per day. 
 
 Sept. 10 to Sept. 24, 1908. . . 
 
 93.6 
 247.0 
 
 510.0 
 
 
 Sept. 27 to Nov. 1 , 1908. . . . 
 Nov. 4, 1908, to Jan. 18, 
 1909. ! 
 
 ISO 
 
 4,2S0 
 570 
 
 48 
 75 
 
 84.0 
 
 Jan. 20, 1909, to date 
 
 86 8 
 
 FiUer No. i,nmninal rate, 
 I0fi00,000 lattoM per day. 
 
 Sept. 10, to Sept. 24, 1908.. 
 Sept.26t0Oot.l9,1908... 
 
 Del. 22 to Dec. 6, 1908 
 
 Dee. 12, 1908, to Mar. 13, 
 1909 
 
 94.4 
 245. S 
 492.0 
 
 968.0 
 
 
 1,373 
 6,69S 
 
 380 
 
 14 99.0 
 29 99.6 
 
 47 87 A 
 
 
 
 
 a Unfinished. 
 
26 WATER SUPPLY OF THE DISTBICT OF COLUMBIA, 
 
 The averages indicate the effluent of No. 4 filter, operated at approximately a 
 10 000 OOO-^allon rate to be equal in quality to the effluent of either of the lower rate 
 filters. The numbers of bacterip, in the effluent during the several runs were no 
 higher, and the percentage efficiencies were generally as great for the high rate as for 
 the lower ones. . ^ ii j j ^ r 
 
 The range of conditions so far met in the expermients does not attord data tor a com- 
 parison of results under conditions any more unfavorable than those shown in the 
 above summary. It is believed that a water containing much greater numbers of 
 bacteria than are normally found in the water applied to the Washington filters 
 would be affected to a greater extent by the increase of the rate from 3,000,000 to 
 10,000,000 gallons per acre per day. But in view of the results of these experiments, 
 and the tendency toward the use of higher rates in practice, the conclusion is entirely 
 warranted that the Potomac water, in the condition in which it is applied to the 
 filters during the greater part of the year, can be satisfactorily filtered at a rate of 
 6,000,000 gallons aiyl probably at 10,000,000 gallons per acre per day. The present 
 plant operated at the lower of these two rates would yield 150,000,000 gallons per day; 
 and the maximum combined capacity of the old and proposed new aqueduct systems 
 would necessitate rates not exceeding 7,500,000 gallons, per acre per day. 
 
 Requisite conditions. — In order that the area of the present filtration plant may serve 
 to purify this greatly increased quantity of water, attention must be particularly 
 directed to certain interfering conditions which demand a remedy. It has just been 
 stated that the rate of filtration could saJely be increased providing the quality of the 
 water as applied to the filters was aa good as that now reaching the filters under ordi- 
 narily favorable conditions. This provision demands artificial improvement of the 
 water in the reservoir system at those times when water of high turbidity and bac- 
 teria is unavoidably admitted to the reservoirs. With the existing reservoir capacity 
 this necessity will increase as the consumption increases. The experiments conducted 
 in 1907-8 upon the preliminary treatment of the Potomac water showed coagulation 
 to be the only method that gave unfailing success in this improvement. Collation 
 must therefore be considered an essential factor in adapting the present filtration plant 
 to the higher rates. The process is entirely flexible and could be applied to whatever 
 extent might be indicated by the results of filter operation. It does not involve any ■ 
 great expense. A project for this improvement has already been submitted to Con- 
 gress, and has received the indorsement of -the public health authorities. 
 
 The greatest physical difficulty in the operation of the filters at the present time 
 is the abnormal shortening of the runs at certain times of the year due to the presence 
 of large numbers of micro-organisms in the applied water. This causes sudden and 
 uncontrollable reductions in the capacity of the plant, requiring heroic efforts lasting 
 for many days to furnish the quality of water demanded by the city, sometimes at 
 some risk to the high standard of quality that is so easily maintained under normal 
 conditions. This difficulty likewise will increase with the rate of filtration. The 
 logical remedy to apply is the elimination of these troublesome micro-organisms from 
 the reservoir system. The trouble is believed to be due principally to one predomi- 
 nating form, Melosira. Treatment with copper sulphate is the common and most 
 effective way of getting rid of such growths, and Melosira is quite sensitive to its 
 action. With careful study and persistent effort this trouble can be avoided. This 
 must be regarded as essential to the operation of the filters at the higher rates proposed. 
 
 Another physical difficulty in the operation of the filters arises from the abnormal 
 fluctuations in the city's consumption during periods of severe cold weather. This is 
 referred to at greater length elsewhere in this report. (See "Waste of water in wiuter 
 of 1904-5.") The remedy, as there stated, is the application of meters to all services 
 in the city, which will not only produce greater uniformity of draft, but will have the 
 effect of increasing the working capacity of the filtration plant and the entire aque- 
 duct system. 
 
 The filters, as stated, were designed for the normal rate of 3,000,000 gallons per acre 
 per day. In order that they might be operated at the higher rates, certain changes 
 would have to be made in the regulator houses and piping to adapt them to the new 
 hydraulic conditiono. 
 
 POINT OF DELIVERY OF ADDITIONAL SUPPLY. 
 
 In consideration of the conclusion just stated, the logical point of delivery of an 
 additional supply of water is the McMillan Park reservoir acfloining the present fil- 
 tration plant. The only exception to this would be in case an additional supply could 
 be developed sit a level high enough to deliver by gravity to the first high service of 
 the city, which is now supplied by pumping. This case is considered, in the Rock 
 
WATEB SUPPLY OS" THE DISTRICT OF COLUMBIA. 27 
 
 Creek and the Seneca projects. In all the other projects studied it is assumed that 
 tiie new supply will be delivered at McMillan Park reservoir. 
 
 In the case of the Great Palls gravity projects, the conditions controlling the 
 hydraulic gradient and certain other features of the plan are therefore determined by 
 the present normal water levels at Great Palls and McMillan Park reservoir. 
 
 SUPPLY FROM THE POTOMAC AT GREAT PALLS. 
 
 The Potomac River at Great Falls has a minimum flow approximately ten times as 
 great as the present consumption of water in the District. The United States has a 
 right to a large enough portion of the flow at that point to supply more than twice the 
 present population of the District. The existing aqueduct dam at Great Falls holds 
 the water at a level suitable for an additional supply from that point. The right of 
 way of the old aqueduct is for the most part wide enough for the location of a new line 
 wiuin its limits. An additional supply from Great Falls would have the benefit of 
 storage in the existing reservoirs. As aheady stated under "Treatment of the addi- 
 tional supply," a considerable part thereof could be filtered at the existing filtration 
 plant without exceeding the limit of its proper capacity. In view of all these facts, 
 the Potomac at Great Falls is a most logical source to consider for an additional supply 
 of water for the District of Columbia. 
 
 Conditions affecting the general location of a new aqueduct from Great Falls. — The gen- 
 eral route followed by the old aqueduct is the only surface route at the proper grade 
 between Great Falls and Georgetown reservoir. About 5,400 feet of this is in tunneL 
 It would be entirely feasible to locate another aqueduct alongside the old one with 
 but little more tunnel than that. In the Report of the Chief of Engineers for 1905, 
 page 2613, Col. Smith S. Leach, then in charge of the Washington Aqueduct, advocated 
 the choice of a new location separated from the old one. He said: 
 
 "With the idea of insurance paramount, the new conduit should not be built along- 
 side the present one, but should follow a radically different course to reduce to a 
 minimum the chance of a single cause disabling both. For this reason all possible 
 alternative routes should be examined." 
 
 This paragraph has been frequently quoted, and the arguments implied therein 
 clearly warranted the study of such a route. 
 
 Because of the fact stated in the first sentence oi the above paragraph, any other 
 location than that parallel to the present aqueduct would have to be largely in tunnel, 
 at a considerably increased cost. Between Dalecarlia and McMillan Park reservoirs 
 there are conditions, however, that render the separation of the old and the new 
 lines an economical advantage. The total length of the old aqueduct system, from the 
 inlet to Dalecarlia reservoir to the discharge into McMillan Park reservoir is about 
 37,500 feet, of which 20,700 feet is a pressure tunnel underneath the city. This length 
 of pressure tunnel could not be avoided nor shortened in a parallel location. Fur- 
 thermore, Georgetown reservoir is far removed from the direct line between Dalecarlia 
 and McMillan Park reservoirs, and a much shorter line can be found than the old one. 
 The new location proposed between these two points has a total length of 30,750 feet, 
 of which 19,400 feet is in tunnel. Its total length is thus less than the other by 6,756 
 feet, and the length in tunnel less by 1,300 feet. As an additional advantage, the 
 tunnel is all grade tunnel. 
 
 The separation of the two lines between Cabin Johns valley and Dalecarlia reser- 
 voir offers difiiculties. From Great Falls to Cabin Johns valley a tunnel location is 
 shorter than a location parallel to the old aqueduct; but from Cabin Johns Bridge to 
 Dalecarlia a line separated from the present one would have to be not only wholly in 
 tunnel, but materially longer. As this seems, on the whole, a reasanably safe section 
 of the line, and as a consideration of the relative advantages of the two locations be- 
 tween Great Falls and Cabin Johns valley will be sufficient for Qongresa to determine 
 upon a policy in regard to this point, but one location is given from Cabin Johns to 
 Dalecarlia, and that one parallel to the old aqueduct. Between Great Falls and 
 the Anglers' Club, a distance of about 2 miles, there are several places where the prox- 
 imity of the Chesapeake and Ohio Canal would render impracticable the construc- 
 tion of a new aqueduct on the south or west side of the old one. There are but few 
 places in this distance where construction on the side away from the canal would 
 not be in tunnel. The parallel location is therefore shown as a straight grade tunnel 
 from Great Falls to a point near the Anglers' Club. 
 
 For much of the distance between the Anglers' Club and Dalecarlia reservoir there 
 would be but little difference in cost of construction on the north or south sides of 
 the old line. There are several stretches, however, aggregating a good many thousand 
 feet, in which the old location hugs the precipitous nillsides, which would make a 
 
28 WATEB SUPPLY OF THE DISTKICT OF COLUMBIA. 
 
 high cost of construction on the north side. This location is therefore shown closely 
 parallel to and on the south side of the old aqueduct. 
 
 Proposed locations for new aqueduct from Great Falh.—The proposed location desig- 
 nated herein as the "parallel location" is as follows: From Great Falls to a point 
 near th* Anglers' Club, at the head of the Conduit road, a straight grade tunnel; thence 
 on the south side of the old aqueduct and parallel to it, cut-and-cover aqueduct, with 
 one short tunnel, to the north end of Dalecarlia reservoir; a by-conduit thence, skirt- 
 ing the south shore and crossing at the narrowest point to the east end of the reservoir; 
 thence a tunnel almost due east to Connecticut avenue extended, thence cut-and-cover 
 aqueduct along the side of the valleys of the branch just north of Linnean Hill and 
 of Piney Branch, to Fourteenth street; and thence a grade tunnel mainly under Spring 
 road and Warder street to the north end of McMillanPark reservoir. The total length 
 of this line, including the by-conduit around Dalecarlia reservoir, is 78,120 feet, of 
 which 30,440 feet is in grade tunnel. 
 
 The proposed location designated herein as the "tunnel location" is as follows 
 From Great Falls to a point on the west side of the valley near Cabin Johns Bridge, 
 a grade tunnel; thence a short section of cut-and-cover aqueduct and a siphon crossing 
 to the east end of the valley; from that point to McMillan Park reservoir the location 
 is the same as already described. The total length of this line is 75,924 feet, of which 
 50,240 feet is in grade tunnel. These general locations are shown on sheet No. 1 . 
 
 Grade of the new aqueduct.— The proposed Great Falls proiects are designed to take 
 water from the river at Great Falls and deliver it into McMillan Park reservoir. The 
 hydraulic gradients at these two points are fixed by existing conditions in the old 
 aqueduct system. Considerable change in these elevations could be made only 
 at great expense, and no advantages are suggested which would warrant such changes. 
 The elevation of the invert of the proposed new aqueduct has therefore been fixed for 
 these two terminal points, as follows: 
 
 The depth of the aqueduct section it is proposed to use is 8.75 feet. The depth of 
 the section of maximum discharge may be taken without much error to be 8.50 feet. 
 AtGreat Falls the crest of the dam is at elevation 150.5 and the lowest recorded stage 
 of the river is 150.7. The condition for maximum flow through the conduit at this 
 point will be fulfilled if the invert be placed at such an elevation that the upper line 
 of the section of maximum discharge coincides with the lowest recorded stage of the 
 river. This fixes the invert at Great Falls at elevation 142.2. 
 
 At McMillan Park reservoir under existing conditions, the normal water level for 
 the maximum capacity of 90,000,000 gallons per day is about 137.5. The conditione 
 for maximum flow at this point will be fulfilled if the invert be placed 8.5 feet below 
 this, or at elevation 129.0. ' 
 
 The elevation of intermediate points upon the systems proposed, depends upon a 
 number of conditions. In the old aqueduct the invert at the outlet of Dalecarlia 
 reservoir is lower than at the inlet by an amount sufficient to make the slope uniform 
 and continuous through the main conduit and by-conduit. For the condition of 
 maximum flow through the new system a similar design would not be advantageous. 
 This condition may always be anticipated some time before maximum draft actually 
 occurs, and the reservoir held in service. The hydraulic gradient across the reservoir 
 will then be flat, and the inlet and outlet elevations of the invert may be the same. 
 This assumption is an advantage. It reduces the total length to which the available 
 fall must be apportioned, and thus permits the use of a steeper slope with a conse- 
 quent diminution in the size of the section required. 
 
 The inlet and outlet gatehouses at Dalecarlia reservoir have therefore been assumed 
 at the same elevation. The proposed project for additional storage at the Stubblefield 
 site introduces the' same question which can be determined only after a decision has 
 been reached as to the adoption of that project. 
 
 No attempt has bee«i made in this investigation to adjust the grades of the tunnel 
 and cut and cover sections at different slopes for the maximum economy in construc- 
 tion. A study of this point before contracts for construction are let may result in a 
 slight reduction in cost. In the projects submitted herewith, the total fall has been 
 apportioned uniformly to the total length? of the lines, excluding the by-conduits 
 around Dalecarlia reservoir and around the Stubblefield reservoir, if provision is 
 made for storage therein. 
 
WATBE SUPPLY OF THE DISTEICT OF COLUMBIA. 
 
 29 
 
 The lei^th of the lines to the several gatehouses proposed are given below for the 
 three conditions to be considered" 
 
 Length of the proposed lines. 
 
 
 
 
 Parallel location. 
 
 Tunnel 
 location. 
 
 
 Without 
 storage. 
 
 With 
 storage. 
 
 Without 
 storage. 
 
 
 Feet. 
 
 
 
 Feet. 
 
 
 17,700 
 21,800 
 47,370 
 50,320 
 78,120 
 
 Feet. 
 
 
 
 Stubblefield reservoir: 
 
 Inlet ... 
 
 
 Outlet 
 
 
 
 Dalecarlia inlet 
 
 47,370 
 80,320 
 78,120 
 
 45,174 
 
 Dalecarlia outlet 
 
 48,124 
 
 HoUillan Park reservoir 
 
 75,924 
 
 
 
 Dalecarlia by-conduit 
 
 2,960 
 
 2,960 
 4,100. 
 
 2,960 
 
 'Stubblefield by-conduit 
 
 
 
 
 
 Total deductions 
 
 2,950 
 75,170 
 0.176 
 
 7,050 
 71,070 
 0.186 
 
 2,950 
 
 Length to which the total fall of 13.2 feet is apportioned.. 
 
 V2,974 
 0.181 
 
 
 
 
 The uniform apportionment of the available fall to these several lengths gives 
 elevations as follows for the gatehouses of the different projects: 
 
 Invett elevations at gatehouses. 
 
 
 Parallel location. 
 
 Tunnel 
 location. 
 
 
 With 
 storage. 
 
 Without 
 storage. 
 
 Without 
 storage. 
 
 Oreat Palls 
 
 Fea. 
 142.20 
 
 Feet. 
 \¥i.ia 
 
 138.91 
 138.91 
 
 134. 16 
 134.16 
 129.00 
 
 Feet. 
 142.20 
 
 Stubblefield reservoir: 
 
 Inlet 
 
 
 Outlet 
 
 
 
 Dalecarlia reservoir: 
 
 Inlet 
 
 133.88 
 133.88 
 129.00- 
 
 134 03 
 
 Outlet 
 
 134. 03 
 
 McMillan Park reservoir 
 
 129 00 
 
 
 
 Aqueduct sections. — The design of surface conduits of the class to which the proposed 
 new aqueduct belongs has in recent years tended unmistakably toward the so-called 
 ' 'horseshoe section, ' ' as combining the advantages of economy in construction and good 
 hydraulic efficiency. The question of the design of aqueduct sections was given 
 careful consideration in the report of the commission on additional water supply for 
 New York City in 1903, and they recommended sections of this general type. This 
 recommendation was adopted, and is now being carried out in the construction of an 
 aqueduct some 90 miles long and more than four times as large in sectional area as the 
 one proposed for this city. 
 
 The tendency in the design of tunnels for this purpose has been to approach a 
 rectangular-shaped section; or rather, a section of proper arch design which fits into 
 an approximately rectangular tunnel excavation. This type of tunnel, too, has been 
 adopted for the extensive New York work just referred to. 
 
 It is proposed to build both the cut-and-cover aqueduct and the tunnel sections of 
 concrete masonry. This is to be quite generally of a single specified proportion. 
 With the low velocities oi the water in the .conduit, the slight pressure, and with 
 proper attention to the density of the concrete, no trouble need be- apprehended 
 either from wear or leakage. The requisite smoothness of the interior surface maybe 
 obtained in general by the proper troweling of the invert; by the use of metal forms 
 of approved design for the entire interior surface above the invert, and by the careful 
 
30 WATEB SUPPLY OF THE DISTEIOT OP COLUMBIA. 
 
 spading and working of the freshly placed concrete in immediate contact with the 
 forms. A careful study of local conditions at certain points on the line, especially 
 in tunnel sections and in deep cuts, may indicate some economy in the use of brick 
 laid in Portland cement mortar. Where this is the case, when contracts are let for 
 this work, brick work may be specified, or perhaps better than that, alternative bids 
 may be asked for brick or for concrete construction. 
 
 For a number of reasons that are now recognized as superfluous, it has been common 
 practice to use a lining of brick in large aqueducts. There are therefore but few data 
 at hand to indicate the hydraulic vEiiue of interior siirfaces such as these proposed. 
 The little information there is indicates a high value, and this is natural to expect 
 from the smoothness of the surfaces that may be obtained. In the design of the vari- 
 ■ous sections the Hazen-Williams formula has been used. 
 
 D=cr "-^^ s ".«* 0.00] -"•"* 
 
 A study of the data at hand indicates that a value for c of 125 will probably be a 
 safe assumption after some deterioration of the surfaces, which is inevitable, shall 
 have taken place. 
 
 The slope of the aqueduct can be definitely determined only after the location 
 has been selected and the question of additional storage decided upon. For the pur- 
 pose of estimating, a slope of 0.176 per 1,000 has been assumed, as determined by the 
 length of the "parallel location" without additional storage. The adoption of the 
 Stuhblefield storage project would increase this slope to 0.186 with a slight reduction 
 in the size of the aqueduct section. With this slope of 0.176 per 1,000, and the value 
 c=125, aqueduct and tunnel sections of the general types aheady described have 
 been designed and are shown on sheets Nos. 14 and 15. 
 
 Principal structures on parallel location. — At Great Falls the intake and gatehouse 
 are combined as shown on sheet No. 16. A 48-inch drain leads from the gatehouse 
 to a point below the dam which will give a free fall, thus providing for the drainage 
 of the upper end of the aqueduct. From the gatehouse a tunnel leads to a point 
 near the Anglers' Club, as shown on the general location sheet. At that point the 
 new aqueduct is to be carried under the old one in a short tunnel from which a 24- 
 inch drain will lead to'the river. (Sheet No. 17.) The new line then follows the old 
 one on the south side. It is proposed to cross Cabin Johns valley in a section some 
 70 feet below the hydraulic grade line, and upon a bridge of concrete masonry imme- 
 diately over the stream. (Sheet No. 19.) The section is to be of steel pipe, entirely 
 surrounded by concrete. A 24-inch blow-off will provide drainage for this portion of 
 the aqueduct. 
 
 At the inlet of Dalecarlia reservoir a gatehouse is to be built. The new line will 
 approach this reservoir on the south side and leave it on the north side of the old line. 
 Some means must therefore be provided for crossing the one over the other. This 
 can be done in the gatehouse at the inlet. - The old by-conduit can be cut and the 
 new gatehouse constructed as shown on sheets Nos. 20^21, and after the new line has 
 been put in service the old aqueduct and the old by-conduit can be reconstructed for 
 -short distances and connected with their respective sets of gates in the gatehouse. 
 From this gatehouse a 48-inch drain is to lead to the old diversion tunnel of Little 
 Falls Branch. 
 
 The most direct and economical location for a by-conduit for the new line around 
 Dalecarlia reservoir involves the crossing of the reservoir at the narrow neck near the 
 south end. For this crossing a section of reinforced concrete has been designed, 
 •supported upon piers and arches of concrete masonry. (Sheet No. 21.) In order 
 that the waters of the reservoir may at all times have an unobstructed flow over this 
 crossing, the by-conduit has been lowered about 4 feet throughout its length. It 
 may be drained through the drain at the inlet gatehouse. 
 
 The outlet gatehouse at Dalecarlia reservoir consists merely of a simple arrange- 
 ment of gates for controlling the flow either from the reservoir or from the by-conduit 
 into the aqueduct tunnel leading east from this point. (Sheet No. 21.) 
 
 This tunnel comes to the surface just west of Rock Creek. The valley of Rock 
 Creek is about 1,400 feet wide at the grade of the aqueduct. The stream itself occupies 
 the bottom of a comparatively narrow gorge which can be crossed by a bridge of moder- 
 ate proportions, some 35 feet below the hydraulic grade line. The crossing has been 
 -designed as a steel pipe ^section carried upon such a bridge as this. (Sheet No. 22.) 
 The bridge has been designed of a width suitable for a highway. A more economical 
 structure could have been shown, but its location in the very heart of Rock Creek 
 Park and the great value it wouldhave as a much-needed high-level crossing of the 
 valley, seemed to warrant the design. A 48-inch blow-oft for drainage is provided 
 at this crossing. 
 
 From the last tunnel beginnii^ at about Fourteenth street and Spring road, the 
 water will be discharged through a gatehouse containing a simple arrangement of 
 
WATEB SUPPLY OF THE PISTBICT OF COLUMBIA. 31 
 
 controlling gates into the north end of McMillan Park reservoir, the most favorable 
 location for deriving the maximum benefit from storage and sedimentation therein. 
 (Sheet No. 22.) There is no point in the vicinity of this gatehouse low enough for 
 Its drainage, and the nearest blow-off on the aqueduct, at the Rock Creek crossing, 
 would leave fully 2 feet of water in the gatehouse and lower end of the aqueduct. A 
 small pumping equipment is designed to take care of the last drainage of this section. 
 It consists of an 8-inch centrifugal pump submerged in a pump pit slightly below the 
 level of the gatehouse floor, driven by a motor deriving its power from the dynamos 
 in the filtration plant pumping station. 
 
 The description just given applies to the structures that would be needed simply 
 to bring an additional supply of water from Great Falls to McMillan Park reservoir, 
 by the "parallel location, with only the present storage facilities. The proposed 
 project for additional storage of the supply at the Stubblefield site is presented else- 
 where in this report. If this should be adopted the above description would not be 
 materially altered. The proposed new reservoir would lie immediately alongside 
 the aqueducts, to both of which it could be easily connected through appropriate gate 
 houses. This would provide for the storage in the proposed reservoir of not only the 
 new but also the old- supply. 
 
 Principal structures on tunnel location. — -For the tunnel location the structures at 
 Great Falls are the same as for the "parallel location." The tunnel from that point 
 takes a more northerly route and leads straight through to Cabin Johns valley near 
 the bridge. Just west of the valley the line drops below the hydraulic gradient to 
 cross on a small bridge passing through under the old Cabin Johns- Bridge. This 
 small structure is similar to the one proposed for the same crossii^ on the parallel 
 location," beii^ located only a few feet away from the latter. The structures on the 
 remainder of the line to McMillan Park reservoir are the same as already described 
 for the "parallel location." 
 
 The tunnel location is tar removed from the proposed new reservoir at the Stubble- 
 field site, and there is therefore no simple way that these two projects could be com- 
 bined. The scheme for storage in Cabin Johns valley could easily be combined with 
 the tunnel location, but its inherent defects, elsewhere described, very properly 
 eliminate that from consideration. 
 
 PRELIMINARY TREATMENT OF THE ADDITIONAL SUPPLY. 
 
 The imperative necessity for coagulation of the water before filtration with the 
 increased rates at which it is proposed to operate the filters has already been pointed 
 out. The coagulation project now ready to present to Congress for use in connection 
 with the existing aqueduct system will provide tor this treatment until the consump- 
 tion exceeds the capacity of that system; and since it will be necessary to apply coagu- 
 lant only for short periods, the one coagulating plant between Dalecarlia and George- 
 town will probably serve satisfactorily for a capacity greater than that. The construc- 
 tion of a coagulating equipment tor the additional supply may therefore be postponed 
 until such time in the future as its need shall be shown. 
 
 The location of a future coagulating basin for the new aqueduct depends upon the 
 ■adoption of the project presented herewith tor additional storage. If a new reservoir 
 is to be built at the Stubblefield site, the plant could be located at its outlet, and a 
 favorable site could be found between that point and Dalecarlia reservoir tor a sedi- 
 mentation basin for the removal of the coagulated impurities. If the additional 
 reservoir project is not adopted, the logical location tor a coagulating plant would be 
 at the outlet of Dalecarlia reservoir with a coagulating basin on the line of the aqueduct 
 below that point. There is only one possible site east of Dalecarlia on the proposed 
 new line. This is just east of Connecticut avenue in the valley of a small branch 
 leading into Rock Creek. The difficulties of construction in this narrow valley are 
 great, and property is expensive, and a preliminary estimate indicates an extremely 
 high cost. If this site be put aside on the grounds of high cost, as seems advisable, the 
 next location to consider for a coagulating plant for the proposed additional supply 
 from Great Falls is the Stubblefield site. This site has not the natural advantage 
 that the other one has in the preliminary sedimentation of the water before treatment 
 by coagulation; and this deficiency would have to be made up for in construction at 
 the Stubblefield site by building a combined basin tor preliminary sedimentation and 
 for coa<?ulation. The total area need be but a small percentage of the area proposed 
 for additional storage alone at that site. An area of 20 acres, properly baffled and 
 paved for drainage and provided with a coagulating equipment and necessary gate- 
 nouses, could be constructed for $300,000. This would give a total capacity of about 
 100,000,000 gallons, and this is believed to be ample for the purpose. In the light of 
 the knowledge that should be gained from the operation of the plant it is proposed to 
 build in connection with the old system at Dalecarlia and Georgetown reservoirs, the 
 
32 WATER SUPPLY OF THE DISTRICT OF OOLUMBIA. 
 
 equipment for the proposed new system at the Stubblefield site might be built for 
 considerably less than the sum above named. 
 
 The coagulation of the old supply below and the proposed new supply above Dale- 
 carlia reservoir would make it necessary to prevent the mixing of the two supplies in 
 Dalecarlia reservoir during coagulating periods. The mixing of .an unsedimented 
 water with a water that has been partially or wholly clarified by sedimentation or 
 preliminary treatment is a decided disadvantage. In order to do this, the inlet gate 
 house at Dalecarlia reservoir would have to be redesigned to provide separate channels 
 for the two supplies through it; and when the Stubblefield coagulatmg plant was in 
 operation, the supply from the proposed new system would be carried through the 
 by-conduit at Dalecarlia and direct to McMillan Park reservoir. The inlet gatehouse 
 at Dalecarlia reservoir, redesigned to meet this requirement, would cost about $72,000. 
 
 If the proposed new line from Great Falls were already in operation, the most eco- 
 nomical and satisfactory arrangement would be to provide one large combined sedi- 
 mentation and coagulation basin for both old and new supplies at the Stubblefield site. 
 In view of the fact, however, that coagulation of the old supply is needed now, the 
 project should be carried out as already proposed, between Dalecarlia and Georgetown 
 reservoirs, and the decision in regard to the separate or combined coagulating plant 
 at the Stubblefield site reserved until the time when its construction shall be required. 
 
 In the estimate of cost of the Great Palls gravity project, a sum is included sufficient 
 to cover the cost of a coagulating equipment for the proposed new supply at the Stub- 
 blefield site, and also a sum sufficient to cover the increased cost of the inlet gatehouse 
 at Dalecarlia reservoir in the event of the adoption of the two separate coagulating 
 equipments instead of the single one for both. 
 
 Costs of the Great Palls projects. — Estimates of cost of the two alternative lines from 
 Great Falls to McMillan Park reservoir have been prepared with all possible care. 
 The unit prices assumed are believed to be generous enough so that any modifications 
 that may be made in the estimates will tend to .reduce rather than to increase the total 
 costs over those indicated. The estimates are given in detail in Appendix E, and a 
 condensed summary thereof is given below. 
 
 The probable prices of land required for the projected work have been obtained by 
 conversation with a number of real estate men and with the county surveyor of Mont- 
 gomery County. An allowance has been made in the estimate for the cost of tunnel 
 rights, the value of a strip on the surface 25 feet wide and directly over the proposed 
 tunnel lines being given for this purpose. 
 
 For the purpose of estimating the quantities of excavation, embankment, masonry, 
 etc., in the masonry aqueduct and tunnels, three types of section were assumed for 
 each, as follows: 
 
 Masonry aqueduct: (1) In soft earth; (2) in compact earth; and (3) in earth and rock. 
 
 Tunnel: (1) Rock, untimbered; (2) rock, timbered; and (3) earth, timbered. 
 
 It is believed that these six sections will constitute a percentage of the total length 
 in actual construction great enough to warrant their use in the estimate of cost for 
 this report. Exact information is not at hand as to the materials that will be pene- 
 trated m construction, but much light is shed upon this point by an acquaintance 
 with the old aqueduct. With this knowledge, supplemented by careful surface obser- 
 vations, these six typical sections were apportioned to the total length of the lines 
 to be estimated. 
 
 The estimates were made up from the profile of the center line of the proposed 
 structures. Tables were constructed giving the quantities for various depths of cen- 
 ter cut, with corrections to be applied for cross slopes. From these tables the quan- 
 tities were tabulated for each 100-foot station of the line, and from these summaries 
 were made of the total quantities in continuous lengths of sections of a given type. 
 
 The following condensed summaries^ give the costs of the two alternative locations 
 for the Great Falls gravity project: 
 
 Condensed summary of estimated cost, parallel location. 
 
 Aqueduct, standard section, 44,622 feet $1, 102, 562. 00 
 
 Tunnel, standard section, 30,440 feet 1,389,650.00 
 
 Crossing and sections of unusual type, 2,918 feet 310, 825. 00 
 
 Gatehouses, other structures, land, etc 626, 346. 00 
 
 Engineering and contingencies 450, 000. 00 
 
 Total estimated cost 3, 379, 383. 00 
 
 Estimated cost of construction per linear foot: ==^=== 
 
 Aqueduct, standard section 24. 71 
 
 Tunnel, standard section \'_ 45! 65 
 
 Crossing and sections of unusual type loe! 50 
 
 Estimated total cost of entire system per linear toot 49, 60 
 
WATEB SUPPLY OF THE DISTRICT OF COLUMBIA. 33 
 
 Condensed summary of estimated cost, tunnel location. 
 
 Aqueduct, etandard section, 22,894 feet $606, 061. 00 
 
 Tunnel, standard section, 50,240 feet 2, 280, 650. 00 
 
 Crossing and sections of unusual type, 2,650 feet 287,336. 00 
 
 Gatehouses, other structures, land, etc 609,639.00 
 
 Engineering and contingencies - 550, 000. 00 
 
 Total estimated cost 4, 333, 686. 00 
 
 Estimated cost of construction per linear foot: 
 
 Aqueduct, standard section 26. 46 
 
 Tunnel, standard section 45. 40 
 
 Crossing and sections of unusual type , 108. 40 
 
 Estimated cost of entire system per linear foot 57. 20 
 
 ADDITIONAL SUPPLY OBTAINED BY PUMPING. 
 
 From the Potomac just above the Little Falls. — A supply of water could be obtained 
 from the Potomac just above the Little Falls, and deuvered by pumping into Dale- 
 carlia reservoir. Such a project would render unnecessary the construction of a new 
 aqueduct between Great Palls and Little Fallg. The first point for consideration in 
 the comparison of this with the Great Falls project is that of cost. 
 
 The costs to be compared are the following: The costs of construction and capital- 
 ized cost of maintenance and operation — 
 
 (1) Of aqueduct from Great Falls down to point of location of pumping station. 
 
 (2) Of pumping station at Little Falls. 
 
 From the estimates already made, the cost of construction of that portion of the 
 Great FaUs gravity project necessary to bring 75,000,000 gallons of water per day to 
 the point of location of the Little Falls pumping station would be about ^1,560,000^ 
 The cost of maintenance and operation of this portion of the Great Falls project would 
 be about $15,000, which capitalized at 3 per cent amounts to $500,000. This makes 
 a total capitalized cost of the Great Falls gravity project to this point of $2,060,000. 
 
 The conditions under which a pumping station would work at Little Falls are as 
 follows: The station would be designed to deliver ultimately 75,000,000 gallons per 
 day, the same as the Great Falls project; this would be lifted against a head of about 
 110 feet, to the hydraulic grade line in Dalecarlia reservoir. 
 
 The cost of complete installation of pumping machinery, station, etc., to lift 
 75,000,000 gallons per day to a height of 110 feet may be roughly estimated at $400,000. 
 No preliminary designs or refined estimates are presented because the capitalized 
 cost of maintenance and operation alone make careful consideration of this point 
 unnecessary. The pumping expense which will be found considered at some length 
 elsewhere in this report (see Appendix B) is taken at $0.04 per million gallons lifted. 
 1 foot high. On this basis the annual cost of pumping 75,000,000 gallons per day, 
 to a height of 110 feet would be about $120,000. This sum capitalized at 3 per cent 
 would amount to $4,000,000, making a total capitalized cost for this pumping project 
 of $4,400,000, or more than twice as much as the Great Falls gravity project from 
 Great Falls to this point. 
 
 From the Potomac near the dty. — A supply of water could be obtained from the 
 Potomac River at some point near the city which would render unnecessary any of 
 the construction involved in the Great Falls gravity project. . Obviouslj; the water 
 taken from the river near the city would have to be both pumped and purified. The 
 purification could be effected either in a new filtration plant near the intake or at the 
 present filtration plant, the capacity of which is discussed more fully elsewhere in this 
 report. (See "Treatment of additional supply.") The hydraulic grade line of the 
 gravity portion of the District distribution system is approximately at elevation 140, 
 or practically the same as the elevation of McMillan Park reservoir, adjoining the 
 filtration plant, the elevation of which may be taken as 143. 
 
 All the water would have to be pumped to this height anyway, and therefore the 
 conptruction of additional filters could be avoided by pumping from the river station 
 to McMillan Park reservoir and filtering the water at the present plant. This fixes the 
 conditions for pumping, which would be as follows: The station would be designed for 
 a nominal capacity of 75,000,000 gallons per day; this would be lifted against a head 
 of about 143 feet plus the friction in the pipe lines. 
 
 A favorable site for such a pumping station may be found on the reclaimed land 
 owned by the United States near the new highway bridge. This site would be espe^ 
 cially advantageous on account of its proximity to the tidal reservoir, a portion of 
 
 H. Doc. 347, 61-2 3 
 
34 WATEB SUPPLY OF THE DISTRICT 01" COLUMBIA. 
 
 which would have to be converted into a basin for the preliminary treatment of the 
 muddy Potomac water before pumping it up to mix with the waters of McMillan Park 
 reservoir which have undergone several days of sedimentation. 
 
 The cost of pumping 75,000,000 gallons against a head of say 150 feet, at a cost of 
 $0.04 per million gallons 1 foot high (see Appendix B), would be $164,200 per year. 
 This capitalized at 3 per cent would amount to $5,473,000. The cost of pumping 
 equipment, pipe lines, preliminary treatment, sewer diversion, etc., would increase 
 this by more than $2,000,000. The capitalized cost of this project, when the total 
 quantity of 75,000,000 gallons per day had been developed, woiild be greater than the 
 capitalized cost of the Great Falls gravity project complete, which would be about 
 $5,879,383. 
 
 ' High cost of pumped supply. — The above paragraphs set forth in a general way this 
 two principal possibilities of obtaining an additional supply of water for the District 
 of Columbia by pumping. In both of them, the total cost, based upon a preUminary 
 pumping cost which is representative of the best class of stations in the country, is far 
 and away higher than the cost of bringing the additional supply by gravity from Great 
 Falls. This makes unnecessary any consideration of the unfailing reliability, the 
 much greater permanence, and certain other advantages in favor of the gravity system. 
 The question of pumping the water supply for the District of Columbia from the 
 Potomac River may therefore be finally dismissed. 
 
 SUPPLT FROM EOCK CKEEK. 
 
 In the investigations of half a century ago which led up to the construction of the 
 present aqueduct system, Rock Creek was considered in some detail as a source -of 
 Bupply. The RockCreek projects were rejected in favor of the Great Falls project 
 iiecause of the practically unlimited supply available at the latter source. The early 
 reports state the limiting quantitv available from Rock Creek under dry-weather con- 
 ditions at about 10,000,000 or 12,000,000 gallons per day. It is readily evident that a 
 quantity considerably greater than that could be developed by impounding, and that 
 source was therefore made the subject of further investigation. 
 
 Area of watershed. — ^The drainage area of Rock Creek is ^ven by the United States 
 Geological Survey as 86 square miles. Only a portion of this could be used for collect- 
 ing a supply for the District of Columbia. At the site which has so far seemed most 
 falvorable for a dam, located in a narrow valley just west of Brightwood and north of 
 the Military road, the drainage area as determined by planimeter from the United 
 States Geological Survey topographical maps is 62.8 square miles. Of this, about 3.5 
 square miles is or promises to be distinctly suburban and should in the event of the 
 adoption of such a project as this be sewered and diverted from the reservoir. The 
 effective drainage area has therefore been assumed as 59.3 square miles. 
 
 Available quantity. — ^The total discharge of streams of the Rock Creek class com- 
 monly lies between 1.0 and 1.5 cubic feet per second per square mile. This would 
 correspond to 40,000,000 to 60,000,000 gallons per day. The dry weather flow, as 
 stated in the old reports, did not exceed 10,000,000 or 12,000,000 gallons per day, and 
 recent observations indicate even smaller discharges. Storage of the flood waters 
 would therefore be essential to the success of this plan. 
 
 The only information available as to the discharge of Rock Creek was in the records 
 of the United States Geological Survey. This was very meager and not sufficient to 
 use as reliable basis for computation of the supply that the stream might furnish. 
 The mattfer was therefore referred to the hydrographic branch of the Geological Survey, 
 and the following memorandum was received in response: 
 
 Memorandum on the mean annual flow of Rock Creek, Maryland. 
 
 Rock Creek drains a total area of 86 square miles. Records of flow were obtained 
 from August, 1892, to November, 1894, wnich showed in 1893 the low rate of run-off of 
 0.703 second-foot per square mile, and even lower rates for the partial years of 1892 
 and 1894. 
 
 There is no apparent reason for discrediting, the accuracy of these records, nor can 
 this period be considered abnormally low. A comparison with the annual run-off of 
 the Susquehanna River at Harrisburg, Pa. (see table), tends to show that the rate of 
 flow in the East was generally normal at this time, although, of course, this can not be 
 considered conclusive as regards the rate of flow of Rock Creek. 
 
 A comparison of flow of streams in the vicinity of Rock, Creek is given in the ap- 
 pended table. With the exception of Antietam Creek, these show a generally high 
 rate of run-off, although Patuxent River, Which is nearest to Rock Creek is somewhat 
 
WATEE SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 35 
 
 lowerthan Monocacy and Patapsco rivers. The discharge data in this tablearecom- 
 pared on the basis of a seven-year period, 1900 to 1906. This period gives the same 
 vsdues as the longeaf period of 1897 to 1907, inclusive. - • u 
 
 If no information were available other than the flow of neighboring streams, it would 
 be advisable to give Rock Creek a lower value than Patapsco, Deer, and Monoeacy 
 rivers, owing to the tendency to a progressive decrease in the rate of run-off of streams 
 from north to south. Furthermore, more weight should be given to the Patuxent 
 River records, because this stream is immediately adjacent to the Rock Creek drainage. 
 On the basis of these neighboring streams alone, it is probable that a run-off per square 
 mile of 1.3 second feet would be a conservative estimate for Rock Creek. Other con- 
 siderations, however, lead me to believe that this value is much too high. 
 - It is noticeable that the axes of all tributaries of the Potomac River are approxi- 
 mately at right angles to the prevailing direction of storms. A study of the flow of 
 the various tributaries shows that in general the rate of flow of the long narrow basins 
 is less than the flow of wide circular basins. The Monoeacy River, which belongs 
 to the latter class, shows a higher rate of run-off than any other stream in the Potomac 
 River drainage with the exception of two tributaries near the source of the Potomac. 
 Antietam Creek and the South Branch of the Potomac River show very low rates of 
 run-off and are long, narrow. basins. ' They average about 1.1 second-feet per square 
 mile. AH small tributaries to the Potomac from Cumberland to Point of Rocks 
 average 1 second-foot per square mile. These are all long and narrow and com- 
 parable with Rock Creek in shape and size. This general law of shape to run-off 
 seems reasonable. Storms in general tend to a high rate of precipitation on ascend- 
 ing slopes and a low rate on descending slopes, hence in long narrow basins at right 
 ajigles to the direction of a storm a larger percentage of the total area would receive 
 a relatively lower rate of precipitation than in the case of a circular drainage. • 
 ,1 have been quite intimately acquainted with the general conditions of flow of 
 Rock Creek for the past five years and particularly so for the past year. Rock Creek 
 is noticeable for its generally low flow, averaging not more than 40 secondrfeet at 
 ordinary stages, and as is the case with other small streams, showing very short, flood 
 
 feriods. It seems to have the general characteristics of other small streams in the 
 'otomac River basin and in addition has more than two years records of flow, which 
 there is no good reason to doubt, which show an abnormally low rate of flow. Hence, 
 in my judgment, it would be unsafe to use a higher figure than 1 second-foot per 
 sqaaire mile for mean annual flow of Rock Creek, and I oelieve that 0.9 secondTioot 
 would be a more conservative value to use. 
 
 R. H. BoLSTEB, Assistant Engineer. 
 
 Mean annual run-off per square mile. 
 
 Year. 
 
 Antie- 
 tam 
 Creek, 
 Md. 
 (Sharps- 
 burg). 
 
 Mono- 
 eacy 
 
 River, 
 
 Md. 
 (Freder- 
 ick). 
 
 Rock 
 Creek, 
 Wash- 
 
 Patux- 
 
 ent 
 
 River, 
 
 Md. 
 
 (Laurel). 
 
 Patapsco 
 River, 
 Md. 
 (Wood- 
 stock). 
 
 Susque- 
 hanna 
 River, 
 Pa. 
 
 (Harris- 
 burg). 
 
 Deer 
 Creek, 
 
 Md. 
 
 (Church- 
 
 ville). 
 
 Little 
 Gun- 
 powder 
 Falls, 
 Md. 
 (Belair). 
 
 Great 
 Gun- 
 powder 
 Falls, 
 Md. 
 (Glen- 
 ooe). 
 
 Drainage 
 area,sq. 
 miles 
 
 , 295 
 
 6C0 
 
 86 
 
 137 
 
 251 
 
 24,000 
 
 141 
 
 43 
 
 160 
 
 1892 
 
 
 
 
 
 
 1.56 
 1.69 
 1.66 
 1.22 
 1.44 
 1.34 
 1.69 
 1.29 
 1.25 
 1.74 
 1.96 
 1.98 
 1.35 
 1.46 
 1.28 
 
 
 
 
 1893 
 
 
 
 0.703 
 
 
 
 
 
 
 1894 
 
 
 
 
 
 
 
 
 1895 
 
 
 
 
 
 
 
 
 
 1896 ! 
 
 
 
 
 
 
 
 
 1897 1 alia 
 
 1.49 
 1.59 
 1.59 
 1.09 
 1.63 
 2.50 
 2.72 
 1.06 
 1.65 
 1.87 
 2.17 
 
 1.77 
 
 
 1.36 
 
 i.es 
 
 al.26 
 
 a 1.69 
 
 a 1.14 
 
 <H.58 
 
 1.92 
 
 2.67 
 
 ■•.83 
 
 O1.40 
 
 1.85 
 
 1.78 
 
 1.62 
 
 
 
 
 1898 
 
 1.00 
 LIS 
 .65 
 ol.OO 
 1.39 
 1.71 
 '•.70 
 1.79 
 
 
 
 
 
 1899. . . 
 
 
 
 
 
 
 1900 
 
 
 
 
 
 
 1901 
 
 
 
 
 
 
 1902.. 
 
 
 
 
 
 
 i9QS 
 
 
 
 
 
 
 1904 
 
 
 
 
 
 
 1905 
 
 
 
 1.49 
 
 i.eo 
 
 6 1.80 
 1.54 
 
 1.60 
 1.77 
 2.12 
 
 1.73 
 
 1.60 
 
 1906. 
 
 
 
 1.48 
 
 190?.. 
 
 
 
 
 6 1.67 
 
 1900 to 1906 
 
 1.09 
 
 
 cl.SO 
 
 1.69 
 
 1.51 
 
 
 
 
 o Partial year, missing days supplied by a coe-llcient derived from the ratio of Antietam to Frederick. 
 * Missing months estimated. 
 ~ ' Based on comparison of 22 months at Laurel with corresponding 22 months at Woodstock. 
 
36 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 As a significant addition to this memorandum, the chief of the hydrographic branch 
 of the Geological Survey made the statement that the only Rock Creek gaugings 
 available (those of 1892, 1893, and 1894, of which 1893, the only coiliplete year is noted 
 in the memorandum), were taken at a time when stream measurement was in an 
 experimental stage. On the strength of his knowledge of the conditions and the 
 work done prior to 1895, he was inclined to discredit the Rock' Creek gaugings quoted 
 and rely upon the indications given by neighboring streams. 
 
 The substance of this is that but little is known of the total discharge of Rock Creek, 
 It may be as low as the conservative estimate of 0.9 cubic foot per second per square 
 mile, or it may be as high as 1.3 cubic feet per second per square mile, as indicated 
 by the neighboring streams. For an area of 59.3 square miles, these two rates of 
 run-oft would yield, respectively, about 35,000,000 and 50,000,000 gallons of water per 
 day. 
 
 The entire flow of the stream could not be diverted for water-supply purposes. 
 Rock Creek flows through Rock Creek Park and the Zoological Gardens, which are 
 visited daily by thousands of the people of Washington for rest and recreation. The 
 stream is a most important feature of these public parks and any diversion which 
 would mar its beauty would undoubtedly meet^with justifiable and strenuous oppo- 
 sition on the part of the park commissioners ariS the people. The utilization of the 
 waters of Rock Creek by storage, for the additional supply of water for the District, 
 would therefore require a definite reserve secure against encroachment, for main- 
 taining a satisfactory flow in the stream as it passes through Rock Creek Park. 
 
 The question of the proper quantity for maintaining the flow of the stream through 
 the park has not been taken up officially with the park commissioners, as such official 
 consideration would not be a controlling factor in determining the value of this 
 project. For the purposes of this investigation it has been assumed that about 18 
 cubic feet per second, or 12,000,000 gallons per day, should be allowed for this pur- 
 pose. An examination of the available stream-flow records for the years 1892, 1893, 
 and 1894 shows that the flow in Rock Creek was less than this assumed quantity for 
 50 per cent of the time during the three dry months, August, September, and Octo- 
 ber, of those three years. The quantity available for additional supply, with the 
 entire discharge of the stream impounded, would then be, upon the basis already set 
 forth, ijetween 23,000,000 and 38,000,000 gallons per day, or roughly from one-third 
 to one-half the nominal capacity of the existing aqueduct system. 
 
 Storage capacity required. — The Geological Survey, recognizing the errors arising 
 from the use of rainfall records in the determination of river discharge, state their 
 figures as given above, in cubic feet per second per square mile, which figures are 
 indepehdent of the rainfall factor. While presumably giviiig a more accurate total 
 discharge, these figures give no indication of the variations thereof and consequently 
 are useless for the estimation of the amount of storage required. Recourse was there- 
 fore had to rainfall records for this purpose. A study was made of the relation between 
 rainfall and run-off, by months, for certain of the neighboring streams for which 
 there were discharge records covering a number of years; and the result of this study 
 was a curve which should afford an approximate indication of the monthly varia- 
 tions in the stream flow of Rock Creek, and of such magnitude that the accumulated 
 discharge for an eighteen-year period should correspond to 1.1 cubic feet per second 
 per square mile, the average of the estimated discnarges already referred to. This 
 curve is given in diagram No. 1. It indicates that in order to have conserved the 
 entire flow of the stream during the eighteen-year period represented, an impounding 
 reservoir must have had a capacity above the lowest working level eq,uivalent to 
 14.6 inches 'of rainfall over the drainage. area of 59.3 square miles; and ,with this 
 storage the average quantity available would have been about 40,000,000 gallons per - 
 day, or about 28,000,000 gallons per day for water supply after providing 12,000,000/, 
 gallons for maintaining the flow in the stream through Rock Creek Park. This com- 
 putation is based upon rather meager data, but the best available. Fourteen and six* 
 tenths inches over the drainage area of 59.3 square miles is equivalent to about 
 2,000,000,000 cubic feet. 
 
 Elevation of storage reservoir. — The elevation of the crest of the dam forming the stor- 
 age reservoir depends upon two principal factors, the elevation of the minimum flow 
 line or lowest working level and the depth of storage capacity required above that 
 level. 
 
 The 
 
 service t 
 
 in the city are well established. The gravity supply from Great Falls fixes that of 
 the gravity service, and no circumstances surrounding either the present supply or 
 any supply that might be derived in future from Rock Creek seem to suggest any 
 advantage in a change in general flow line elevations. Of the total consumption in 
 the District of Columbia, about two-thirds go to the gravity service, about one- 
 
 'he general elevation of the minimum flow line depends principally upon the 
 /ice to which the water is to be delivered. The flow lines of the different services 
 
WATEK SUPPLY OP THE DISTBICT OF COLUMBIA. 37 
 
 fourth to the first high, and the comparatively small remainder to the higher services. 
 It would be a distinct advantage to he able to deliver a supply derived from Rock 
 Creek at elevation 210, the flow line of the first high service, and do away with a large 
 amount of pumping now required. This could be done. Furthermore, the necessity 
 of having an impounding reservoir of large capacity to make this source feasible, 
 would fix the minimum flow line at a high level anyway. To the extent to which 
 this project has been carried it has, therefore, been assumed that the supply would 
 be developed at such a level that it could be delivered to the city by gravity at eleva- 
 tion 210. 
 
 This advantage of greater elevation would necessitate purification works separate 
 from and at a higher elevation than the existing ones. An allowance of 30 feet for the 
 loss of head through the filters and through the pipe line to the city, and for aeration 
 if necessary, fixes the minimum flow line in the reservoir at elevation 240. 
 
 The capacity of the valley immediately above elevation 240 is given in diagram No. 
 1. Two thousand million cubic feet of water, the storage necessary in order to secure 
 the maximum supply from the stream, would fill the valley from elevation 240 to ap- 
 proximately elevation 257.5. For the purpose of a preliminary investigation this 
 
 ovation was assuqied for the spillway level and the top of the dam at elevation 260. 
 
 General features of the project. — Having approximate figures for the quantity avail- 
 able for water supply from Rock Creek, the location and crest elevation of the dam, 
 etc., a general lay-out was made of the location of structures essential to the full devel- 
 opment of this project. This is given in sheet No. 25. The project would require 
 the construction of a dam at the site and with the elevation already mentioned. Just 
 below the dam, filters would be built through which the water would flow by gravity 
 to a filtered water distributing reservoir immediately adjoining. From this reservoir 
 a steel pipe line would lead to a connection with one of the 48-inch mains of the 
 first high service, with a short branch to McMillan Park reservoir through which a 
 large quantity of raw water could be delivered for a limited time from Rock Creek to 
 that reservoir in the event of accident or repairs to the old aqueduct system. The 
 constant diversion at the dam for maintaining the stream flow in Rock Creek would 
 be utilized for the development of power for the operation of the filtration plant. 
 The flooding of the valley above the dam would necessitate the purchase of some 
 4,000 acres of land and all improvements thereon, the relocation of several miles of 
 county roads, railroads, and electric roads, with the construction of embankments and 
 bridges over the reservoir. A regard for the sanitary quality of the supply would 
 demand also the diversion or the disposal of the sewage from the suburban districts 
 adjoining the reservation site. An electric power station at Chevy Chase Lake would 
 have to be removed and reconstructed. 
 
 Cost of the project. — A preliminary estimate of the cost of the various portions of the 
 project is presented in the table below. For a fair comparison with the costs of the 
 other projects, the capitalized value of the annual cost of operation of the entire system 
 is included, thus giving the total capitalized cost of the project. A deduction is necese 
 sary for the cost of pumping 28,000,000 gallons of water per day 70 feet high, as the 
 adoption of the Rock Creek project would virtually effect a saving of this amount over 
 the cost of the Great Falls project at the lower level. 
 
 Prelimivfiry estimate of cost, Rock Creek project. 
 
 Main dam and spillway .• $1, 250, 000 
 
 Electric generating station and machinerj? 10, 000 
 
 Filtration plant and reservoir 750, 000 
 
 Pipe line to first high service 450, 000 
 
 Sanitary drainage of towns 250, 000 
 
 Land 800, 000 
 
 Damage to buildings. Forest Glen Seminary, etc 275, 000 
 
 Reconstruction of Chesvy Chase power station 100, 000 
 
 Relocation of county roads, with embankments and bridges 450, 000 
 
 Relocation of railroads, with embankments and bridges 730, 000 
 
 Relocation of electric roads, with embankments and bridges 250, 000 
 
 Engineering and contingencies 800, 000 
 
 Estimated total construction cost 6, 115, 000 
 
 ■vAimual operation, filtration plant, $30,000, which cai)italized at 3 per cent . 1, 000, 000 
 Annual operation, rest of system, $15,000, which capitalized at 3 per cent . 500, 000 
 
 7, 615, 000 
 950, 000 
 Estimated net capitalized cost for comparison with Great Falls project. . . 6, 665, 000 
 
 Annual cost of pumping 28,000,000 gallons per day 70 feet high, at $0.04 
 per million gallons 1 foot high, $28,600, which capitalized at 3 percent. 
 
38 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 The enormous cost of this plan, and the limited quantity of water that could be 
 obtained for an additional supply, make any further consideration of this project 
 unnecessary. 
 
 Supply fro-m Rock Creek at lower level. — In the estimate just given, the large items 
 of expense are for the high dam required, the filtration plant, and the relocation of 
 roads, railroads, and electric roads. The development of a supply from this source at 
 a lower level, with delivery direct to the existmg filtration plant, would eliminate 
 some of this expense and reduce all. 
 
 A supply could be developed from Rock Creek with its minimum working level at 
 elevation 185, which would give ample head for its delivery by gravity to the filters 
 at elevation 165. A computation from the available data indicates that a dam with 
 crest at elevation 227 would give adequate storage. This development would require 
 the purchase of only about 1,750 acres of land instead of the 4,000 estimated above. 
 The dam required would be smaller. From the dam a pipe line would lead to a con- 
 nection with the raw water distribution system to the filters. 
 
 A preliminary estimate of the cost of the several portions of this development is as 
 follows; 
 
 Main dam and spillway $500, 000 
 
 Pipe line to filtration plant 416, 000 
 
 Sanitary drainage of towns 250, 000 
 
 Land 355, 000 
 
 Relocation of roads, railroads, etc 545, 000 
 
 Engineering and contingencies 300, 000 
 
 Total 2, 366, 000 
 
 Since writing the above in regard to the available capacity of this source, authori- 
 tative statements have been received, based upon wide experience, that indicate the 
 available capacity considerably higher than stated above, namely, about 40,000,006 
 gallons per day after making allowance for stream flow through Rock Creek Park. 
 On the basis of this increased quantity the development of a supply from Rock Creek 
 at the lower level, with its much lower cost, seems attractive. 
 
 The increasing value of real estate, with the rapid suburban growth in the direction 
 of the upper valley of Rock Creek may put a different light upon this point, due to 
 the prohibitively high cost of the large area required, by the time new works will 
 be necessary. At the least, any conclusion upon this point must be left to be deter- 
 mined by future conditions. 
 
 PROJECT FOR BRINGING WATER AT HIGH LEVEL FROM SENECA CREEK. 
 
 Seneca Creek, which flows into the Potomac about 8.5 miles above Great Falls, 
 has a drainage area of about 127 square miles. The available portion of this area would 
 yield, with storage, 75,000,000 gallons per day or more. This could be impounded at 
 an elevation considerably higher than the Potomac at Great Falls , and could be 
 delivered to the city at least as high as the first high service, elevation 210, thereby 
 effecting a saving in the pumpage of between 25 and 30 per cent of the total supply. 
 This would involve, the purification of this portion in new filters located at a high 
 level, and the delivery of the balance of Jhe supply to McMillan Park reservoir for 
 purification in the existing filters. A suitable site could be found for filters and 
 reservoir on the high ground west of Georgetown. 
 
 The principal structures required for this development would be a dam across 
 Seneca Creek; an aqueduct to the city; filters and reservoir; connecting pipe lines 
 to the city and to McMillan Park reservoir. 
 
 Dam near Seneca. — No topographical maps are available for the Seneca region 
 except those of the United States Geological Survey on the scale of 1:125,000. The 
 much greater distance and the roughness of the topography above Great Falls, indi- 
 cated without much question that the development could not be as economical as 
 the Great Falls project, and therefore the extensive surveys that "Hrould have been 
 required for its full presentation were not undertaken. Without attempting, with the 
 meager data in hand, to obtain an economical balance between slope of aqueduct and 
 height of dam, the slope and section were assumed the same as for the Great Falls 
 aqueduct. This condition would fix the crest of the dam at Seneca at about eleva- 
 tion 270, after making allowances for reservoir fluctuations and losses of head through 
 aqueduct, filters, and pipe lines. The maps of the Geological Survey indicate that 
 a dam just above Seneca with a crest at elevation 270 would have a length of about 
 1,400 feet and a maximum height of 130 feet. Such a dam would cost m excess Of 
 
WATEE SUPPLY OP THE DISTRICT OP COLUMBIA. 3& 
 
 $2,000,000. Land for reservoir, relocation of roads, clearing, draining, etc., would 
 cost perhaps $500,000 more. As the available data on this point are very meager, a 
 
 fenerous assumption is made for the purpose of comparison of this with the Great 
 'alls project, that the dam and reservoir at Seneca would cost no more than in Cabin 
 Johns Valley— that is, about $1,400,000. 
 
 Aqueduct from Seneca to Washington, — The lengths of the different types of struc- 
 tures for the aqueduct would be about as follows: 
 
 Feet. 
 
 Length in cut-and-cover aqueduct , 78, 500 
 
 Length in ttmnel 11,500 
 
 Length in inverted siphon, bridge, etc., including all valleys 50 feet deep 
 
 or more 20,000 
 
 Total length , 110, 000 
 
 These would cost, on the basis of the average estimated cost of similar types of sec- 
 tions for the Great FaUa project: 
 
 78,500 X $24.72..... $1, 940,000 
 
 11,500X$31.23 359,000 
 
 20,000X$78.50 1,570,000 
 
 Total cost of aqueduct 3,869,000 
 
 Filters, reservoirs, etc. — Filters for purifying the supply for the first high service 
 would have to have a capacity of about 30 per cent of the total supply. After the 
 increase of the supply this would be about 45,000,000 gallons per day. A filtration 
 
 C' nt of 9 acres would properly purify this quantity. This would cost, together with 
 d, gatehouse, sand-washing equipment, etc., about $1,000,000. A distributing 
 reservoir for filtered water would cost $100,000. 
 
 Connections with city and McMillan Park reservoir. — Pipe lines to the city would cost 
 about $200,000, and to McMillan Park reservoir about $300,000 more. 
 
 Cost. — A summary of the approximate cost of bringing a supply of water front 
 Seneca Creek is given below. In order to make it comparable, with the estimated 
 cost of the Great Falls gravity project, the maintenance and operation charges are' 
 included, giving the total capitalized cost; and a deduction is made for the cost of 
 pumping which would be saved in the event of the adoption of a high level project, 
 such as this. 
 
 Summary of estimated cost of Seneca project. 
 
 Dam, reservoir, lands, etc., near Seneca, at least.'. $1, 400, 000 
 
 A3ueduct 3, 869, 000 
 
 Filters, reservoirs, etc 1, 100, 066 
 
 Connections with city and with McMillan Park reservoir 500, 000 
 
 Engineering and contingencies i. 820, 000 
 
 Total construction 7,689,000 
 
 Maintenance and operation per year: 
 
 Aqueduct system $30, 000 
 
 Filtration plant •. 30, 000 
 
 Total maintenance and operation 60, 000 
 
 Which capitalized at 3 per cent 2, 000, 000 
 
 Total capitalized cost of Seneca project 9, 689, 000 
 
 Credit for cost of punjping saved: 45,000,000 gallons. per day, 70 feet high, 
 at $0.04 per million gallons 1 foot high, $46,000 per year, which capi- 
 talized at 3 per cent 1, 533, 000 
 
 Estimated net capitalized cost of Seneca project, for comparison 
 with Great Falls project 8, 156, 000 
 
 The estimated cost of the Great Falls gravity project, complete, is $3,879,383. If the 
 maintenance of this system beestimated at $30,000 per year, and the additional cost 
 of maintenance and operation at the filtration plant be $30,000 more, the total capi*- 
 talized cost of this project in its final development would be about $5,879,383. The 
 estimated capitalized cost of the Seneca project is so greatly in excess of this that it 
 may be dismissed from any further consideration in this report. 
 
40 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 QEOUND-WATBR SUPPLY. 
 
 As a source of a large supply of water for the District of Columbia the ground water 
 can not be seriously considered. A ground-water supply for this city would mean 
 one drawn from wells driven into the water-bearing gravel deposits underlying the 
 District. Very few large cities have supplies of this kind, and it may be stated that 
 only under most exceptional conditions would a city of the size of Wadiington depend 
 upon such a supply if a supply of surface water were available. 
 
 The city of Brooklyn, N . Y., draws half or more of its supply from the sands of Long 
 Island. The geological formation of Long Island is unusually favorable for a large 
 storage of ground water in readily available form for a -jvell supply. 
 
 The experience with so-called "deep wells" in the District of Columbia has demon- 
 strated the presence of water-bearing strata quite widely distributed, but the con- 
 ditions are far less favorable than on Long Island. The average yield from 21 of the 
 deep wells of the District is about 40 gallons per minute. It would require 1,350 
 wells of this capacity to supply 75,000,000 gallons per day. 
 
 Flowing wells can not be obtained in this locality. Such a supply as this would -in- 
 volve continuous pumping, and in general the projects we have outlined involving 
 continuous pumping cost more than the gravity projects. The pumping, moreover, 
 would have to be at a number of isolated stations, as the wells obviously could not 
 all be grouped together, and this would tend to make the cost of pumping still higher. 
 
 It may be suggested as an advantage in favor of a ground-water supply that no fil- 
 tration would be required. The data at hand do indicate, it is true, that a deep-well 
 supply, locally so'called, would be of excellent sanitary quality without filtration; 
 but the quantity of iron in this water would make its general use prohibitive without 
 subjecting it to some process for the removal of the iron. Presumably aeration and 
 filtration would be necessary and suflBcient and therefore the high sanitary quality 
 of such a supply would not be an economical advantage. 
 
 From the known facts regarding the mutual interference of wells placed in groups, 
 and the uncertainty of any predetermination of the quantity of water that might De 
 drawn from the water-bearing strata underlying the District of Columbia, such a plan 
 could not be entered upon with any assurance of success as to quantity developed 
 nor as to relative economy. A single well or a small group of wells may make an ideal 
 supply for private consumers or for industries of some magnitude, but it docs not aflord 
 a generally satisfactory solution of the problem of securing a sufficient and unfailing 
 supply for a large community. 
 
 ADDITIONAL SUPPLY PROM THE UPPER BRANCHES OF THE POTOMAC HIVBH. 
 
 The idea of bringing a supply from the upper branches of the Potomac River has 
 been suggested. It would be possible, of course, and the idea appeals to the layman 
 who thinks only of the purity of a stream near its source. 
 
 The confluence of the two forks of the south branch of the Potomac which lies just 
 above the town of Petersburg, W. Va., is about 115 miles from 'Washington in a direct 
 line. Assuming that construction over this line were a possibility, the cost of bring- 
 ing water from that source, at the same total cost per linear foot as the Great Falls 
 gravity project, would be about 124,000,000. The advantage argued for such a dis- 
 tant source is the pristine purity of the water, but that term is only relative. The 
 possibilities of the pollution of supply from even the most remote source are so great, 
 and the protection afforded by filtration has been so thoroughly demonstrated, that 
 there is a strong tendency toward the filtration of all surface supplies. The purifica- 
 tion of the water from the Potomac at Great Falls has been so eminently satisfactory 
 as to bring forth favorable comment from every authority. Cities that are going great 
 distances for supplies from mountain sources are commonly doing so not because of a 
 fear of the efficiency of purification processes but because of the insufficiency of nearer 
 sources. The expenditure of this great amount of money for an additional supply 
 would be entirely unwarranted under existing conditions. 
 
 RAISING THE DAM -AT GREAT PALLS. 
 
 The quantity of water flowing throurfi the present aqueduct system could be 
 increased by raising the dam at Great Palls, which would increase the slope of the 
 hydraulic gradient. Such a. plan is limited by two considerations: 
 
 (1) Interior pressure on aqueduct. 
 
 (2) Damages due to raising the water level above the dam. 
 
 The old aqueduct, where not in tunnel, is of brick masonry and was not designed 
 for nor expected to resist any interior pressure. In some places it is actually under 
 
WATEE SUPPLY OP THE DISTKIOT OE COLUMBIA. 41 
 
 a pressure of as much as 3 or 4 feet but the strength to resist this pressure lies in the 
 .eaJth embankment over it. From this consideration it would therefore not be safe 
 to make much of a raise in the hydraulic gradient at Great Falls. 
 
 The principal parties that might be damaged by raising the water level above the 
 dam at Great Falls are the Great Falls Power Company and the Chesapeake and Ohio 
 Canal Company. The Great Falls Power Company owns Conns Island, which would 
 be partially or entirely submerged. This island has very little intrinsic value, but 
 its submergence might be made the pretext for another expensive damage suit by that 
 company against the United States. The Chesapeake and Ohio Canal (Simpany holds 
 property along the Maryland side of the stream above Great Falls. Any considerable 
 damage done them would be at times of freshet in the river. The question of the 
 limiting height of the dam from this point of view has been considered before. Mr. 
 Josiah Pierce, jr., chief engineer of the Great Falls Power Company, in 1902 or earlier, 
 conferred with Mr. Nicolson, chief engineer of tBe Chesapeake and Ohio Canal Com- 
 pany, with reference to the safe elevation for the crest of a dam at this point, and they 
 reached the conclusion that elevation 153, or 2.5 feet above the present crest, was as 
 high as it would be safe to build it. 
 
 If the dam were to be raised 2.5 feet, the capacity of the aqueduct system would be 
 increased fey less than 10,000,000 gallons per day. The dam has already been raised 
 once, and to raise it again simjjly by placing more material on top would lower its 
 factor of safety to a dangerous point. It would be necessary, therefore, to increase the 
 mass of the dam throughout its height, and the cost would probably be about $500,000. 
 
 In view of the limitations to such a plan as this, together with ite cost and the small 
 increase it would make in the capacity of the system, its consideration seems of small 
 importance. 
 
 WASTE OP WATER IN WINTER OT 1904^5. 
 
 During the winter of 1904-5 there occurred the greatest draft that the aqueduct 
 system has ever been subjected to. The maximum daily consumption rose almost to 
 100,000,000 gallons, corresponding to a quantity in excess of 300 gallons per capita 
 per day. This abnormal draft lasted the greater part of January and February, and 
 made the mean consumption for the year 70,600,000 gallons per day, whereas the 
 mean for nine months of the year, excluding the winter months, was only 67,100,000 
 gallons per day. 
 
 The days of abnormal draft occurred invariably when the temperature was low. 
 It has been the comipon experience of a great many cities to have the consumption 
 jump up to a considerable percentage over the normal upon the occurrence of freezing 
 weather. There is no mystery as to where the water goes. Fixtures are opened 
 freely in all parts of the city to create a current in the house service pipes sufficient.to 
 prevent the freezing thereof. The high consumption here referred to was without 
 doubt caused by just such waste as this. There is nothing inherent in freezing weather 
 that would tend to increase the legitimate use of water. The tendency would, if any- 
 thing, be in the other direction. There is nothing, either, which would tend to 
 increase the waste from fixed leaks in mains or service pipes. The greater part of the 
 excess of the abnormal draft over the normal consumption may Qierefore, without 
 much chance of error, be charged to the waste of water from open fixtures to prevent 
 freezing. 
 
 The effect of the increase in the consumption of water in the city is to draw down 
 the levels of the reservoirs. In an aqueduct system a certain slope of the water sur- 
 face is required to produce the flow of a certain quantity, and the greater this quan- 
 tity the greater must be the slope. If the quantity taken from the lower end of the 
 system be increased, the greater quantity flowing through the aqueduct system to 
 replace it adjusts itself to the steeper slope required. Since the water level at the 
 upper end of the system is not susceptible of control, it therefore follows that this 
 adjustment to the steeper slope involves the lowering of the water levels in the lower 
 end of the system, including the reservoirs. This may be seen from the "Discharge 
 dia^am of the Washington Aqueduct," Appendix D, whereon are platted lines indi- 
 cating the slopes the water takes for various volumes of discharge. 
 
 The greatest disadvantage resulting from this abnormal draft upon the system 
 arises from the conditions fixing the safe annual average consumption that the system 
 is good for. As has been stated, the maximum capacity of the system is 90,000,000 
 gallons per day, and under existing conditions an occasional abnormal draft may be 
 expected 45 per cent in excess of the mean daily consumption for the year. Nearly 
 a third of the ultimate capacity of the system must therefore be held in reserve to 
 meet these periods of abnormal waste. Kemedying the condition by cutting down 
 the percentage excess will obviously have the effect of increasing the safe working 
 capacity of the entire system. 
 
42 WATER SUPPLY OF THE DISTBICT OF COLUMBIA. 
 
 Another disadvantage resulting from the drawing down of the reservoirs of the aque- 
 duct system is that it results in a reduction of the available reserve that would consti- 
 tute the only source of water supply for the city in the event of the interruption of 
 flow in the aqueduct. It reduces also the reserve available for maintaining the sup- 
 ply to the city when the consumption exceeds the maximum quantity which can flow 
 through the aqueduct. The reduction of the quantity of water in the reservoirs, 
 ogether with the increase in quantity flowing, combine to make a great reduction in 
 the time it takes for water to pass through flie reservoirs, and consequently in the 
 degree of purification effected thereby. 
 
 The high consumption of January and February, 1905, drew down the levels of the 
 reservoirs to an alarming extent. It reduced the total quantity of water that could 
 possibly be drawn from the three reservoirs in the event of an interruption of flow 
 above iDalecarlia reservoir from 400^000,000 to about 275,000,000 gallons, or from six 
 days' supply at the normal rate of consumption to less than three days' supply at the 
 rate existing at the time of the depletion. This total quantity, practically speaking, 
 would not have been available, as the rate at which it could be drawn would have 
 dropped off rapidly. Under these conditions water could have been supplied to the 
 city in quantity sufficient for ordinary uses and for effective fire protection for only 
 thirty to forty hours. 
 
 The high consumption also reduced the quantity that could be drawn at or in excess 
 of the limiting capacity of the system from 156,000,000 to 32,000,000 gallons. In 
 other words, it the cold weather and the maximum rate of consumption had persisted 
 five days longer, this small reserve would have been used up; and thereafter until the 
 reservoirs were again filled, any demand in excess of the limiting capacity could not 
 have been supplied. The relations between the abnormal consumption, the tem- 
 perature, and the depleted storage are shown graphically in Diagram No. 2. 
 
 Such a condition as this might arise, in greater or less degree, any winter. There 
 has been nothing approaching the consumption of February, 1905, during the winters 
 since then, as they have been, by comparison, very open and free from severe and 
 long continued cold weather. An examination of the temperature records kept in 
 this office for about forty years indicates that a winter no less severe than that of 1904-5 
 has occurred in about one year out of five. ■ 
 
 Remedies for this condition. — The problem of predicting a recurrence of such a con- 
 dition as this with possible disastrous results is entirely indeterminate in view of the 
 highly indeterminate character of the two factors controlling it, namely, the severity 
 of the winters and the amount of the flagrant waste that goes on at such times. An 
 interruption of the flow in the aqueduct when conditions such as these might prevail 
 is equally indeterminate, but the possibility thereof is set forth elsewhere in this 
 report, and in that possibility lies the contingent one of a disastrous water famine. 
 
 There are two remedies that might be applied, both of which aim at holding the 
 reservoirs as full as possible. This could be done either by lifting the entire supply 
 by means of pumps located at the inlet of the first, or Dalecarlia, reservoir; or by tak- 
 ing proper measures to prevent the extreme waste in^freezing weather, thus holding 
 the total quantity of water demanded down within a reasonable percentage of the 
 normal and maintaining naturally the normal conditions of a storage. 
 
 Pumping station at inlet of Dalecarlia reservoir. — All the functions desired of a 
 pumping station at this point would be fulfilled by an equipment having a capacity 
 of 90,000,000 gallons per day against a head not exceeding 6 feet. The quantity of 
 water pumped can not exceed the limiting capacity of the aqueduct above that point; 
 and the head need not exceed that necessary to lift it up to the elevatiop of the spillway 
 of Dalecarlia reservoir, which is 146.5. 
 
 Preliminary designs have been prepared for a pumping station of this capacity, and 
 are shown on sheet No. 24. The station would consist essentially of three 36-inch 
 vertical shaft centrifugal pumps, each directly connected to a horizontal shaft engine. 
 These would be placed in a bay along the east side of the station and would discharge 
 directly into the reservoir. From the corner of the building a dam would extend 
 south across the inlet channel, and in this dam would be placed gates which would 
 ordinarily be wide open, but would be closed when it was desired to raise the reservoir 
 level by putting the pumps in operation. The engines would be driven by steam 
 generated in a boiler plant in the west half of the building. 
 
 In the construction of this plant the dam, with its gates, could be built complete in a. 
 cofferdam, after which the gates could be left wide open and the pumping station com- 
 pleted ready for operation without interruptirig the flow through the conduit; or 
 the reservoir might be by-passed and drained and the construction carried on with the 
 reservoir out of service throughout the construction period. 
 
WATER SUPPLY OF THE DISTRICT OP COLUMBIA. 43 
 
 The cost of such a pumping equipment is estimated approximately as follows: 
 
 Pumping station building, coal storage, sluice gates, etc $20, 000 
 
 Three pumps and engines complete '. 18, 000 
 
 Two lOO-horsepower steam boilers, with automatic stokers 6, 000 
 
 Stack, piping, valves, crane, etc 12,000 
 
 Engineering and contingehcies 9.i 000 
 
 Total construction 65, 000 
 
 Measures for preventing cold-weather waste. — ^The cold- weather waste that causes the 
 condition now under discussion is a result, as has been stated, pittely of wanton waste 
 from house fixtures. This is entirely under the control of the individual water taker, 
 who, in the absence of a meter on the house service, can carry on this practice unde- 
 tected. If his service is metered, he will not, as a rule, let the water run to prevent 
 freezing of pipes; if he does prefer this method, at his own expense, he will at least 
 waste it judiciously and pay for what he wastes. In the waste problem in the District 
 of Columbia, if there were no meters on any of the services, but the pitometer investi- 
 gations were pushed until the entire city had been covered, the most aggravated 
 form of waste then remaining would be this cold-weather waste from house fixtures, 
 which puts such a drain upon the limiting daily capacity of the system. Using this 
 as the basis for argument, and assuming no other means for protecting the city against 
 possible water famine some winter, the remedy clearly would be a meter on every 
 service in the city. 
 
 Relative costs and advantages. — ^The complete pumping equipment at Dalecarlia 
 reservoir for lifting the water high enough to Keep the reservoir system filled up, would 
 cost $65,000. The annual operating charges it is believed would not exceed $3,000, 
 which, capitalized at 3 per cent, would be $100,000, making a total capitalized cost 
 of protecting the city by this means of about $165,000. 
 
 The cost of applying meters to every service in the District is estimated elsewhere 
 at about $775,000. 
 
 A comparison of these two costs should not be construed as indicating an advantage 
 in favor of the pumping station for the purpose of remedying this cold-weather condi- 
 tion. If this were the one and only benefit to be derived from the universal use of 
 meters, the pumping station would be the'more economical remedy, but the other 
 benefits of metenng are many, as may be seen by referring to that subject. The 
 pumping station project has no advantages over the other, but from its very nature 
 IS designed to encourage the waste of water by furnishing a supply by means of which 
 householders could protect themselves as they do now against the annoyance of frozen 
 pipes. The United States and District governments are not benevolent institutions, 
 and should not in justice be expected to afford protection against structural defects 
 in the houses of water takers, each one of whom could protect his own plumbing at, 
 small expense. 
 
 There are, moreover, certain disadvantages of the pumping-station project in com- 
 parison with the poicy of universal metering: 
 
 (1) The possibility of an excess draft of 40 to 50 per cent over the normal would 
 reduce to a large extent the opportunity for excluding muddy water from the reservoir 
 system in time of freshet. 
 
 (2) It would cause an increased cost of pumping at the filtration plant. 
 
 (3) It would cause an increased cost of filter operation at the filtration plant. 
 
 (4) It would brinjg an- increased burden on the filters at a time when all natural 
 conditions are least favorable for a high degree of purification. 
 
 (5) It would cause an increase in the cost of pumping that portion of the water going 
 to the high services of the District. 
 
 (6) It would cause diminished pressures in the city mains. 
 
 (7) It would not effect the remedy that would be to a large degree effected by the 
 imiversal metering of services, in reducing the maximum percentage excess over the 
 normal consumption and thereby increasing the safe working capacity of the systen;i 
 and postponing the time when new works would be required. 
 
 These are all disadvanta,ge8 caused by abnormal fluctuations in the daily demand. 
 The greatest economy in many respects and the greatest efficiency of the filters may 
 be obtained by uniformity in the daily consumption, and this can be obtained in no 
 way so well as by the universal use of meters. Although it would involve a greater 
 cost, the universal application of meters to house services would produce benefits 
 which are indispensable to the District, none of which would be brought about by 
 the pumping-station project. And these facts, together with the arguments already 
 presented in favor of the universal metering of services for other reasons than the 
 
44 
 
 WATER SUPPLY OF THE DISTKICT OP COLUMBIA. 
 
 abnormal winter consumption, point emphatically to this as the most advantageous 
 policy to be followed for the avoidance of risk of shortage of water during severe and 
 long-continued freezing weather. 
 
 PROJECT FOR ADDITIONAL STORAGE OF WATER SUPPLY. 
 
 The existing aqueduct system includes three reservoirs through which under ordi- 
 nary conditions the water passes successively before going to the filters. The total 
 storage area is about 125 acres. The period of effective storage therein for a flow of 
 65,000,000 gallons per day is about four days. Thte natural purification effected in 
 the three reservoirs is shown in the following table: 
 
 
 Percentage ot 
 turbidity re- 
 maining after 
 passing reser- 
 volra. 
 
 Percentage of 
 bacteria re- 
 maining after 
 passing reser- 
 voirs. 
 
 Percentage of 
 0.10 c. c. sam- 
 ples containine 
 B. coli after 
 passing reser- 
 voirs. 
 
 
 1907. 
 
 1908. 
 
 1997. 
 
 1908. 
 
 1907. 
 
 1908. 
 
 Balecarlia inlet 
 
 100 
 40 
 32 
 25 
 
 100 
 45 
 38 
 27 
 
 100 
 40 
 35 
 13 
 
 100 
 43 
 
 47 
 20 
 
 61 
 68 
 51 
 33 
 
 45 
 
 
 32 
 
 
 34 
 
 
 22 
 
 
 
 It is evident at a glance that there is great value in the storage to which the present 
 supply is subjected. The percentage removal of turbidity and of bacteria is large. 
 With no storage whatever, the operation of the filters would be impracticable during 
 certain periods and the excellent sanitary quality now obtained in the filtered water 
 would be impossible. 
 
 The degree of natural purification thus effected depends upon the length of the 
 period ot effective storajge. The future increase in the rate of flow through the reser- 
 voirs with increasing consumption will cause a reduction in the storage period and a 
 consequent reduction in the efficiency of removal of turbidity and bacteria. When 
 the draft on the system shall have increased to twice the present consumption, the 
 results of all parts ot the purification process, both in the reservoirs and in the filters, 
 will have become decidedly poorer than those now obtained. Besides the purifica- 
 tion due to storage, a large influencing factor in this is the length of the period during 
 which freshet water of high turbidity and high bacteria can be excluded from the 
 system. This becomes shorter and shorter as the draft increases, thus requiring the 
 introduction from year to year of increasing quantities of water of this kind, and it 
 - is this water which yields the poorest quality ot filtered water. It is possible under 
 present conditioijs to exclude freshet water tor about three days, after which the gates 
 must be opened to avoid a dangerous depletion of the reserve supply. This is suffi- 
 cient in many cases to avoid the worst of the turbid conditions. It is rarely enough 
 to avoid all, and frequently tails to exclude the worst ot it. Before the consumption 
 reaches the limit ot the capacity ot the two systems it may be possible to exclude the 
 turbid water tor but little more than one day, which will greatly aggravate the 
 above conditions. 
 
 In order to maintain the high standard ot quality ot the water supply_ of the District 
 which has been set during the three years ot operation of the filtration plant these 
 conditions will demand some remedy, the necessity tor which will increase as the 
 years go "by. 
 
 The advocates ot additional storage have in mind not so much the mere maintaining 
 of present high standards under conditions of increasing complexity, but rather the 
 actual improvement of the quality of the raw water. Much has been said and written 
 upon the relations between typhoid fever and polluted water in Washington. While 
 it is generally accepted that the Potomac water before filtration was not as large a 
 factor in the causation ot typhoid fever in the District as was formerly believed, the 
 probability of the occasional infection of the supply and the consequent appearance 
 ot the disease under the old conditions is denied by no one. And while from the 
 standpoint of the best bacteriologic knowledge the effluent of the filters has in general 
 been excellent and, as tar as could be ascertained by an able commission, not respon- 
 sible for the spread of the infection, still it is recognized that the process of filtration 
 does not produce a sterile water, and that at certain times of the year large numbers 
 of bacteria are found in the effluent of the filters. The storage of a water supply con- 
 taining the specific germs of typhoid fever causes a progressive reduction in its num- 
 bers. Additional storage is therefore advocated as an added safeguard against the 
 possibility of the introduction ot disease-producing organisms into the water supply. 
 
WATEE SUPPLY OF THE DISTRICT OF COLUMBIA. 45 
 
 Storage possibilities investigated. — There are two sites, either of which could be 
 adapted to use as a reservoir of large capacity. One of these, Cabin Johns valley, 
 could be used with eiliier the parallel or the tunnel location. The other, the so-called 
 "Stubblefield site," could be used only with the parallel location. 
 
 The Stubblefield site is a comparatively flat area of some 180 acres lying between 
 the old aqueduct and the Chesapeake and Ohio Canal about 4 miles below Great 
 Falls. A reservoir could be built upon this site by the constniction of earth embank- 
 ments around three sides, somewhat similar to the present Georgetown reservoir. 
 These long embankments or dikes would be the principal item of expense of this 
 project. There would also have to be two gatehouses and a drain. The details of 
 this project are shownon sheet No. 23. 
 
 The designs for this project show a by-conduit extending from inlet to outlet gate- 
 houses, a distance of 4,100 feet. Experience in the operation of the old aqueduct 
 system has shown the necessity for by-conduits around the reservoirs, and most par- 
 ticularly around the one nearest the intake; and their capacity should be great enough 
 to offerilittleor no more restriction to the flow than would be. met in an equal length 
 of mainiaqrieduct under conditions of maximum discharg^- This condition would 
 necessitate the construction of the by-conduit, as shown, some tinie- before the con- 
 sumption should have reached the limiting working capacity of the old and new 
 Systems combined. Its construction at present would be unnecessary, however, and 
 this would effect a saving of $100,000 or more in the first cost of construction. Gate 
 openings should be left in the ra,tehouses and stubs built to facilitate the future con- 
 struction of the by-conduit. The cost of the by-conduit is not deducted fromthe^ 
 estimate, as it would some day be required to complete the system. 
 
 A detailed estimate of the cost of this project is given in Appendix E, a summary 
 of which foUows: 
 
 Summary of estimated cost of reservoir at Stubblefield site. 
 
 Land $36, 200 
 
 Clearing and grubbing 5, 790 
 
 Embankment, pacing, etc 1, 028, 396 
 
 Inlet gatehouse 50, 025 
 
 Outlet gatehouse , 50, 025 
 
 Sluice tower, etc 30, 152 
 
 Engineering and contingencies 180, 000 
 
 Estimated total cost of construction 1, 380, 588 
 
 The valley of Cabin Johns Creek could be turned into a reservoir by- the construc- 
 tion of a dam above Cabin Johns Bridge. This possibility seemed seductive at first. 
 The area of the valley at high- water level of the reservoir is about 230 acres. The dam 
 to form this reservoir could be built for less than $650,000. The important difficulty 
 in the problem lies in the disposition of the freshet flow of the stream. This project, 
 does not purposely contemplate the use of the water at Cabin Johns Creek for the sup- 
 ply. The flow of the stream is probably often less than 5,000,000 gallons per day. It 
 would either flow into the proposed reservoir, however, or else expensive diversion 
 works would be required to prevent this. Any objection to its discharge into the 
 reservoir would be on account either of physical or bacteriological impurities. 
 
 The bacteriological condition of the water of this stream is shown by the following 
 data obtained from the examination of weekly samples covering a period of some 
 seven months; and for the sake of comparison, similar figures for the Potomac River 
 water for approximately the same period: 
 
 
 Cabin 
 Johns 
 Creek. 
 
 Potomac 
 Eiver. 
 
 
 1,350 
 23 
 
 3,230 
 24 
 
 Peroefitagfe of samples of 10 oiibic ceatlmeters or less which contained Bacillus coli. . . 
 
 This indicates the water of Cabin Johns Creek to be at least as good and probably 
 somewhat better in sanitary quality than that of the Potomac River. This considera- 
 tion does not make it either necessary or desirable at the present time to exclude the 
 w^ter of this stream from the reservoir. The increase in population in the future 
 might, make it desirable to do so. 
 
 In physical condition the water of Cabin Johns Creek is similar to that of the Poto- 
 ipac. At ordinary stages it carries but little turbidity, but in trerfiet it carrieaa great 
 deal,. In a general way, maximum turbidity and maximum volume al discharge, are 
 
46 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 coincident; and the flood flow of this stream would at times be sufficient to displace all 
 the clear water in the reservoir with very turbid water in a short time, thus defeatine 
 one of the main objects of additional storage. The ordinary flow of the stream could 
 be easily diverted, but the diversion under the difficult conditions of topography 
 there found of a possible freshet flow approaching 4,000 cubic feet per second would in- 
 volve an expense that would be absolutely prohibitive. The project has been thor- 
 oughly studied and a scheme for the storage and gradual diversion of ordinary freshets 
 worked out in detail. This scheme woulJ not be effective in the total exclusion of the 
 waters of the stream during high or prolonged floods. The excess over the combined 
 capacity of the diversion channel and the flood storage basins would overflow into the 
 reservoir, thus completely defeating at such times, as stated before, one of the impor- 
 tant objects sought in additional storage. 
 
 Designs have been made for all the works necessary to carry out the Cabin Johns 
 Valley storage project along the above lines. Because of the defects in the project, 
 these are not presented with this report, but are filed in the office of the Washington 
 Aqueduct. An estimate shows that its cost, if used in connection with the paraUel 
 location, would be $1,610^102, or about 17 per cent more than the project for storage 
 at the Stubblefield site with the same aqueduct location. 
 
 Storage in Cabin Johns Valley without complete diversion of the waters of that 
 stream would not accomplish the objects desired. A scheme for its partial diversion 
 only would make the cost of the project considerably higher than that for storage at 
 the Stubblefield site. The latter has none of the complicating disadvantages oi the 
 former, but would make an excellent reservoir in every respect. For the additional 
 storage of Potomac water for the supply of the District of Columbia the Stubblefield 
 site therefore offers the best advantages to be found. The reservoir outlined for this 
 site on sheet No. 23 has an area of about 140 acres, which is somewhat larger than the 
 combined area of the three existing reservoirs of the aqueduct system. 
 
 Opinions of authorities upon the question of additional storage. — The value of this 
 portion of the report dealing with the additional storage of the water supply should be 
 increased by the comment of authorities upon this point. During the summers of 1906 
 and 1907 a board of officers of the United States Public Health and Marine-Hospital 
 Service was engaged in an exhaustive study of the typhoid fever situation in the 
 District of Columbia. Through this study they became thoroughly familiar with the 
 question of the storage of the Washington water supply and the benefits resulting 
 therefrom. A request was therefore made of the Surgeon-General of that service for an 
 expression from that board or some member thereof upon the sanitary advantages to be 
 derived from additional storage of the supply. 
 
 In response to this request, the following communication was received: 
 
 Sukgeon^GbnbhaIj, Public Health and Marine-Hospital Service, 
 
 Treasttby Department, 
 Washington, D. C, April 17, 1909. 
 Sir: I beg leave to acknowledge the receipt of your letter of the 23d ultimo, rela- 
 tive to plans for increasing the water supply of the District of Columbia, and request- 
 ing the opinion of the board of officers engaged in the investigations of typhoid fever 
 in the District of Columbia as to the advantages of providing additional storage fo? 
 the water supply of Washington. 
 
 In reply, there is forwarded herewith a .copy of a letter from Passed Asst. Surg, 
 L. L. Lumsden, acting chairman of the above-mentioned board, containing the 
 opinion as requested. 
 
 Respectfully, Walter Wyman, 
 
 Surgeon-General, Public Health and Marine-Hospital Service. 
 Maj. Jay J. Morrow, 
 
 United States Engineer Office, Washington, D. C. 
 
 Public Health and Marine-Hospital Service, 
 
 Hygienic Laboratory, 
 Washington, D. C, April U, 1909.- 
 Sir: Referring to a letter of March 23 from Major Morrow, Corps of Engineers, in 
 jregard to the advisability of additional storage to meet the increasing wa;ter supply 
 of the District of Columbia, and referred on March 29 by the bureau to the chairman 
 of the typhoid fever board for an expression of an opinion, we have the honor to sub- 
 mit the following: 
 
 The results of our bacteriologic studies have shown that from 55 to 86 ner cent of 
 the total bacteria in the Potomac River are removed by storage in the three sedimen- 
 tation reservoirs before the water is applied to the filter beds. 
 
 Besides the removal of the bacteria, the storage clears the water of much sediment 
 which would clog the' sand filters, making necessary more frequent cleaning of the beds 
 
WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 47' 
 
 and BO causing greater difficulty in the operation of the Alters to accomplish the de- 
 sired purification of the water. Believing the storage to be an important part of the 
 system for the purification of the water, we recommended in our report on typhoid 
 fever in the District of Columbia for 1907 (Hygienic Laboratory Bull. No. 44, p. 60) 
 "that in order to furnish a water supply of a satisfactory grade of purity throughout 
 the year, additional storage reservoirs shall be constructed, or a coagulant (alum) 
 shall be used during periods of high turbidity." 
 
 We believe that, with an increase in the water supply, it would be highly advisable 
 to have storage reservoirs of increased capacity, and we would recommend that such 
 increased storage capacity of the reservoirs be made sufficient to give the water a 
 considerably loiter period of storage than is done with the present supply. 
 RespectfuUy, 
 
 L. L. LtJMSDEN, 
 
 Passed Assistant Surgeon, Acting Chairman Typhoid Fever Board. 
 
 The SuROEON-GENERAt, PuBLIC HEALTH AND 
 
 MARINE-HoSPrTAL SERVICE. 
 
 The health officer of the District of Columbia was also asked to express his views 
 on the subject of additional storage, and he submitted the following letter: 
 
 Health Department, 
 Washington, D.'C, April 3, 1909. 
 
 Dear Sir: In reply to your letter of the 22d ultimo, I beg to say that I am heartiljj 
 in favor of action looking toward an increase in the storage capacity of the watfe^ 
 supply system of this District. The terrors of a water famine in a community of this 
 size are so apparent as to render an effort to describe them unneces^sary; and the fact 
 that an increase in the storage facilities would, in proportion to the extent of such 
 increase, diminish the danger of a water famine in event of an accident to the conduit 
 or conduits above the storage reservoirs would alone justify a very considerable ex- 
 penditure for the purpose of providing such an increase. The occasions, however, 
 upon which the storage facilities of our water supply system will be called into use for 
 the purpose of averting a water famine will presumably, in the future as in the past, 
 1)e few and far between. It is important, then, to realize that the advantages to be 
 derived from an increase in such facilities would not be limited to such rare occasions. 
 
 The figures already published by your office show beyond question the extent to 
 which storage alone will remove bacteria and other suspended matter from the water, 
 and demonstrate that an increase in the storage facilities would be of permanent and 
 continuous sanitary benefit to the community. Even though the sources of our water 
 supply may not be at all times polluted with infective organisms, and even though 
 our water supply as it comes into the city may not be responsible for the undue preva- 
 lence of typhoid fever in this community, nevertheless we know that the sources of 
 our water supply are exposed to infection and that after infection occurs it is too late 
 to begin to adopt preventive measures. Increased storage facilities should, there- 
 fore, be provided without waiting for the occurrence either of a water famine or of a 
 water-borne outbreak of disease. 
 
 The fact that we have an efficient filtration plant standing between the city and 
 the danger of a water-borne outbreak of disease does not do away with the desirability 
 for an increase in our storage facilities. The more extensive storage facilities the 
 less will be the work thrown upon the filter beds and the more efficiently and economic- 
 ally can they be operated; and this will be true not only duriiig periods of excessive 
 turbidity of the Potomac River, when a large storage capacity would permit the 
 river water to be shut off until the excessive turbidity had subsided, but it will be 
 true also at other periods.. The fact that increased storage will diminish the work 
 to be performed by the filtration plant will, moreover, postpone the time when it will 
 be necessary to enlarge our filtration facilities, which in any event, with the growth 
 of the community, must sooner or later be done. The increased purification of the 
 water, resulting from the extension of storage facilities, would moreover, in event of 
 the occurrence at any time of any accident that would prevent the use of our filtra- 
 tion plant, render it possible to deliver with greater safety unfiltered water to the 
 community. 
 
 As a protection against water famine, as a measure that will tend to assure at all 
 times a pure, safe water to the community, and as a measure that will render it possible 
 more economically to operate our filtration plant and that will postpone the time 
 when an extension of that plant will be necessary, the establishment of additional 
 storage facilities seems to me to be highly desirable. 
 Very respectfully, 
 
 Wm. C. Woodward, M. D., Health Officer. 
 
 Maj. Jay J. Morrow, 
 
 United States Engineer Office, Washington, D. C. 
 
48 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 COAGULATION AS A SUBSTITUTE FOR ADDITIONAL STORAGE. 
 
 The objects of additional storage capacity may be briefly summed up as follows: 
 
 (1) To compensate for the detrimental effect upon the physical and bacteriological 
 condition of the water due to the shortening of the period of storage with increasing 
 draft on the system, by providing for a longer period of exclusion of undesirable 
 water and a longer period of sedimentation and storage for all water. 
 
 (2) To effect an actual additional improvement, aside from the compensating 
 effect just referred to, in the physical and especially in the bacteriological condition 
 of the water, by the same means above mentioned. 
 
 (3) To increase the reserve available for the supply of the city in the event of an 
 interruption of the flow in the aqueduct above the point of storage. 
 
 The necessity for attaining at least the first of these objects has been elsewhere dis- 
 cussed, and here need only be emphasized again. It is pertinent to consider, how- 
 ever, whether an additional reservoir is the only means or the best means to this end; 
 and also whether any other means suggested would give the improvement mentioned 
 in the second object. 
 
 In the disCuBsion of the increase in the capacity of the existing filtration plant by 
 a simple increase in the rate of filtration, strong emphasis was laid i^on the necessity 
 for the coagulation of the water during occasional turbid periods. The whole subject 
 of the preliminary treatment of the water was investigated and reported upon at 
 length by this office in February, 1908, and coagulation was found to be unquestion- 
 ably the only feasible method. The conclusions of that report have been properly 
 approved. The Medical Society of the District of Columbia has publicly given its 
 indorsement to that method of treatment. A project and estimates for the coagula- 
 ting treatment are ready to be submitted to Congress. In view, therefore, of the 
 necessity for this treatment in connection with increasing the capacity of the filters, 
 its great importance in the improvement in quality of the water during periods 
 of high turbidity, and its unqualified approval by the proper engineering officials 
 and by the medical society, coagulation may be considered an assured adjunct to the 
 aqueduct system and filtration plant in the future. 
 
 Coagulation can be irade to give exactly the objects desired in (1) above by effect- 
 ing the sudden precipitation and removal of a large percentage of both turbidity and 
 bacteria in any quantity in which they are likely to enter the system. In tact, coag- 
 ulation is more flexible and its eflEciency could be more constantly depended upon 
 than could that of any system of reservoirs feasible to construct as a part of the Wash- 
 ington Aqueduct system. This fact is generally recognized, and data are given in 
 detail in support of the case in hand, in the report of February, 1908, on the "Pre- 
 liminary treatment of the Potomac water." Under present conditions it is proposed 
 to coagulate the water only during periods of high turbidity. Under conditions 
 of greatly increased consumption the same specification would hold, only the periods 
 would be longer and more severe. 
 
 These arguments apply with equal fo!hce to the attainment of the second object (2). 
 The investigations and report of February, 1908, were carried out with that specific 
 object in view and the conclusions of that study remain unchanged. The only point 
 in which coagulation, as proposed, could not effect an improvement equal or better 
 than could be brought aijout by additional storage is in the bacteriological condi- 
 tion of the water accompanying low turbidity. It is not proposed to use any coagu- 
 lant at such times. Admitting the efficiency of the coagulation treatment at times 
 of high turbidity, the advisability of additional storage for the purpose of improving 
 the bacteriological condition of the water depends upon the strength of the argu- 
 ments in favor of this particular point. 
 
 The conditions that exist in the aqueduct system during these favorable periods are 
 characteristically as follows: The water is relatively clear. The bacteria in the river 
 water commonly number only a few hundred, and B. coliis not detected in samples 
 as large as 10 c. c. for periods of two or three weeks at a time. In the water that has 
 passed through the three existing reservoirs there are periods as long as four months 
 during which the bacterial counts do not exceed 100 c. c, and frequent periods of a 
 month or longer during which B.coU is not detected in 10 c. c. or less. This is cer- 
 tainly proof positive that there is no large and constant sewage pollution in the unfil- 
 tered water. This water, fair in quality, as it is, is then filtered through filters of the 
 most modern type. More than three years of experience has demonstrated the effi- 
 ciency and reliability of these filters. There is no reason to believe that thev would 
 not give results beyond criticism in the filtration of Potomac water direct from the 
 river under the favorable condition just assumed. 
 
 In view of the excellent purification that can be and is now effected in the pas- 
 sage of the water through the three reservoirs and then through the filtration plant. 
 
WATER SUPPLY OP THE DISTRICT OF COLUMBIA. 49 
 
 it would be a nice question, involving deeper knowledge than we now have|Of th^ 
 laws governing the persistence of disease-producing organisms in natural and artifir 
 cial processes of water purification, to be a,ble to sfiite definitely that the danger p£ 
 infectioQ of the filtered supply was real and great.enough to warrant any such expend- 
 iture as would here be required to provide another reservoir in this aheady efficient 
 gystem. 
 
 The excellent results of filtration, as judged from the, highest bacteriological stand- 
 ards, lead strongly to the belief that this would not be warranted for the sole purpose 
 of improving the bacterial quality of the water in the absence of any considerable 
 turbidity. And since coagulation is proposed only for turbid water, and is all suffi- 
 cient for that condition, an additional reservoir ia not necessary for the attainment 
 of the objects (1) or (2) set forth above. 
 
 The argument of relative cost is strongly in favor of coagulation. The first cost of 
 the completion, construction, and equipment for this process for the old aqueduct 
 system has been generously estimated at $130,000. The maintenance and operating 
 cost, likewise estimated high in order to cover the worst conditions that might occur 
 the first year, is $14,120. This capitalized at 3 per cent is $470,000, making the 
 total capitalized cost not in excess of $600,000. The first cost of the storage project 
 is estimated at considerably more than twice that amount. 
 
 The proposed new reservoir, if constructed, would be 17,700 feet from the intake 
 at Great Falls. This new reservoir, to fulfill the object defined in (3) would, in 
 other words, afford some measure of protection against water famine, due to an acci- 
 dent in less than 25 per cent of the length of the aqueduct system. Considered aa 
 insurance, the cost of this would be enormous, amounting to about $78 per linear foot 
 for the 17,700 feet above the reservoir; and, furthermore, it would be but imperfect 
 insurance, because it would be limited by the available capacity of the reservoir. 
 It the argument of insurance were of great enough importance, perfect insurance 
 could be provided by other means at much lower cost than that. 
 
 From a consideration of the foregoing tacts there 'appears to be no argument that 
 imperatively demands the construction of an additional reservoir in connection with 
 the present or the proposed new aqueduct system. 
 
 LANDS REQUIRED POR THE PROJECTED WORK. 
 
 The lands required for the construction of the proposed aqueduct system are gen- 
 erally shown on the location plans. For the parallel location but little additional 
 land will have to be taken west of Dalecarlia reservoir, because the new line will be 
 located for much of the distance within the present limits of the aqueduct right of 
 way. In order, however, to guard against the possibilitv of damage suits due to con- 
 struction operations requiring more room than lies within these limits, the estimate 
 for land includes a sum large enough to purchase a strip averaging 25 feet wide from 
 the Anglers' Club to Dalecarlia reservoir. There are five small areas, aggregating 
 2.7 acres, that are clearly necessary for the construction, and these are indicated on 
 the location plans. 
 
 East of Dalecarlia reservoir the aqueduct goes directly into tunnel until Rock 
 Creek Valley is reached. It is not proposed to purchase land over the tunnel Kne,. 
 but an amount is included in the estimate as compensation tor tunnel rights. 
 _ Across Rock Creek Valley the location plans and estimates indicate a 100-foot 
 right of way for the aqueduct in cut and cover. This amounts to 6.8 acres. 
 
 In the event of the adoption of the project for additional storage at the Stubblefield 
 site, an area of about 181 acres will have to be taken. Thia ia indicated on the 
 Stubblefield location plan. 
 
 , The fact that it ia considered unnecessary to recommend the immediate construc- 
 tion of the proposed works tor increasing the water supply should not deter the United 
 States from acquiring the lands necessary for these works. Certain of the lands,, 
 especially those lying within the District of Columbia, are constantly increasing in 
 value and should be purchased at an early date. 
 
 SUMMARY AND CONCLUSIONS. 
 
 This investigation of the problem of increasing the water supply of the District of 
 Columbia was undertaken in the belief that the increasing consumption was danger- 
 pusly near to and would soon exceed the safe limit of the capacity of the existing 
 aqueduct system. Up to a year or ao ago thia belief was entirely justified, as the pub- 
 lished figures on the consumption of water clearly show. In the reports of the officers 
 in charge of the Washington Aqueduct in the past several years, recommendations 
 
 H. Doc. 347, 61-2 i 
 
50 WATEE SUPPLY OP THE DISTRICT OF COLUMBIA. 
 
 have been made for various measures designed to meet this condition. Among 
 others were recommendations that steps be taken to suppress the large waste of water. 
 The water department of the District began active prosecution of this work about two 
 or three years ago. Because of the uncertain efficiency of measures for the supptes- 
 Bion of waste, dependence was not placed, however, upon this alone, but the pre- 
 liminary work for the increase of the supply herein reported upon was undertaken. 
 The efforts of more than two years m the elimination of waste have produced 
 results. From these results, whicn are set forth in some detail in Appendix A, the 
 following facts may be stated. A large part of the water supplied to the city is wasted. 
 A great reduction has been effected in this waste. A still further reduction can 
 unquestionably be effected by the means now in the hands of the water department. 
 With this large amount of waste once suppressed, the maintenance of a moderate 
 force for the specific purpose of the examination of the distribution system by the 
 means now in use will insure the city against any considerable increase in the under- 
 ground waste in the future. It is shown, also, that while there exists a tendency for 
 the per capita daily consumption to increase from year to year in most cities, this 
 tendency does not seem to exist in cities that have become thoroughly awakened to 
 the possibilities of waste prevention. These facts warrant the conclusion — 
 
 (1) That the per capita daily consumption and waste of water in the District of 
 Columbia in the next few decades will be somewhat lower than it has been during 
 the last few years. 
 
 The figures given in Appendix A, showing the actual results that have been accom- 
 plished; the high degree of efficiency of the corps engaged in the work of waste elimi- 
 nation, as shown by the current weekly reports of the results of their work; the large 
 area of the city so far untouched by the waste surveys; the consistent reduction from 
 month to month throughout the past year in the quantity supplied to the city from 
 the filtration plant; and the consideration of the question of future population of the 
 District; these facts give safe grounds for the conclusion — 
 
 (2) That the total consumption and waste of water in the District in the next few 
 years will be far enough below the safe working capacity of the existing aqueduct 
 system to make it entirely safe to postpone the construction of new works until fuller 
 and more definite results are available of the measures for the prevention of waste, 
 which will afford a more reliable figure for the future per capita consumption than we 
 are now justified in assuming. 
 
 The consideration by the officer and the assistants in charge of the Washington 
 Aqueduct, and by the writer of this report, of the stability of the structures of the 
 old aqueduct system has led to the conclusion — 
 
 (3) That there is no weakness in the existing aqueduct system sufficiently serious 
 or threatening to justify at the present time a recommendation for the construction 
 of the proposed new works for the purpose of insurance alone. 
 
 Obviously the increase of the supply, though not now necessary, is only a matter of 
 time. One most important measure may be taken in this direction before the ccm-r 
 struction of a new aqueduct is actually undertaken. The application of meters to 
 all services in the District would check a large part of the sudden and abnormal 
 increase in waste that occurs from time to time, especially in extreme cold weather, 
 This would have the effect of reducing the percentage of excess capacity that must 
 be maiptained, which would virtually increase the working capacity of the system. 
 Universal metering is advocated not alone on the grounds just stated. Other well 
 recognized reasons are given at length in Appendix A. In addition to these reasons 
 and their valid bearing upon the reduction of waste and the ideal apportionment of 
 charges against water takers, the importance of the fact above stated justifies the 
 conclusion — 
 
 (4) That the first and most important step that could be taken at the present time 
 for the purpose of increasing the available capacity of the aqueduct system would be, 
 the application of meters to all services in the District. 
 
 'There is a wide variation in the relative costs of construction and of maintenance 
 and operation of the several projects touched upon in this report for increasing the 
 water supply of the District of Columbia. In the case of the gravity projects the 
 construction cost constitutes the larger part. Information is not at hand for the 
 accurate estimation of the cost of maintenance and operation, but the wide difference 
 in the costs of the several projects makes the consideration of this point unnecessary. 
 In the case of the projects involving pumping, the capitalized operating costs alone 
 are in excess of the construction costs of the gravity projects, and the constructloii 
 costs of the former would increase this difference in favor of the gravity supply. 
 
WATEB SUPPLY 01" THE DISTRICT OF COLUMBIA. 
 
 51 
 
 The following table gives the estimated total construction costs of the grayity pr.oj- 
 .ects, and the estimated operating costs of the pumping projects, capitalized at 3 
 .percent: 
 
 Source. 
 
 .I^Otomac at Great Falls: 
 
 Parallel location 
 
 Tunnel location 
 
 igeneoa Creek 
 
 Rock Creek: 
 
 Higli level 
 
 Low level 
 
 (Potomac l?y pumping from- 
 
 Little f-alls 
 
 Tidal basin 
 
 Assumed 
 normal 
 capacity 
 per day. 
 
 Cfallons. 
 76,000,000 
 75,000,000 
 75,000,000 
 
 40,000,000 
 40,000,000 
 
 75,000,000 
 75,000,000 
 
 Estimated 
 cost of 
 project 
 
 construc- 
 tion. 
 
 879,383 
 333,686 
 
 689, oqo 
 
 115,000 
 366,000 
 
 000,000 
 473,000 
 
 a Operating charges capitalized. 
 
 The choice between the two lines from Great Falls, namely the parallel and the 
 tuiinel locations, admits of no clearly drawn conclusion. It is purely a matter of 
 judgment in determining the policy to be pursued on the part of the appropriating 
 committees in Congress. The most that can be given here is an expression of pcrjon^ 
 opinion as to the insurance value there might be in the fact that the two aqueducts, 
 iew and old, were separated by a considerable distance. The alternative line desig- 
 nated as the tunnel, location is presented iu accordance with the areumeiit previously 
 jstatfid — "to reduce to a minimum the chance of a single cause disabling both." The 
 adoption of the tunnel location would afford all the benefits implied in this argunjent, 
 short of the calamity that might result from earthquake disturbajaces. The separation 
 of the two lines might or might not give protection against that. 
 
 On the other hand, the construction upon the parallel location would add materially 
 to the strength of the old system, stable as that now is. Each embankmeiit, under- 
 run by a culvert that is now looked upon as a point of weakness, would be increased 
 in width and consequently in stability. The construction of a ma,SBive concrete 
 masonry aqueduct on the south or lower side of the existing aqueduct would effectually 
 protect the less stable brick ring against the dangers arising from the remote possibility 
 of flood waters overtopping and washing away the aqueduct embankment; and the 
 repairs which it is proposed to make in the old aqueduct as soon as a new one is built 
 Will eliminate the possibility of leaks or collapsing sections over embankments. 
 
 It is pertinent to refer, too, to a point which can hardly be reckoned in an estimate, 
 and yet which is of much financial weight. Compared with work in the open, tunnel 
 construction affords large opportunities for inferior workmanship on the part of un- 
 scrupulous contractors. The experiences with the CrOton Aqueduct, and with the 
 Lydecker Tunnel here in Washington, are sufiiciently illustrative of this possibility. 
 Between Great Falls and Cabin Johns Bridge there are on the tunnel location 29,000 
 linear feet of tunnel, while on the parallel location there are only 9,200 linear feet, or 
 less than one-third as much. In view of these facts the writer of this report feela 
 justified in expressing the opinion that the separation of the two lines by the adop- 
 tion of the tunnel location at an increased expense of nearly $450,000 would not be 
 advisable. 
 
 From the point of view of cost, the choice among the above, if works were soon to be 
 constructed, would lie between the Great Falls gravity project, parallel location, and 
 the low level development from Rock Creek. Since the actual construction, of new 
 works is a problem for the future rather than for the present, a consideration of the 
 relative economical advantages of the two under present conditions would not be 
 applicable at that time, due to the changing conditions in the valley of Rock Creek. 
 The residence district of Washington is unquestionably growing in that direction, and 
 a few years may find the necessary land too valuable for this use. 
 - Mention has been made of the periodic occurrence under existing conditions of high 
 ■bacterial counts and visible turbidity in the filtered water. The increased draft upon 
 the system will shorten the period during which the objectionable water can be ex- 
 cluded from the aqueduct; it will shorten the time for sedimentation and natural 
 •purification in the reservoirs; and in general it will greatly aggravate the conditions 
 as they now exist and make the necessity for some remedy still- more imperative. 
 'There are two remedies for this, namely, additional reservoir capacity and preliminary 
 
52 WATER SUPPLY OP THE DISTKICT OF COLUMBIA. 
 
 treatment by coagulation. Of these two an additional reservoir of the capacity it is 
 feasible to build will not afford a complete remedy for these conditions. It has been 
 clearly demonstrated, both upon other waters and upon the Potomac water, that pre^ 
 liminary treatment by coagulation can be made to give a complete remedy and at a 
 much lower cost. 
 
 In connection with the increase of the supply it has been proven- entirely feasible 
 and safe to operate the filters at rates considerably higher than those for which they 
 were designed. An imperative requirement of this adaptation of the existing plant to 
 higher rates is an adequate form of preliminary treatment. The storage that could be 
 obtained in connection with the ola and the proposed new aqueduct system could not 
 be depended upon at all times to give the desired results. Coagulation would un- 
 questionably give a satisfactory preparation. In a question upon which depends the 
 maintaining of the supply at critical times, there is only one choice, and that the one 
 that will give perfect results. The conclusion is therefore justified — 
 
 (5) That preliminary treatment by coagulation must be considered an essential 
 part of the purification process in the future as applied to the water supply drawn 
 from the Potomac River for the District of Columbia; and since coagulation is more 
 flexible, more certain in its action, and cheaper than additional storage, that the con- 
 struction of a new reservoir for that purpose is unnecessary. 
 
 RECOMMENDATIONS. 
 
 It is therefore recommended : 
 
 (1) That provision be made for the installation of meters on all the services in the 
 District of Columbia. 
 
 (2) That provision be made for the preliminary treatment of the existing supply 
 by coagulation in accordance with the project already prepared for that purpose. 
 
 (3) That in view of the fact that changing conditions may change the relative eco- 
 nomical advantages of the projects, the choice of a source and route for an additional 
 supply be deferred until such time as its actual construction shall be required. 
 
 BespectfuUy submitted. 
 
 Fkancis F. Longley, 
 Assistant Superintendent. 
 Maj. Jat J. Morrow, 
 
 Corps of Engineers, U. S. Army. 
 
 Appendix A. 
 
 THE USE AND WASTE, AND THE FUTURE CONSUMPTION OF WATER IN THE DISTRICT 0» 
 
 COLUMBIA. 
 
 The subjects of this study are taken up in the following order: 
 
 (1) Use and waste of water in the District of Columbia: Legitimate domestic uses; 
 commercial uses; public uses; waste; sufficient quantity for the per capita daily con- 
 sumption from the above considerations. 
 
 (2) Prevention of water waste in the District of Columbia: The use of meters; waste 
 survey of the distribution system: causes affecting the variations in per capita daily 
 consumption in the District of Columbia in recent years. 
 
 {3) Increase in per capita daily consumption. 
 
 (A) Future per capita daily consumption in the District of Columbia. 
 
 (5) Summary of discussion relating to use and waste and per capita consumption 
 of water. 
 
 (6) The future population of the District of Columbia. 
 
 (7) The future consumption of water in the District of Columbia. 
 
 USB AND WASTE OF WATER IN THE DISTRICT OF COLUMBIA. 
 
 A large part of the total quantity of water supplied to the District of Columbia is 
 wasted. It was the intention of the designers of the existing aqueduct system in 
 1853. that the cities of Washington and Georgetown, as they then were, should be fur- 
 nished with "an unfailing and abundant supply of pure and wholesome water." 
 Their estimate of per capita consumption, in a day when 20 to 35 gallons per day was 
 commonly considered ample for this purpose, was set at the extravagant figure of 90 
 gallons per capita per day. This quantity included provision for domestic and pub- 
 lic uses, washing of streets, etc., and upon this figure they predicted the total quantity 
 
WATEB SUPPLiT OF THE DISTBICT OF COLUMBIA. 
 
 53 
 
 seeded by the cities for the forty years from that date. . At the end of that forty years 
 the per capita daily consumption was practically twice as great, and in 1905 it had 
 reached a maximum of 218 gallons or nearly 2.5 times as great as their almost unprece- 
 dented assumption of 90 gallons in 1853. 
 
 Water supplied to a city is disposed of in one of the following ways: (1) Legitimate 
 domestic uses; (2) commercial uses; (3) public uses; (4) loss or waste. 
 
 (1) Leaitimate domeitic uses. — ^The quantity of water chargeable to legitimate domes-, 
 tic uses "haa been investigated, directly or indirectly, in a number of cities. It is 
 dependent apparently upon such factors as the class of dwelling, the general character 
 of the city, and the character of the population; and dependentj too, most logically, 
 upon the proper care and restrictions placed upon the distribution of the water, m 
 which respect the use of meters on service pipes is an important measure. The best 
 analysis available of the quantities of water going to these different uses is probably 
 that of Dexter Brackett for the city of Boston in 1895. His studies, together with the 
 other information at hand, indicate a wide range of quantities for domestic use, vary- 
 ing from less than 10 gallons to nearly 50 gallons per capita per day under fair and 
 reasonable conditions of restriction. 
 
 . The investigations of the water department of the District during the last three 
 years afford interesting information about the per capita consumption. These studies 
 have now covered about one-half the services in the District. In each section studied 
 data are recorded from which the per capita consumption can be determined. A 
 nuumary of these data is given in the following table: 
 
 Separated. 
 
 Accumulated. 
 
 City 
 blocks. 
 
 Per cent 
 of total 
 so far in- 
 vestigated. 
 
 Per capita 
 daily con- 
 sumption. 
 
 City 
 blocks. 
 
 Per cent 
 of total 
 so far in- 
 vestigated. 
 
 Per capital 
 daily con- 
 sumption. 
 
 99 
 65 
 60 
 66 
 70 
 24 
 43 
 
 24 
 16 
 14 
 13 
 17 
 6 
 10 
 
 Oallons. 
 30 
 47 
 65 
 91 
 132 
 165 
 304 
 
 99 
 164 
 224 
 279 
 349 
 373 
 416 
 
 24 
 40 
 54 
 67 
 84 
 90 
 100 
 
 Oallom. 
 30 
 37 
 
 It 
 69 
 75 
 99 
 
 416 
 
 100 
 
 This table shows clearly that the total per capita daily consumption in a large 
 section of the city is low; in 24 per cent of uie city blocks so far investigated it aver- 
 ages 30 and does not exceed 40 gallons; in 40 per cent it averages 37 and does not 
 exceed 60 gallons. These comparatively low figures for total per capita consumption 
 probably represent domestic use in large part; certainly domestic use does not exceed 
 these amounts. They may not be representative of the entire District because the 
 areas first covered by the investigations were the high service areas which are com- 
 paratively new and largely residential. A per capita quantity suflScient tor the 
 domestic uses of so large a part of the population, however, should in all reason be 
 sufficient for the entire population, and therefore the table should afford a good 
 criterion of the actual quantity chargeable to this use for the entire District. It 
 seems probable that 40 gallons per capita per day is suflBcient for this purpose. 
 
 (2) Commercial uses. — ^The per capita quantity of water used for commercial pur- 
 poses depends upon the commercial and manufacturing development of the city and 
 Upon the relative extent to which the public supply and other sources are used for 
 this purpose. The few statistics available on this point indicate the prevalence of 
 figures for a number of cities varying from 20 to 35 gallons per capita. The figure for 
 Boston in 1892 was about 30 gallons; Syracuse, N. Y., 1889, about 30 gallons; Yonkers, 
 N. Y., 1897, about 27 gallons; New York, N. Y., about 24 gallons. 
 ; Special circumstances often make this lower. Fall River, a manufacturing city,: 
 in 1892 had a commercial consumption estimated at only about 2 gallons per capita 
 per day, owii^ to the fact that most of the factories took their own water directly from' 
 the river. 
 
 The development of manufacturing in the District of Columbia is relatively small. 
 During the fiscal year 1908 the total daily quantity of water used by all consumers 
 
64 
 
 WATER SUPPLY OF THE DISTBICT OF COLUMBIA. 
 
 taking about 100,000 gallons per day or more was 2,800,000 gallons, which is equivalent 
 to about 8 gallons per capita per day. This does not include the great many stores and 
 office buildings which would come in this class usiBe less than that quantity. It would' 
 seem as though 20 gallons per capita per day would be sufficient for commercial uSeS 
 in this city. 
 
 (3) Public tises. — Investigations by the District water department in 1905 indicated' 
 a quantity of water supplied to the United States Government buildings amountiilg 
 in round numbers to about 8,500,000 gallons a day, which was equivalent to a, ^er 
 capita daily amount of about 27 gallons. In addition to this there are large quantities 
 chargeable to the various buildings coming under the District government. It has 
 recently been shown that the schools, police stations, engine houses, etc., take large 
 a,niounts, and much water is of course used for street sprinkling, fountains, sewer 
 flushing, etc. 
 
 The quantities of water that have been used for fire purposes in the last few years 
 have been: 1904, 26,800,000 gallons; 1905, 23,300,000 gallons; 1906, 22,700,000 gallons; 
 19Q7, 28,600,000 gallons; the maximum of which is equivalent to less than 0.25 gallons 
 per .capita per day. 
 
 The large item of this use is the quantity taken by the public buildings of the 
 Government and of the District. In the past this has been excessive, much of it' 
 havii^ been unwarranted waste, as the records of the water department clearly show. 
 
 Public uses in other cities are extremely low compared to that in Washington. In 
 Boston in 1892 it was estimated at less than 4 gallons and in Fall River in 1899 at less' 
 than 6 gallons per capita per day. 
 
 In all reason the quantity of water for public uses in the District of Columbia should 
 be greatly diminished by reasonable measures for the suppression of useless waste. 
 Considering this fact, and not losing sight of the legitimate demands of the District 
 and the United States Government in the matter, it seems that 30 gallons per capita 
 per day should be an ample allowance for this purpose, erring on the side of extrava- 
 gance rather than of parsimony. 
 
 (4) Waste. — From the preceding considerations it would seem that an allowance of 
 about 90 gallons per capita per day was ample for the legitimate domestic, commercial, 
 and public uses of this city; that is, if all unnecessary waste of water were suppressed 
 and only that quantity demanded which would serve some useful purpose, even 
 though this actual use be lavish, the total quantity of water required should not 
 exceed, but should probably be somewhat less than 90 gallons per capita per day. 
 This, it will be remembered, was the assumption, then considered highly extravagant, 
 made by the designers of the existing aqueduct system half a century ago. 
 
 The excess of the enormous per capita consumption over this amount in recent years 
 has been waste; has performed no useful services. What this has amounted to may 
 be seen from the following table: 
 
 Fiscal year. 
 
 Total eon- 
 sumption 
 per day. 
 
 Popula- 
 tion. 
 
 Actual 
 per cap- 
 ita daily 
 consump- 
 tion. 
 
 Quantity 
 required, 
 per dav at 
 rate of 90 
 gallon&per 
 capita per 
 day. 
 
 Quantity > 
 wasted . 
 on these 
 assump-' ■ 
 tionsper : 
 day. ., 
 
 1900 . . . 
 
 Gallons. 
 48,600,000 
 50,700.000 
 56,000,000 
 58.300,000 
 63.000,000 
 70.600.000 
 67,700.000 
 66,900.000 
 64,900,000 
 
 296,000 
 300,600 
 306,200 
 311,800 
 317, 400 
 323,100 
 326, 400 
 329,600 
 339,400 
 
 GaUona. 
 164 
 169 
 183 
 187 
 199 
 218 
 207 
 203 
 191 
 
 Gallons. 
 26,600,0C0 
 27,000,OCO 
 27,600,000 
 28.100,000 
 28,500,000 
 29,100,000 
 29,400 000 
 29,700,000 
 30,600,000 
 
 Gallons. 
 21,900,00Cf 
 
 1901 
 
 23,700,009; 
 
 1902 
 
 28,400,0COi 
 
 1903 ; 
 
 30,200,000 
 
 1904 
 
 34,600,000 
 
 1905 
 
 41,500,000 
 
 1906 
 
 38,300,000 
 
 1907 
 
 37,20O,Q0ft 
 
 1908 
 
 34,300,009 
 
 
 
 This table i? based upon the e-stimited populations for the District given elsewherSi 
 in thi? appendix, which are believed to be more consistent than those assumed for this 
 purpo3e in the annual report of the Waghington Aqueduct for 1906. The fifth columfl^ 
 indicates that if waste had been effectively suppressed and the per capita consump;-' 
 tion kept within the 90-gallon limit, the aqueduct system as it now stands would have* 
 a capacity more than double the demand. 
 
^ 
 
WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 55 
 
 , ..Waste, e ven of this magnitude, is not of any great importance while the capacity of 
 Hie aqueduct system is greatly enough in excess of the proper demands to keep up 
 the supply at practically no increase in cost. But when an enormous waste like this 
 Biikes necessary the expenditure of large sums of money for improvements and exten- 
 Hons which would be absolutely- unnecessary in the absence of such waste, then it 
 assumes an importance which can not be too greatly emphasized. 
 ^Washington does not stand alone in this useless and excessive waste of water. A 
 ponsiderable percentage of the total supply is wasted in practically every city, and the 
 waste in such cities as Pittsburg, Philadelphia, Buffalo, Chicago, and Denver, is 
 
 ?irobably egual to or greater than it is here, as may be seen from the following table 
 see also Diagram No. 3): 
 
 city. 
 
 feuflalo. N. Y 
 
 ^alt Lake City, Utah . 
 
 Mttsburg, Pa 
 
 lenver, Col 
 
 ■bany, N. Y 
 
 licaio.IU 
 
 ajhln^ton, D.C 
 
 'hiladelphia, Pa 
 
 ■ ie, Pa 
 
 itrdlt, Mich 
 
 sey citi', N.J 
 
 loston, Mass 
 
 tjtand Rapids, Mich. . 
 Iftltimore, Md 
 
 '»dlBg, Pa 
 
 irln^neld, Mass 
 
 ^veland, Ohio 
 
 ihmond, Va 
 
 imbrtdge, Mass 
 
 loilngton, Del 
 
 Louis, Mo 
 
 .waukee, Wis 
 
 Idhester, N. Y 
 
 m Bedford, Conn. . . 
 inton,Mass 
 
 » Joseph, Mo 
 
 roTldence, R.I 
 
 rymi.Mass 
 
 7oreester, Mass 
 
 ^ewton. Mass 
 
 Sartford, < onn 
 
 f lanfapolis, Miun 
 
 Liowell. Mass 
 
 Itadison, Wis 
 
 Uanchester, N. H 
 
 Fall River, Mass 
 
 iulncy. Mass 
 
 ■KJckton, Mass 
 
 f 
 
 Popula- 
 tion. 
 
 410,000 
 
 70, 900 
 
 294,000 
 
 191,000 
 
 100,000 
 
 232,000 
 
 330,000 
 
 S22,0C0 
 
 67,800 
 
 412,000 
 
 236,000 
 
 614,000 
 
 106,600 
 
 556,000 
 
 93,600 
 
 83,100 
 
 501,000 
 
 112,000 
 
 100,000 
 
 87,800 
 
 728,000 
 
 363,000 
 
 191, 700 
 
 88, 400 
 
 28,500 
 
 100, OCO 
 
 225,800 
 
 85,500 
 
 140, 800 
 
 38,500 
 
 113, OCO 
 
 300,000 
 
 95,500 
 
 27,500 
 
 '65,900 
 
 112,400 
 
 39,300 
 
 62,000 
 
 Total con- 
 sumption. 
 
 Gallons. 
 
 132,000,000 
 
 17,000,0t0 
 
 68,491,000 
 
 40,655,700 
 
 21,377,000 
 
 455, 000, OCO 
 
 66,900,000 
 
 320, 000, OCO 
 
 11,860,000 
 
 68,817,000 
 
 37,500,000 
 
 96,423,000 
 
 14,772,000 
 
 71, 726, OCO 
 
 11,915,000 
 
 9,960,000 
 
 58,880,000 
 
 12,480,000 
 
 11, 065, OCO 
 
 8,3.53,000 
 
 69, 200, OCO 
 
 33, 730, OCO 
 
 16,616,000 
 
 7, 430, OCO 
 
 2, 129, OCO 
 
 7,200,010 
 
 16, 227, OCO 
 
 6, 016, OCO 
 
 9, 434, OCO 
 
 2,316,000 
 
 6, 682, OCO 
 
 17,591,000 
 
 5, 526. OCO 
 
 1,457,000 
 
 3,400,0(0 
 
 4,950,000 
 
 1,414,000 
 
 2, 020, OCO 
 
 Per cap- 
 ita con- 
 sumption 
 per day, 
 1907. 
 
 Oallons. 
 
 322.0 
 
 240.0 
 
 232.7 
 
 214.0 
 
 214.0 
 
 204.3 
 
 203.0 
 
 201.7 
 
 175.6 
 
 166.7 
 
 169 
 
 157.0 
 
 138.6 
 
 129.0 
 
 127.3 
 
 120.0 
 
 117.5 
 
 113.0 
 
 110.0 
 
 95.4 
 
 95.0 
 
 93.0 
 
 86.6 
 
 84.0 
 
 74.7 
 
 72.0 
 
 72.0 
 
 70.5 
 
 66.8 
 
 60.3 
 
 59.1 
 
 68.6 
 
 68.0 
 
 53.0 
 
 51.6 
 
 43.9 
 
 36.0 
 
 36.0 
 
 Percent- 
 age of 
 meters. 
 
 3.4 
 
 3.0 
 15.4 
 
 1.0 
 18.0 
 
 4.2 
 14.3 
 
 S.O 
 
 2.3 
 30.5 
 
 9.0 
 
 S.E 
 30.2 
 25.0 
 
 9.2 
 50.8 
 88.6 
 53.0 
 22.0 
 25.5 
 
 6.6 
 S6.0 
 69.7 
 29.0 
 47.0 
 50.0 
 88.0 
 35.0 
 98.2 
 88.1 
 98.2 
 71.8 
 73.5 
 98.2 
 75.0 
 97.9 
 54. 
 9LS 
 
 ■ This table shows, too, in a general way, the beneficial effect of a reasonable measure 
 for the restriction of waste. The larger the percentage which the metered services are 
 of the total, the smaller in general is the per capita daily consumption. This tend- 
 ency is unmistakable. The question of meters will be dealt with at a greater length 
 Under "Waste prevention." 
 
 , Waste falls naturally under two heads — that which can be prevented and that 
 which can not. The line of demarkation between these two clases is, from the nature 
 of the problem, indistinct. Owing to the intricacies of a system for the distribution 
 Of water to a large community, the enormous length of pipe with Joints every few feet, 
 and the whole system subjected to heavy internal pressure; the fact that the entire 
 iiystem is liable to corrosion from natural causes, and that it is entirely hidden from 
 Bight under the ground, which does not in any way resist, but does mo'^t effectively 
 conceal any tendency to leakage, it is no surprise that some water should be actually 
 lost. The distribution system of the District of Columbia includes nearly 100 miles 
 of water mains from 10 to 75 inches in diameter, about 350 miles of smaller sizes, 
 and some hundreds of miles of house-service pipes. 
 
56 
 
 WATER SrrPPLY OF THE DISTRICT OF COLUMBIA. 
 
 The statistics of certain cities in which more than usual care is taken to account 
 for all the water supplied show a considerable percentage still unaccounted for; 
 The following figures represent this portion for the cities named for the year 1908: 
 
 city. 
 
 Percentage 
 of total serv- 
 ices me- 
 tered. 
 
 Percentage 
 of total con- 
 sumption 
 that 
 passes 
 through 
 meters. 
 
 Percentage 
 
 of total 
 
 which Is not 
 
 accounted 
 
 for. 
 
 Milwaukee, Wis 
 
 96.0 
 100.0 
 94.0 
 98.2 
 98.2 
 86.6 
 98.2 
 88.6 
 
 E8.5 
 62.8 
 72.4 
 79.2 
 80.4 
 74.6 
 85.4 
 79.0 
 
 31 
 
 Yonlrers.N.Y 
 
 37 
 
 Brockton; Mass 
 
 23 
 
 Worcester, Mass 
 
 19 
 
 Hartford, Conn 
 
 18 
 
 Woonsooket, K. I 
 
 14 
 
 Fall Elver, Mass 
 
 13 
 
 Newton, Mass 
 
 11 
 
 
 
 The figures in the last column are obtained by subtracting from 100 per cent those 
 in the second column, after correction, on the assumption that the quantity passing 
 through the small remaining portion of unmetered services is proportional to that 
 through the metered services. There are many reasons given for the inability to 
 account for the entire supply. Among these are certain jjublic uses, such as tor parks, 
 fountains, watering troughs, street cleaning, sewer flushing, fire protection, etc., and 
 also certain other undesirable and uncontrollable losses, such as leakage from street 
 mains, etc., leakage from reservoirs, and stealing. 
 
 In those cities having the Urger percentages unaccounted tor, the explanations offered 
 for this fact indicate that a considerable portion should, if measured, be included in 
 "public, uses. " Beyond this, these figures afford a measure of the unavoidable waste 
 that has taken place in these cities. After making some allowance for "public uses, " 
 etc., these figures indicate that a city should by proper measures be able to restrict its 
 unavoidable waste to less than 25 per cent of the total supply, and 33 per cent for this 
 purpose would seem a most generous assumption. 
 
 (5) Sufficient quantity for the per capita consumption from the preceding condderor 
 tions. — On the basis of 90 gallons per capita per day for the legitimate domestic, pub- 
 lic, and commercial uses of the community, and of one-third of the total supply toi 
 waste under reasonable conditions of restriction, the total tor consumption and waste 
 in the District of Columbia should not exceed 135 gallons per capita per day. 
 
 PEEVENTION OF WATER WASTE IN THE DISTRICT OP COLUMBIA. 
 
 Actual loss of water may take place due to any one of the following conditions: 
 , (1^ Flagrant waste from faucets, hose, etc. 
 
 (2) Defective plumbing in houses. 
 
 (3) Leaks in mains or service pipes. 
 
 In addition to these there may be an apparent loss of water due to slippage in the 
 pumps where stated quantities are determmed from plunger displacement. This is 
 sometimes a serious matter, but affects the economical operation of the pumping sta- 
 tions more directly than it does the matter of capacity of the supply. AS^ter may also 
 be stolen by large consumers, by by-passing meters, unauthorized use of fire mains, etc. 
 This water, though surreptitiously taken and not paid for, may serve some useful pur- 
 pose, and hence may not be actually lost; but it is an abuse that merits the attention 
 of the water registrar if detected. 
 
 (1) Use of meters. — ^The use of meters on house services puts an effective control 
 on the first two items of loss noted above. With a meter on the service pipe, substan- 
 tially all the water going into a house is measured, to be paid tor by the taker, whether 
 it be used tor proper domestic purposes, whether it wastes from leaky fixtures, or 
 whether it be willfully wasted as is so frequently done by allowing it to run all night 
 in winter weather to prevent freezing, and in summer by considerable waste in the 
 effort to obtain cooler water by "letting it run." Metering a house does not imply 
 any desire upon the part of a water department to restrict the use, even the lavish 
 use of the water, but merely to let each taker assume the burden of any extravagant 
 waste that may exist upon his premises. A taker who uses only reasonable care in 
 this respect need not suffer by paying for his water by measure. In view of the mil- 
 
WATER SUPPLY OF THE DISTEIOT OF COLUMBIA. 
 
 57 
 
 lioasof dollara invested in the system of water supply of a large city that is, in fact, the 
 only Ic^cal basis upon which he should pay for it. 
 
 The use of meters has an unmistakable tendency to reduce the per ca,pita consump- 
 tion of water. This is shown by the table of consumption in various cities for the year 
 1907 (p. 55), which indicates clearly that the cities in which a large percentage of the 
 services are metered have eenetally the lowest per capita consumption. It is shown 
 more specifically in the following table, which gives the reductions in per capita con- 
 sumption effected In a number of cities by the use of meters: 
 
 City. 
 
 Cleveland, Ohio 
 
 Orapd Baplds, Mich 
 
 Rlohmond, Va 
 
 ililwaulcee, Wis 
 
 New Bedford, Mass . 
 
 Hartford, Cionn 
 
 Ulnneapolis, Minn. . 
 Lowell, Mass 
 
 Year. 
 
 1900 
 1908 
 1899 
 1908 
 1890 
 1907 
 1891 
 1903 
 1899 
 1908 
 1900 
 1907 
 1899 
 1908 
 1893 
 1907 
 
 Per eent 
 of 
 
 meters. 
 
 Per 
 capita 
 daily 
 con- 
 sump- 
 tion. 
 
 OaUom. 
 
 174 
 135 
 168 
 113 
 112 
 80 
 107 
 84 
 85 
 59 
 94 
 68 
 83 
 68 
 
 Reduc- 
 tion per 
 capita 
 per day. 
 
 aallmi. 
 I 70 
 
 } 38 
 
 55 
 
 32 
 
 36 
 25 
 
 Opinions vary as to the distinct advantage to be gained by the universal use of 
 meters on the service pipes of a city. One argument frequently used follows from an 
 examination of curves snowing the general relation existing between the per capita 
 consumption and the percentage of services metered, such as the one given under 
 ''Use and waste of water, " for thirty-eight cities for tiie year 1907. It is in substance, 
 that the use of a small percenta^ of meters, judiciously placed, will effect a very 
 marked reduction in the per capita consumption; and beyond that the reduction ber 
 cornes less and less for a given additional percentage of meters. The curves might 
 logically lead to the question as to whether there were in the general case enough ad- 
 vantage to be derived from the extension of meters beyond 40 to 50 per cent of all 
 services to warrant the expense of that extension and in fact, an adverse view is not 
 uncommonly taken upon this point. It is pertinent to repeat here that such curves 
 as these do not indicate a definite relationship that, because of any existing conditions, 
 should exist between per capita consumption and the percentage of metered services; 
 but rather the average relationship that actually happens to exist under the varying 
 conditions in different cities. They do not indicate the actual reduction in per capita.' 
 consumption that would follow upon the progressive installation of meters, nor the 
 actual numerical values we should expect in any case. Generally speaking, it would 
 seem that the quantity of waste suppressed should be approximately proportioned to 
 the number of meters installed. The suppression of waste on the services of manufac- 
 turers and other large takers would undoubtedly be relatively large, but beyond that 
 it would be most diflBcult to discriminate between the 50 per cent of takers whose 
 waste was serious and the 50 per cent whose waste was not worth suppressing. This 
 view of the case has been taken, too, by investigators of the water- waste problem in 
 Pittsburg, who have predicted a reduction dimini^ing slightly for successive increases 
 in the percentage of metered services, but not departing very much from the propor- 
 tional relation referred to above. There should be perhaps an even greater tendency 
 for this to hold true in Washington, where there are but few manufacturers and other 
 large takers compared to most cities of its class. 
 
 The most important argument in favor of universal metering of services in the Dis- 
 trict is touched upon elsewhere. Under existing conditions of waste, the maximum 
 daily consumption may at times be 40 to 50 per cent in excess of the mean daily con- 
 sumption for the year. _ During the fiscal year 1905 there occurred an excess draft of 
 about 45 per cent. This means that a large percentage of the capacity of the system 
 must lie entirely useless except when drawn upon to supply this enormous and un- 
 warranted waste; yet, useless as it is, the excess capacity must be maintained under 
 
58 
 
 WATEE SUPPLY OF THE DISTRICT OF C01.TJWBIA. 
 
 these extravagant conditions, else whan theaJmormal draft does occur the supply will 
 not equal the demand. Plainjjr.in order to secure the benefits of as large a percentage 
 as possible of the capaeiiy rtfthe aqueduct system, the essential thing to obtain is uni- 
 formity of demand tor the water in the city. This will be accomphshed better than 
 id any otber way by the application of a meter to every service m the District. In 
 fact, that is the only way. 
 
 To illustrate the benefit of this, let it be assumed that the universal use of meters 
 would reduce the maximum variation above referred to from 45 per cent to 15 per cent. 
 With the limiting capacity of the aqueduct system, 90,000,000 gallons per day, that 
 would be equivalent to raising the safe annual average that could be depended upon 
 from 62,000,000 to 78,000,000 gallons per day. And this would virtually have the 
 effect of increasing the working capacity of the system by 25 per cent. The effect of 
 this argument upon the necessity for the construction of new works needs no comment. 
 
 The important advantages of the universal use of meters, then, would be as follows: 
 
 (1) It would suppress the waste from house fixtures, which in the absence of the 
 meters would be entirely undetected. 
 
 (2) It would give the water department of the District the only reasonable and 
 logical basis upon which to determine proper charges against all water tiker.s. 
 
 (3) It would afford the greatest attainable degree of uniformity in the daily con- 
 sumption which would virtually have the effect of increasing the working capacity of 
 the aqueduct system by a substantial percentage. 
 
 The record of the number of meters in the District in the last few years has been as 
 follows: 
 
 Year. 
 
 Service 
 connec- 
 tions. 
 
 Water 
 
 meters 
 
 in service. 
 
 Percentage 
 
 wtiioli 
 
 metered ' 
 
 services , 
 
 are of total. 
 
 1905 
 
 64,035 
 65,721 
 67,672 
 69,532 
 
 2,104 
 2,401 
 -8,267, 
 12,611 
 
 3.9 
 
 1906 
 
 4.3 
 
 1907 
 
 14.3 
 
 1908 
 
 21.2 
 
 
 
 In 1905 and 1906 there were comparatively few metered services in the Dietricti 
 Following upon the abnormal consumption of thoee years the water departnent 
 began to install meters at the rate of about 5,000 a year, but since 1908 there ha\e been 
 no funds provided specifically for that purpose. They now have authority to use for 
 installing meters only the surplus from the regular maintenance appropriations. 
 This is being ueed to install meters on all new houses, as far as possible, especially in 
 the high-service districts. The funds are not more than sufficient to keep up with 
 the new services, however, and there will therefore continue to be some 47,000 unme- 
 tered services in the District unless money is provided for their installation. 
 
 The cost of purchasing and installing meters in the District for the lait year or two 
 has been about 116.50 each. At this price it would cost about $775,030 to equip every 
 service in the city with a meter. Estimates for the installation of meters are sub- 
 mitted for the next fiscal year by the District water department, and they merit the 
 unqualified approval and support of Conjren. 
 
 (2) Waste survey of distribution system. — The detection of leakage in the mains and 
 service pipes involvBS either the examination of the outside of the pipes for visible 
 evidence or its indirect determination by the measurement of the flow at a great many 
 points on the system. The examination of the outride of hundreds of miles of pipe 
 buried in the ground is obviouijly out of the question. To accomplish this purpose,' 
 then, we are limited to the second means, which also is a difficult proposition. 
 
 This is effected in practice by the isolation of a convenient district, shutting all 
 valves and cross connections except one, through which all the water supplying that 
 district must pass. The flow throuofh this is then determined. If there is waste as 
 indicated by a large flow during the night, this is localized by shutting off succes; 
 sively different portions of the district and observing the effect on the rate of flow 
 through the main inlet. If a portion in which there is much wa,;te be shut off, there 
 will be a corresponding reduction at the principal point of measurement. The waste 
 maybe still further localized to the individual house services, which are, of course, the 
 limit of such an investigation. 
 
 The flow into the isolated district may be determined by either of two devices, the 
 Deacon waste meter and the Cole-Flad pitometer. Both of these record, automatic- 
 ally and continuously, the rate of flow through the pipe to which they are attached'.. 
 
WATER SUPPLY 01' THE DISTRICT OF COLUMBIA. 
 
 5Bi 
 
 The Deacon device is of thirty-five years' standing, hut it is very expensive and mot 
 largely used in this country. The pitometer is a development of recent years, and is 
 (icnuing into common use with its eflSciency well demonstrated.- The pitometer has; 
 been used in the recent Washington investigations. . 
 
 The work of the pitometer division of the jBaatribet water department commenced ini 
 July, 1906. It has heen pursued conlanuoualy since that time, and has produced: 
 excellent results. JBetween July 1, 1908, and January 16, 1909, the following leaks) 
 i^mB-dsbexstieA and repaired: 
 
 Hi broken services, hydrant connections, etc.: 
 
 Less tian 1,000 gailons per day 
 
 1,000 to 2,000 gallons per day 
 
 2,000 to 3,000 gallons per day 
 
 3,000 to 5,000 gallons per day 
 
 ' 6;0D0 to 10,000 gallons per day 
 
 10,000 to 20,000 gallons per day 
 
 llore than 20,000 gallons per day 
 
 Imbroken andleaking mains, fire hydrant valves, etc 
 total - 
 
 Number 
 of leaks. 
 
 99 
 64 
 67 
 
 105 
 86 
 
 114 
 40 
 
 Total 
 quant ty 
 per day. 
 
 Oallons. 
 
 . 43,800 
 
 67,000 
 
 143,700 
 
 348,600 
 
 £26, too 
 
 1,489,100 
 
 1,044,000 
 
 565 I 3,663,200 
 44 i 613,700 
 
 609 , 4,276,900 
 
 Avera^ 
 quant ty 
 per leaki 
 per day. 
 
 Qallons. 
 
 440 
 
 1,240' 
 
 2,r40.' 
 
 3)320' 
 
 6,1201 
 
 13,0/0' 
 
 26, 100 
 
 6,490 
 13,940 
 
 7,020 
 
 A summary of the results since the investigations began in 1906 is as follows: 
 
 
 Class 1. 
 
 Class 2. 
 
 Class 3. 
 
 Year. 
 
 Num- 
 ber of 
 leaks. 
 
 Quantity 
 per day. 
 
 Average 
 quantty 
 per leak 
 per day. 
 
 Num- 
 ber of 
 leaks. 
 
 Quantty 
 per day. 
 
 Average 
 quant ty 
 per leak 
 per day. 
 
 Num- 
 ber of 
 leaks. 
 
 Quant'ty 
 per day. 
 
 Average 
 quantity 
 per leak 
 per day. 
 
 1903-7.. . 
 
 20 
 205 
 565 
 
 Gallons. 
 
 343,000 
 
 3,0 6,000 
 
 •3,670,000 
 
 Gallons. 
 
 3 
 
 55 
 
 , 44 
 
 Gallons. 
 
 149,000 
 1,188,000 
 
 614, 000 
 
 Gallons. 
 
 6 
 7 
 
 GalUms. 
 
 422,000 
 
 2,338,000 
 
 Gallons. 
 
 100 r 8 
 
 
 
 ^ 
 
 IWM (partial).- 
 
 
 
 
 
 
 
 
 ■; ■ Total 
 
 790 
 
 7,069,000 
 
 8,9.0 
 
 102 
 
 1,951,000 19,100 
 
 13 
 
 .2,760,000 
 
 212,00fl^ 
 
 RECAPITULATION. 
 
 Number of leaks , 905 
 
 Quantity per day gallons. . 11, 780, OOO 
 
 Xrerage quantity per leak per day do 13, 000 
 
 Class 1 includes all broken or leaking service pipes, defective wiped joints on service 
 pipe connections, defective goosenecks on corporation cocks; hydrant connections, 
 etc., and, in general, all points between the mains and the buildings served. 
 
 Claw 2 includes all breaks or leaks in mains and flre hydrants, fire-hydrant valves, etc. 
 
 Class 3 includes all other sources of waste detected, such as extraordinary flow to 
 gtrvernment buildings, schools, police stations, parks, and other District property; 
 improper use by large consumers, etc. In other words, in the two and one-half years 
 that these investigations have been under way there has been a suppression, accord- 
 ing to the records of this division of the water department, of nearly 12,000,000 gallons 
 per day of useless waste. 
 
 This estimated saving is quite consistent with the actual reduction in the total con- 
 flumption. The year 1904^5 was abnormally high, due to the extreme cold-weather 
 consumption in January and February. Aside from that year the average yearly' 
 increa'ie in the total consumption for six years had been at the rate of 3,200,000 gal-i 
 tens per day. At this rate, figuring from July, 1906, when the pitometer investiga^ 
 tions began, the present total consumption would have been about 77,000,000 sallonS 
 per day. The mean consumption for this fiscal year will not exceed 65,000,000 gal-f 
 tens per day, and will probably be slightly less, and this difference is approximately 
 Sqtistl to the reduction m wa=te as estimated bv the water department. 
 
 The records of the pitometer division do not show directly just what percentage of 
 fflte total area of the distribution system has been covered by their investigations. Ans 
 
60 WATER SUPPLY OP THE DISTRICT OF COLUMBIA. 
 
 examination of their maps seems to indicate that they have covered about two-thirdB 
 of the area. An indirect computation from their population statistics would seem to 
 indicate that their work so far had included about one-half of the total number of 
 house services. These two estimates may be consistent and approximately true, as 
 the area so far covered has included most of the newer outlying districts where the 
 area tor a gj ven number of house services is relatively larger than in the older parts of 
 the city still to be covered. 
 
 It is the inteiition of the water department of the District to continue these most 
 profitable investigations as long as they show any efficiency in the suppression of waste. 
 
 (3) Causes affecting the variations in per capita daily consumption in the District of 
 Columbia in recent years. — ^The increase in per capita consumption in Washington up 
 to the year 1903-4, as shown in the table under "Waste of water," seems explainabte 
 by the arpfuments elsewhere set forth regarding the general tendency in all cities 
 toward this increase. The per capita consumption reached its highest mark in the 
 year. 1904^5. The high average for that year was due to the abnormal cold weather in 
 February, 1905, which made the maximum draft on the system nearly 100,000,000 
 gallons per day, corresponding to a per capita for consumption and waste in excess of 
 300,000,000 gallons. The effect of this cold-weather draft uj)on the total daily con- 
 sumption in comparison with the total for the following year is evident from the fact 
 that while the total for 1904-5 was 70,600,000 gallons and that for 1905-6 was only 
 67,700,000 gallons per day, the average for the nine months, exclusive of the winter 
 months, for 1904-^ was 67, 100,000 gallons and for 1905-6 was 68,300,000 gallons. There- 
 tore since the installation of meters and the waste investigations did not commence dur- 
 ing the latter year, the decrease in per capita consumption for that year can be ascribed 
 only to the natural cause of a less severe winter. The winters since then have all 
 been more open. In 1904-5 there were 18 consecutive days on which the temper- 
 ature at 8 a. m. was not above freezing. In 1905-6 the longest consecutive freezing 
 period was 8 days; in 1906-7, 6 days; and in 1907-8, 5 days. We may further con- 
 clude then that the decrease in consumption effected by measures for the suppressioii 
 of waste have not been complicated by any extreme cold weather consumption. 
 
 Both the measures intended to reduce the waste, namely, the installation of meters 
 and the jjitometer investigations, were begun early in the fiscal year 1906-7. From 
 the effectiveness of these measures, and in wie absence of any other known contributing 
 causes, it is reasonable to conclude that the decrease in per capita daily consumption 
 since 1905-6 has been the direct result thereof. 
 
 INCREASE IN PER CAPITA CONSUMPTION. 
 
 The statistics of water consumption in practically every city in the United States 
 show an unmistakable tendency to increase from year to year. This is due to the 
 increased numbers of faucets, bathtubs, water-closets, etc., in modem houses; to the 
 tendency toward better service pressures; to the advances in sanitary science, etc. 
 The increased number of house fixtures and the greater pressures tend to increase 
 both the legitimate consumption and the useless waste. A measure of this increase is 
 complicated by the question of meters, as the percentage of metered surfaces is growing 
 larger in most cities. In order to eliminate this factor, the increase has been dem- 
 onstrated by the following procedure. 
 
 Figures giving the per capita daily consumption and the percentage which the 
 metered services are oi the total number of services have been obtained from some 
 thirty-five to forty cities of the United States tor the period 1890 to date. These have 
 been divided into periods of five years each. All the figures for each of these periods 
 have been plotted with the percentages of meters as abscissae and the per capita daily 
 consumptions as ordinates. The general distribution of these points may be seen from 
 the diagram already given showing per capita consumptions, etc., tor thirty-eight 
 cities for the year 1907. By an application of the principles of the method of least 
 squares, the most probable line passing through each of these series of points has been 
 constructed. Each one of these lines therefore shows for that period the average 
 relation between the per capita consumption and the percentage of metered services; 
 They do not indicate a definite relationship that, because of any existing conditions, 
 should exist between per capita consumption and the percentage of metered services, 
 as the deviations from the curves are for some cities very wide; but they do indicate 
 the average relationship which actually does exist under the widely varying conditions 
 in different cities. 
 
 In Diagram No. 4 the average lines thus determined for the four periods have been 
 plotted, omitting, to avoid confusion, the several hundred points from which they 
 were determined. They indicate that for any given condition as regards the use of 
 meters there has been a large increase in per capita consumption from each five-year 
 
62 
 
 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 ■lor the per capita daily consumption in this city under reasonable conditions for thp 
 (prevention of waste. The dia^m given under "Increase of per capita consump- 
 tion" indicates that there is little, if any, tendency toward this increase in cities that 
 make determined efforts for the suppression of waste; and the experience of many 
 cities in the past decade bears out this belief. A per capita daily quantity of 135 
 gallons therefore seems to be clearly indicated as sufficient for the needs of the Dis- 
 trict for some time to come, on condition that the measures for the prevention of waste 
 be vigorously prosecuted. 
 
 SUMMARY OP DISCUSSION RELATING TO USE AND WASTE AND PER CAPITA CONSUMPTION 
 
 OP WATER. 
 
 The present consumption of water in the District of Columbia is extravagantly higl^ 
 amounting to nearly 200 gallons per capita per day. It has been clearly demonstrated 
 that a large part of this is useless waste. A generous consideration of the actual needs 
 ■of the community, based upon existing facts and conditions, leads to the conclusion 
 that 90 gallons per capita per day should be ample for all the legitimate uses of the 
 District. Besides legitimate use there is in every city a variable amount of waste. 
 Some of this can be prevented by proper measures, but not all of it. On the condition 
 that these measures are vigorously carried out, the unsuppressed waste in the District 
 rof Columbia should certainly be limited to one-third, of the total consumption, which 
 would make the total reasonable requirements of the District amount to 135 gallons 
 per capita per day. 
 
 The measures which should be taken for the suppression of waste are the metering 
 •of services and the surveys for the detection and suppression of waste on the distribu- 
 tion system. About one-fifth of the services in the District are now equipped wift 
 meters. This is not enough to bring about the benefits that should be derived from the 
 meter system. The pitometer surveys for the detection of waste have covered about 
 half the city and are being carried on under the ordinary appropriations of the water 
 department 8f the District. In order to positively reduce the present enormous waste 
 fl,nd to maintain it within the reasonable limit stated above, that department should 
 be given unstinted help in this work. The good showing they have made so far with 
 the limited funds at their disposal has been a positive demonstration of the results that 
 may finally be accomplished under favorable conditions. 
 
 The work of waste prevention has so tar resulted in a saving of about 12,000,000 
 gallons of water per day, with about one-half of the distribution system still to be 
 covered. There should in the end be a total suppression of at least 24,000,000 gallons 
 per day, which would reduce the per capita consumption to about 135 gallons per day, 
 the same figure reached above from a consideration of the reasonable requirements of 
 ithe community. 
 
 THE FUTURE POPULATION OP THE DISTRICT OP COLUMBIA. 
 
 It is difficult to make a reliable estimate of the future population of any city, because 
 it is influenced so greatly by local conditions. The two factors that frequently cause 
 irregular or abnormal increases in population are (1) acquisition of new territory, by 
 annexation or otherwise; (2) extraordinary commercial or manufacturing activity. In 
 the case of the District of Columbia the territorial limits of the population are defi- 
 nitely fixed by the boundaries of the District, and the nature of a city given over gp 
 largely to governmental functions is such as to preclude the likelihood of commercial 
 or manufacturing activities that will greatly alter the natural increase in population. 
 
 A statement from the Director of the Census gives the following populations for the 
 area included wiliiin the territorial limits of the District as they now exist: 
 
 Census year. 
 
 1800 
 1810 
 1820 
 1830, 
 1840. 
 1850. 
 
 PoDula- 
 tion. 
 
 «,144 
 15,471 
 23,33fi 
 30,2el 
 33,745 
 51,687 
 
 Increase 
 (iu^i^g 
 decade. 
 
 7,327 
 7,865 
 6,925 
 3,4S3 
 17,942 
 
 Percent- 
 age of 
 increase. 
 
 Census year. 
 
 1860. 
 1870. 
 1880. 
 1890. 
 1900. 
 
 PoDula- 
 tion. 
 
 75,080 
 131,700 
 177,624 
 230,392 
 278,718 
 
 Increase 
 during 
 decade. 
 
 23,393 
 56,620 
 45,924 
 62,768 
 48,326 
 
 Percent- 
 age of 
 increase, 
 
 45 
 75 
 
 30 
 21 
 
WATEE SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 63 
 
 i The estimates of population made by the police department in recent years are ^ven 
 tmlow: 
 
 Tear. 
 
 Fonula- 
 
 tiE: 
 
 Estimated 
 
 from the 
 
 United 
 
 States 
 
 census. 
 
 Year. 
 
 Popula- 
 tion. 
 
 Estimated 
 from the 
 United 
 States. 
 census. 
 
 |g85 
 
 203,459 
 218, 157 
 .258,431 
 
 204,000 
 220,000 
 240,000 
 265,000 
 264,000 
 
 1905 
 
 323,123 
 326,435 
 329,591 
 339,400 
 
 304,000 
 
 Jg88 
 
 1906 
 
 309,000 
 
 1892 
 
 1907 
 
 314,000 
 
 J896 
 
 270,619 
 277,782' 
 
 1908 
 
 320,000 
 
 1897 
 
 
 
 
 
 The population for corresponding years, estimated approximately from the United 
 States census figures, is also given in this table. The populations indicated by the 
 police censuses since 1892 are several thousands greater than indicated by the estimate 
 made from the United States census of 1900. Plainly, they can not both be right. To 
 judge of the relative accuracy of the two sets of figures, we have from 1885 to the present 
 year two figures from the United States census and nine figures from the police census. 
 These nine figures are quite consistent among themselves. The enumerators em- 
 ployed to obtain them have been, in general, the- policemen on their own beats, whose 
 acquaintance with the houses and inhabitants of their several localities might probably 
 be much better than of census enumerators, who were comparative strangers to the 
 localities. 
 
 It is believed, therefore, that the figures of the police census more nearly represent 
 fhe true population of the District than do the others for the period covered. Inter- 
 polation between the figures given by the police census gives populations of about 
 238,000 for 1890 and 295,000 for 1900, which are, respectively, about 8,000 and 16,000 
 greater than indicated for those years by the United States census. 
 
 A population table can therefore be constructed, using the United States census 
 figures up to include 1880 and the two figurea given above for 1890 and 1900. 
 
 CeneUB year. 
 
 Popula- 
 tion. 
 
 Increase. 
 
 Percent- 
 age of in- 
 crease. 
 
 Census year. 
 
 Popular 
 tion. 
 
 Increase. 
 
 Percent- 
 age of In- 
 crease. 
 
 1860 
 
 75,080 
 131,700 
 177,624 
 
 
 
 1890 . . 
 
 238,000 
 295,000 
 
 60,000 
 57,000 
 
 34 
 
 1870 
 
 57,000 
 46,000 
 
 76 
 36 
 
 1900 
 
 24 
 
 IffiO 
 
 
 
 
 
 The increase during the four decades, 1860 to 1900, has been fairly uniform. The 
 maximum increase during any decade was 60,000, the minimum 46,000, and the 
 average 55,000. The future population of cities is sometimes estimated by assuming 
 a fixed percentage increase which has been found to be true in those particular cases 
 for several successive decades. But a glance at the column above, showing percent- 
 age increases in the population of the District of Columbia through successive decades, 
 shows plainly that there would be no justification for such a method of estimating 
 in the case under consideration. 
 
 In the absence of any constancy in the percentage rate of increase in the District 
 of Columbia, a logical procedure would seem to be to construct a table of estimated 
 future populations based upon the comparatively uniform numerical increase that 
 has taken place during the last four decades, maximum, minimum, and average. 
 
 ji-^fstimated future population of the District of Columbia upon above assumptions. 
 
 • Year. 
 
 Maximum.' 
 
 Minimum. 
 
 Average. 
 
 Year. 
 
 Maximum. 
 
 Minimum.' 
 
 Average. 
 
 1900 
 
 295,000 
 356,000 
 415.000 
 475,000 
 
 295,000, 
 341,000 
 387,000 
 483,000 
 
 295,000 
 350.000 
 405,000 
 460,000 
 
 1940.. 
 
 535,000 
 595,000 
 655,000 
 
 479,000 
 525,000 
 571,000 
 
 615,000 
 
 1910 
 
 1950 
 
 670,000 
 
 1920.. 
 
 1960 .... 
 
 625,000 
 
 1930.... 
 
 
 
 
64 
 
 WATEB SUPPLY OF THE DISTHIOT OF COLUMBIA. 
 
 The increase in population of large cities from one decade to the next is often an 
 increasing quantity. This is true of New York, Chicago, Philadelphia, and other 
 cities, especially where their importance as great commercial centers has given 
 impetus to their growth. It is not generally true, however. The probable popula- 
 tion of the District in 1910, as estimated from the police census from 1900 to 1908, 
 will correspond closely to the population estimated in the last table for that year based 
 upon the average increase. This fact, together with the conditions stated at the 
 beginning of this discussion which seem to mark a tendency to uniformity of growth 
 of population in the District of Columbia, make it seem advisable to use tor our esti- 
 mates of future population the figures given in the last column of the table above^ . 
 based upon the assumption of an approximately uniform numerical increase during 
 each decade equal to the average increase during the past four decades. It is believed 
 that these figures will represent the population of the District reasonably well for 
 the immediate future periods which we must consider. 
 
 THE FUTURE CONaUMPTION OP WATER IN THE DISTRICT OF COLUMBIA. 
 
 From the estimates of future population and of future per capita daily consumption 
 in the District of Columbia, the following table may be constructed showing the 
 probable quantities of water that will have to be provided for: 
 
 Year. 
 
 Popula- 
 tion. 
 
 Mean an- 
 nual con- 
 sumption 
 per day. 
 
 Maximum 
 
 capacity to 
 
 provide tor 
 
 per day. 
 
 Year. 
 
 Popula- 
 tion. 
 
 Mean an- 
 nual con- 
 sumption 
 per day. 
 
 Maxiniimi' 
 
 capacity t» 
 
 provide to* 
 
 per day. 
 
 1920 
 
 405,000 
 4fi0,000 
 615,000 
 
 Gallone. 
 55.000,000 
 62,000,000 
 70,000,000 
 
 Oallom. 
 80,000,000 
 90,000,000 
 101,000,000 
 
 1950 
 
 670,000 
 626,000 
 
 OaUorii. 
 77,000,000 
 84,000,000 
 
 Oallom. 
 112.000.000 
 
 1930 
 
 1980 
 
 122,000,000 
 
 1940 
 
 
 
 .]* . 
 
 The actual capacity to be provided for in any system is not the mean annual con- 
 sumption. It has been pointed out that a maximum draft of 45 per cent in exceeg 
 of the mean annual average has occurrSd and may be expected to occur again hi 
 this city. For the mean annual consumption given in the third column this would 
 indicate the need of the maximum capacities stated in the last column. The limiting 
 capacity of the old aqueduct system has been found to be practically 90,000,000 
 gallons per day. It must be borne in mind that the probability of having to provide 
 only the quantities stated in this table is contingent upon the reduction of waste, to 
 has already been emphasized. With this practically assured by the experienclfll 
 the last two years, however, the limit of the capacity of the old aqueduct sy^m 
 should not be reached until about 1930. ' 
 
 Assuming the approximate correctness of the prediction as to the future per capita 
 daily consumption in the District of Columbia, there still remains a factor of safety 
 in favor of the above statement. The means advocated for the highest de^eej'of 
 restriction of waste should have the effect of reducing the 45 per cent assumed in the 
 above table for the excess of the maximum rec(uired capacity over the mean annual 
 consumption. This would reduce the figures in the last column, or, in other wordSi 
 should still further postpone the date when the maximum demand will equal m 
 exceed the maximum capacity of the system. $ 
 
 Appendix B. 
 
 THE COST OP PUMPING WATER. 
 
 The scope of this investigation does not confine us to the consideration of gravity 
 supplies. With the Potomac flowing right past the city, the possibility of drawing 
 the supply from the river at some nearer point than Great Falls and lifting it by means 
 of pumps to the height required for distribution seemed a possibility not to be ignored. 
 The pumping of the large quantity of water, against the considerable heads necessary, 
 would constitute so large a portion of the annual charges against such a project thai 
 the cost of pumping water became a matter of importance. 
 
 The cost of pumping water, as herein considered, is made up of the following parts: 
 
 Cost of materials, including coal, oils, grease, waste, etc. 
 
 Cost of labor, including engineers, firemen, oilers, and other pumping station em- 
 ployees. 
 
WATER SUPPLY OF THE DISTBICT OF COLUMBIA. 
 
 65 
 
 Cost of ordinary maintenance repairs, etc. 
 
 It does not include interest on investment, depreciation, extraordinary repairs, 
 etc., at station, nor any portion of maintenance of distribution system nor general 
 administration. 
 
 Of the first part, the item of coal constitutes the larger part; in the general ease it 
 may be from 85 to 90 per cent of the total cost of materials. This total for materials 
 is commonly 40 to 50 per cent of the total pumping expenses. The item of coal alone 
 is therefore important. With the advances that have been made in recent years in 
 the design of pumping machinery and the greater economy resulting therefrom, the 
 item of coal has decreased. It is not possible, however, to predict, even from the 
 guaranteed duty of a pumping equipment, what its actual station duty in service 
 will be. As has been aptly stated in the pumping investigations for the commission 
 on additional water supply for New York City, "the discrepancy between the annual 
 duty in regular running and that shown by the official tests of different engines, in 
 most cases, seem to indicate that the designing engineers and builders of pumping 
 engines are far in advance of the pumping station managers." So much depends 
 upon the intelligence and watchfulness of the employees intrusted witlf the imme- 
 diate care of the station that the prediction of the amount of coal used per unit of 
 water pumped is not attended by any certainty. 
 
 Statistics of the cost of pumping water at stations in large cities afford a good means 
 of arriving at the probable cost in actual operation. A study has been made of the 
 cost records of a number of cities, supplemented by information from the engineers 
 in charge. The costs as given in reports are generally not comparable with the costs 
 in the District of Columbia for a number of reasons, among which are the following: 
 
 (1^ Different heads pumped against. 
 
 (2) Different types of machinery used. 
 
 (3) Different kinds of ccal used. 
 h) Different freight rates on coal. 
 
 The differences in (1) are equalized as far as possible by computiug the pumpage 
 of each station compared to an equivalent quantity pumped 1 foot high. The equal- 
 izing of the differences in (3) and (4) offeia the greatest difficulties. The cost of coal 
 used at any point depends upon its quality and upon the distance of the point of 
 consumption from the coal fields. The first step in this equalization process seemed 
 to be to eliminate the variable factor of freight rates. This could not be done -with 
 precision, except at an unwarranted expense of time, but an approximate correction, 
 which is believed to be generally accurate, was ajipjied after a study of the freight 
 tariffs in the files of the Interstate Commerce Commission. 
 
 Following. upon this, the assumption was made that in a general way the value* 
 of the different coals for pumping were proportional to these equalized coal prices. 
 The coal cost for each station was thus changed in the ratio of the cost of good steam- 
 ing coal in the District of Columbia to the cost at that station, and upon this and the 
 other proper pumping-station expenses the cost was computed of pumping l,000,000i 
 ^lons of water 1 foot high. Without going into any greater refinement, which, with 
 the meagre data would have been difficult and perhaps no more accurate, these reeult- 
 ing costs were assumed to be approximately what the costs would have been at the 
 various stations if they had been operated with coal of the quality and cost used hero 
 in the District. 
 
 Average costs of pumping water at various stations for entire periods considered (.5 to tO 
 years), after freight corrections, etc., per 1,000,000 gallons 1 foot high, coal at current 
 Washington prices (fS.50 per ton). 
 
 Philadelphia, Pa.: 
 
 Spring Garden $0,049 
 
 Belmont 047 
 
 Queens lane 033 
 
 Roxborough ..059 
 
 Frankford (new) 030 
 
 Boston, Mass.: 
 
 Chestnut Hill, low service ... .040 
 Chestnut Hill, high service. . . 036 
 Spotpond 044 
 
 Cleveland, Ohio: 
 
 Kirtland street 040 
 
 Fairmont street 094 
 
 Division street 093 
 
 Chicag;o, 111.: 
 
 Fourteenth street 035 
 
 H. Doc. 347, 61-2 5 
 
 Chicap, 111.— Continued. 
 
 Sixty-eighth street $0. 049 
 
 Twenty-second street 052 
 
 Chicago avenue 054 
 
 Springfield avenue 043 
 
 Central Park avenue 043 
 
 Harrison street 049 
 
 Lake View 054 
 
 Detroit, Mich 032 
 
 Milwaukee, Wis 043 
 
 Providence, R.I 039 
 
 Fall Riverj Mass 070 
 
 Minneapolis, Minn 030 
 
 Washington, D. C: 
 
 Trumbull street station 048 
 
 Filtration plant station 056 
 
S6 
 
 WATER SUPPLY OF THE DISTRICT OF COLUMBIA, 
 
 ll•f?J^^ V"'^''^*''®^ '^'^^^ ranging from $0.03 to nearly $0.10 per 1,000,000 gallons 
 ijitted 1 toot high under Washington conditions. The average cost is about $0,049 and 
 there are but few below $0,040. The cost of pumping at the Trumbull street station 
 Here in the District for the fiscal years 1907 and 1908 was $0,048, or practically the 
 same as the average given above. The cost at the pumping station at the filtration 
 .plant was slightly higher than the average, or $0,056 for the fiscal years 1907 and 1908. 
 1 he assumption of the cost, for purposes of estimating, as $0,040 per 1,000,000 gal- 
 lons 1 toot high, would seem to be as low as was advisable, and would put any stations 
 ■built on that assumption in the class with the best stations in the country. This 
 iigure therefore has been assumed. 
 
 Appendix 0. 
 
 separate supply op unfiltbeed water for public buildings, large consumers, 
 
 piee service. etc. 
 
 The qua,ntity of water used by the Government, for all of which the water depart- 
 ment receives nothing, is very great. At the navy -yard and in many of the public 
 .buildings enormous streams are flowing continuously, without restriction, for purposes 
 such as condensing water, etc., for which filtered water is unnecessary. It seemed 
 desirable, therefore, to inquire into the propriety of a separate supply of unfiltered 
 water for these and for certain large private consumers, and for fire service. 
 
 inquiry of persons in charge of the various public buildings, etc., and of private 
 "consumers, brought out the fact that in a general way one-half of the water supplied 
 to those consumers might be unfiltered water. There were general expressions of 
 protest against the idea, however, and especially against the expense it would involve 
 in every case of separating the filtered and unfiltered supplies within the building. 
 
 The government buildings and the large private consumers use a total of about 
 12,000,000 gallons per day. Of this, the indications are that about 6,000,000 gallons 
 .per day might be unfiltered water. The points at which this quantity would be used 
 are widely scattered and the plan would necessitate an extensive and entirely separate 
 ^set of mains. 
 
 This supply might be obtained bjr pumping from a small plant at the river, or by 
 •gravity from McMillan Park reservoir, wiuiout filtration. Of these the latter would 
 obviously be simpler and cheaper. 
 
 The saving that would be effected by this plan would be the reduction in total ex- 
 penses at the filtration plant due to not having to filter this quantity of water. The 
 cost of filtration is about $3.10 per 1,000,000 gallons, but the saving m total expenses 
 Would be much less than this for a reduction of 6,000,000 gallons in the quantity fil- 
 iered. This reduction would make no change in office and laboratory expense, which 
 Kionstitutes 22 per cent of the total. The change in pumping expense would be prac- 
 tically confined to the fuel item, which is about 20 per cent of the total. This would 
 be reduced in proportion to reduction in quantity pumped, or, say, six sixty-fifths of 
 20 per cent, or 2 per cent of totals. The change in filter operation expense may be 
 ^assumed to be reduced in proportion to the reduction in quantity pumped, or, say, 
 six sixty-fifths of 32 per cent, or 3 per cent of total. The reduction would therefore be 
 about 5 per cent of tne total expense. For the year 1908, 5 per cent of $82,400 would 
 ^e $4,120, which is equivalent to $1.88 per 1,000,000 gallons as the actual saving 
 effected by reducing the quantity filtered by 6,000,000 gallons. . , 
 
 The capitalized value of $4,120 at 3 per cent is $137,000. In order for the plan to 
 be financially attractive, therefore, $137,000 should cover all expenses incident to 
 laying the complete new set of niains for this separate supply, and all changes in the 
 'various public buildings and private properties necessitated by the separation of the 
 !two supplies. It is believed that this sum would be jnsufficieiit for the purpose. 
 
 This plan is not carried out in any more detail, because it is plain that it has no direct 
 ibearing upon the principal point of this investigation, which is for an additional sup- 
 ply for the entire city. The quantity of unfiltered water that might be used in the 
 «ity wiiJiout inconvenience is obviously too small to warrant the construction of a 
 separate system of mains and the maintenance of a special pumping station at the 
 Tiver, and the plan for a separate system from McMillan Park reservoir, as described 
 above, is, simply a part of the larger problem of procuring an additional supply for the 
 fentire city. 
 
 , ilre sermce- — The reports of the fire department of the District give the total quan- 
 (tity of water used for fire service in recent years as between 25,000,000 and 30,000,000 
 gallons per year, or about one-tenth of 1 per cent of the total consumption. As far as the 
 Actual value of the water is concerned, this quantity is relatively insignificant, and the 
 saving effected by use of unfiltered water may be ignored. 
 
WATEB SUPPLY OF THE DISTBICT OF COLUMBIA. ^7 
 
 The two things above all others that are essential for eflacient fire protection are am- 
 ple capacity and suflficient pressure. The problem resolves itself into one of obtaining 
 a large quantity of water under high pressure at any place and at any time requiredj 
 regirdless of the actual value of the water used. 
 
 In the District of Columbia this might be effected in either of two ways: 
 
 (1) High-service fire mains laid from Fort Reno reservoir (flow line elevation 423, 
 or practically 175 pounds pressure in business district). 
 
 Q) Fire-pumping station at river front, and flre-service mains to business district. _ 
 
 The first involves (a) the laying of mains of proper size from Fort Reno reservoir 
 to the city and to the District of Columbia pumping station; (b) pumping with present 
 equipment into Fort Reno reservoir, entirely as at present; (c) the only maintenance 
 charges, aside from the pumping above specified, would be those incident to the 
 inspection and repair of the fixed high-service system. 
 
 The second involves (o)_ the laying of mains of proper size from a fire pumping sta- 
 tion near river into the city; (6) construction of pumping station having capacity of 
 18,000,000 to 25,000,000 gallons of water per day against a pressure of 184 pounds 
 (equivalent to the elevation of the fiow fine, 'Fort Reno reservoir); (c) maintenance 
 and operation charges, including inspection and repair of high-service system; the 
 maintenance of complete pumping station force all the time, keeping boiler pressure 
 up at all times, ready for use. 
 
 Of these two ways, the former would probably be the simpler and more economical, 
 and has been studied and reported upon by the superintendent of the water depart- 
 ment. Like the last plan, as far as we are concerned in the present investigation, 
 this is therefore a part of the larger problem of procuring an additional supply tor the 
 entire city. A high pressure supply designed purely for fire service in the business 
 portions of the city can not be too emphatically recommended; but if a considera- 
 tion of the subject fails to show any relief in the general problem of increasing the 
 supply of water for the city, it is clearly not within the scope of this investigation, 
 but lies rather with the District water department who already have a project for 
 this purpose before Congress. 
 
 Appendix D. 
 capacitt op the existing aqueduct system. 
 
 With the mean daily quantity of water demanded by the city for consumption and 
 waste not far from the safe limit, and the maximum daily quantity so great as to cause 
 a most serious depletion of the reservoirs, it becomes important to know the limiting 
 capacity of the present aqueduct system. 
 
 Only one careful study of the flow in the aqueduct has previously been made. 
 This was in 1896-97, by Capt. (now Lieut. Col.) D. D. Gaillard, Corps of Engineers, 
 U. S. Army, and the details thereof are reported in full in the annual report of the 
 Chief of Engineers for 1897, pages 4004-4009. No significent changes have been made 
 since then in the structures of the system. The result qf his studies, stated briefly, 
 was that the ultimate capacity of the system with the Potomac at its lowest stage and 
 the water in the Georgetown reservoir at elevation 144, was about 76,500,000 gallons 
 per day. Georgetown reservoir was at that time the distributing point to the city 
 mains, and McMillan Park reservoir was not yet in service. 
 
 Those studies did not includq any consideration of the maximum quantity the 
 aqueduct would deliver if the water in Georgetown reservoir were drawn still lower. 
 Tne assumed elevation of 144 left more than 2 feet of water above the crown of the 
 conduit at that point, and obviously the discharge could be increased by drawing it 
 lower. The problem now consists in the determination of the ultimate capacity of 
 the system at the lowest attainable levels, from the intake at Great Falls to tne outlet 
 of McMillan Park reservoir, or, in other words, to the filtration plant pumping station. 
 
 Since the studies of 1896-7, estimates have been made of this limiting quantity. 
 Curves have also been constructed to show the discharge corresponding to a given loss 
 of head through a certain section of the conduit. These have all been based upon the 
 coefScients derived in the studies of twelve years ago. Until October, 1905, there had 
 never been a direct and continuous measurement of the discharge. Since that date 
 all the water flowing into the distribution system of the District has passed through 
 the Venturi meters at the filtration plant. The accuracy of these meters has been 
 tested over and over, and they are safely within a linaiif of error of 2 per cent. For 
 some time before October, 1905, the total recorded consumption was determined from 
 the loss of head through a certain section of the conduit, as referred to above. Shortly 
 after that date it was observed that this method of determining the discharge did not 
 check with the records obtained from the Venturi meters. This indicated a change 
 
68 
 
 WATEB SUPPLY OF THE DISTKICT OF COLUMBIA. 
 
 in the discharge tor a given loaa of head, resulting probably from a change in the 
 hydraulic quality of the interior surfaces of the conduit since the determinations of 
 1896-7. The recent study was taken up with a view to the determination of the 
 coefiScients under present conditions, and for a knowledge of the ultimate limiting 
 capacity of the entire system. 
 
 The earlier studies were carried out by current meter measurements, from which 
 the mean velocities, the values of the coefficients, etc. , were determined. The formula 
 used was J)=c\/r8. The continuous records of discharge through the Venturi meters 
 over a period of more than three years afford accurate data as to quantity for the 
 present study. 
 
 The data and procedure followed, together with a r6sum6 of the old experimental 
 data, are given below. The formula used, and to which the old figures have been 
 adapted to make them comparable with the present ones, is the exponential formula 
 devised by Messrs. Hazen and Williams, now widely used for this work: 
 
 i,=cr''««3 s""'" O.OOl-''*'" 
 
 Experiments in 1896-7. — From Colonel Gaillard's report (Report of Chief of Engi- 
 neers, U. S. Army, 1897, Part 6, pp. 4004^4009), we have— 
 
 Section A— Dalecarlia to Georgetown reservoir, 10,150 feet. 
 
 Section B— Gauging station No. 2 to Dalecarlia reservoir, 22,719 feet. 
 
 Section C— Bridge No. 2 to gauging station No. 2. 17,626 feet. 
 
 Section D— Manhole No. 1 to bridge No. 2, 7,201 feet. 
 
 Gauging station No. 1, probably at ventilators just south of south connection, Dalecarlia reservoir. 
 
 Gauging station No. 2, between tunnels just west of bridge No. 3. 
 
 Slopes and velocities determined by Colonel Gaillard August and September, 1896, and 
 
 June and July, 1897. 
 
 
 Length. 
 
 Mean 
 velocity. 
 
 Fall in 
 length 
 stated. 
 
 1,000. 
 
 c, 
 
 H.&W. 
 
 
 Feet. 
 10,150 
 
 22,719 
 
 17,526 
 7,201 
 
 1.169 
 1.267 
 1.279 
 1.334 
 
 1.055 
 1.063 
 1.379 
 1.382 
 1.534 
 1.545 
 1.765 
 1.800 
 1.826 
 1.898 
 
 . 1.765 
 1.800 
 1.826 
 1.898 
 
 1.765 
 1.800 
 1.826 
 1.898 
 
 0.569 
 .649 
 .639 
 .739 
 
 .717 
 .717 
 1.207 
 1.217 
 1.527 
 1.627 
 2.145 
 2.162 
 2.217 
 2.217 
 
 1.54 
 1.54 
 1.64 
 1.64 
 
 0.066 
 .064 
 .063 
 .073 
 
 .032 
 .032 
 .053 
 .054 
 .067 
 .067 
 .094 
 .095 
 .097 
 .098 
 
 .0884 
 .0884 
 .0935 
 .0935 
 
 .150 
 .150 
 .143 
 .143 
 
 105 
 
 
 106 
 107 
 103 
 
 
 105 
 127 
 
 
 128 
 327 
 128 
 124 
 125 
 119 
 121 
 121 
 125 
 
 
 124 
 
 
 123 
 
 
 126 
 123 
 128 
 
 
 125 
 
 
 93 
 
 
 
 95 
 
 
 
 99 
 
 
 
 103 
 
 
 
 
 
 97 
 
 SUMMARY. 
 
 Section A to C J"5 
 
 Section B toC }^ 
 
 Section C to C l« 
 
 SecUonDtoC "i 
 
WATEB SUPPLY OP THE DISTKICT OF COLUMBIA, 
 
 69 
 
 B and C may be taken together as 125. D is evidently a decidedly restricted section. 
 B, C, and D combined, give'C equal 119, which may be used as long as the aqueduct 
 
 ia flowing full 
 
 Flbw under present conditions through aqueduct. 
 [Section, Great Falls (M. H. No. 1) to DaleoarUa, 47,«6 feet.] 
 
 Period. 
 
 September 12 to October 15, 1906 . . . 
 November 11 to December 19, 1906. 
 
 Januarr 1 to March 3, 1907 
 
 May 29 to June 30, 1907 
 
 Mardi 16 to March 28, 1906 
 
 Meaavalueof C 
 
 Quantity 
 (mil. gals, 
 per day). 
 
 70.9 
 64.6 
 70.9 
 63.4 
 64.6 
 
 L. H. 
 
 6.35 
 4.79 
 5.19 
 4.91 
 4.84 
 
 Slope per 
 1,000. 
 
 0.134 
 .099 
 .109 
 .103 
 .102 
 
 97 
 103 
 108 
 
 99 
 102 
 
 102 
 
 The above periods were chosen for their uniformity of conditions. The quantities 
 were obtained from Venturi records, corrected for changes in level of reservoirs during 
 the periods. The losses of head were obtained from aqiKduct records. The values 
 of C were derived from the H. & W. formula. 
 
 The values of C are very low in comparison with Colonel Gaillard's value, about 119 
 for this section. 
 
 Possible errors in recent data: Leakage from reservoirs, conduits, or tunnels; evapo- 
 ration; water for sand washing. 
 
 Leakage from reservoirs, etc., if any, would be practica;lly constant and might ac- 
 count for an apparent reduction in capacity, as recent measurements of Q are ob- 
 tained after passing through entire reservoir system,_ while Colonel Gaillard's were 
 obtained by measuring, the capacity at a point in this section. This does not seem 
 likely, however, as we nave no evidence of any loss of water from this cause. 
 
 There might have been some evaporation during the first and the fourth periods; 
 enough possibly to have increased C for those periods by 1 point each. 
 
 A large quantity of water for sand washing was used during the first period; enough 
 to have increased C for that period by 2 points. The amounts during the other periods 
 were small. 
 
 Applying these corrections, the average value for C equals 103 with variations not 
 greater than 5 per cent each way. 
 
 Colonel Gaillard's observations were made shortly after a most thorough cleaning 
 of the conduit. Some 750 cubic yards per mile of deposits were removed, and "men 
 provided with hoes and stiff brooms scraped and swept the conduit and iDy-conduits 
 for their entire length." (Report for 1896, p. 3910.) 
 
 As far as the records show, this is the only comprehensive and thorough cleaning 
 the conduit has ever had. It has been cleaned at frequent intervals since then by 
 removing any large masses of solid matter, fallen rock, roots, logs, etc., and simply 
 stirring up the soft deposits and attempting to flush them out upon first turning in 
 the water. Some sweeping of the walls is done, but evidently, much less thoroughly 
 than Colonel Gaillard's cleaning, for which $14,000 was specially appropriated. The 
 insufficiency of flushing for removing deposits is set forth in Report oi the Chief of 
 Enrineers, U. S. Army, 1896, page 3910. 
 
 Tne deposits now are very small, and would not diminish the cross-section area 
 by any significant amount; but this, together with the deterioration of the condi- 
 tion of the wall surface which has taken place in the absence of scraping and sys- 
 tematic treatment, probably makes the present value of C lower than Colonel Gaillard's 
 figures would indicate. 
 
 Section D has about 25 per cent unlined tunnel. 
 Section C has about 7 per cent unlined tunnel. 
 Section B has about 6 per cent unlined tunnel. 
 
 The five periods above worked out seem to indicate a value C equal 103 for the 
 entire Great Falls-Dalecarlia section flowing full. The low coefficient in the D sec- 
 tion is due plainly to the 25 per cent of rough unlined tunnel. This value of C for this 
 will diminish but little, if at all, because it is already so rough. Assume that it remains 
 97 and compute the coefficient C for B and C sections. The coefficient for B and C 
 sections seems tp be at present from this computation about 105, which would be a 
 reduction of 13.5 per cent in the carrying capacity of the B and C sections since' 
 Colonel Gaillard's observations. 
 
70 
 
 WATER SUPPLY OP THE DISTEIOT OP COLUMBIA. 
 
 [Section, Dalecarlia to Georgetown, 10,150 Xeet.) 
 
 • 
 
 
 
 Period. 
 
 Quantity 
 (mil. gals, 
 per day). 
 
 L.H. 
 
 1.60 
 1.18 
 1.44 
 1.28 
 1.24 
 
 Slope per 
 
 C 
 
 September 12 to October 15, 1906 
 
 70.8 
 64.5 
 70.9 
 63.4 
 64.5 
 
 0.157 
 .116 
 .142 
 .126 
 .122 
 
 89 
 
 November 11 to December 19, 1906 
 
 96 
 
 January ItoMaroha, 1907 
 
 
 May 29 to June 30, 1907 
 
 89 
 
 March 16 to 28, 1906 : 
 
 
 
 
 
 Mean value of C 
 
 
 
 1 
 
 
 
 The data were obtained as heretofore explained.- A correction of +0.09 foot has 
 been applied to the recorded loss of head based upon level determinations made upon 
 the aqueduct system ,in the spring of 1908, in which the gauge at Dalecarlia outlet 
 was found to be 0.09 foot too high. This change accounts for about 3, per cent of the 
 total 13 per cent of apparent reduction in capacity since Colonel Gaillard's observa- 
 tions, which gave C equal lOS! The remaining 10 per cent can be accounted for only 
 by deterioration of the interior surfaces. 
 
 [Section, Ge&getowu to McMillan Park reservoir, 20,700 feet.] 
 
 Standard section: 
 
 Area =76.34 square feet. 
 Perim.=31.72 feet, 
 r =2.40. 
 
 There are three abrupt changes in direction, at the points where the east and west 
 shafts join the tunnel. For average conditions of flow, allow a total of 0.2 foot for 
 the loss of head at these points, etc. 
 
 A correction of +0.'57 foot has been applied to the recorded loss of head based upon 
 recent level determinations on the aqueduct system, in which the gauge at the east 
 shaft house was found to be 0.57 foot too low. 
 
 Period. 
 
 Quantity 
 (mil. gals, 
 per day). 
 
 Velocity. 
 
 L. H. cor- 
 rected. 
 
 Slope 
 per 1,000. 
 
 September 12 to October 15, 1906. . . 
 November 11 to December 19, 1906. 
 
 January 1 to March 3, 1907 
 
 May 29 to June 30, 1907 
 
 March 16 to 28, 1906 , 
 
 June 4 to 7, 1908 
 
 June'21to24, 1908 , 
 
 June 30 to July 4, 1908 
 
 July 14 to 18, 1908 
 
 August 1 to 6, 1908 
 
 September 9 to 13, 1908 
 
 October 10 to 16, 1908 
 
 70.8 
 64.3 
 70.9 
 63.4 
 64.1 
 67.9 
 68.0 
 72.2 
 74.0 
 71.4 
 64.7 
 65.1 
 
 1.44 
 1.31 
 1.44 
 1.28 
 1.31 
 1.38 
 1.38 
 1.47 
 1.50 
 1.45 
 1.31 
 1.32 
 
 1.63 
 1.51 
 1.81 
 1.56 
 1.64 
 1.37 
 1.37 
 1.91 
 1.74 
 1.67 
 1.76 
 1.18 
 
 0.079 
 .073 
 .089 
 .075 
 .079 
 .066 
 .066 
 .092 
 .084 
 .081 
 .086 
 .057 
 
 100 
 95 
 9S 
 92 
 91 
 106 
 106 
 94 
 108 
 100 
 
 Mean value of C. 
 
 These observations indicate a value for equal to about 98 for the Lydecker tunnel; 
 and to the loss of head resulting therefrom must be added 0.2 foot for loss of head at the 
 shafts. It is believed that the wide variations from the mean value for are due to 
 the fluctuations in the level of water in McMillan Park reservoir. The draft at the 
 pumps is so irregular that conditions of uniform flow through the tunnel for several 
 consecutive days practically never occur. 
 
 Actual test of hmiting capacity of the Washington Aqueduct. — During the week or 
 November 23 to 28, 1908, the following procedure was carried out to determine by 
 actual trial the limiting capacity of the system. Computations from the coefiicienta 
 derived above indicated that this would be about 90,000,000 gallons per day. The 
 requirements for the test then were — 
 
 (1) Adjustment of the reservoir levels to those indicated by the preliminary compu- 
 tations for a discharge of 90,000,000 gallons per day. 
 
 (2) Draft of 90,000,000 gallons per day at filtration plant pumping station, where 
 the quantity was recorded by the Venturi meters. 
 
 (3) Observations of levels and other conditions throughout aqueduct system after 
 steady flow at this computed limiting rate had been established. 
 
WATEE SUPPLY OF THE DISTBICT OF COLUMBIA. 
 
 71 
 
 The reservoir levels were adjusted by closing the gates at Great Falls at 6 p. Ba.' 
 November 23, which caused the levels in all toe reservoirs to drop. On Tuesday^ 
 November 24, an inspection trip was made through the conduit from Great Falls to-. 
 Dalecarlia. When Dalecarlia reservoir had reached elevation approximately 142.6" 
 the stop planks were put in at both ends. When Georgetown reservoir had reached, 
 elevation 140.25 it was shut oft to prevent further drop. Meanwhile, after the con- 
 duit had been inspected and repaired, it was slowly filled with water. When McMik 
 Ian Park reservoir had nearly reached elevation 137.5 the three pumps were started' 
 at approximately a combined rate of 90,000,000 gallons per day, the gates at Great' 
 Falls were opened wide, and Dalecarlia and Georgetown reservoirs put in service' 
 again. 
 
 Unforeseen condition in Georgetown reservoir. — ^When Georgetown reservoir was 
 drawn down to this level a condition was noted which had not previously been takeii' 
 into account. In the middle of the division embankment in the reservoir there is, 
 an opening 15 feet wide, the sill of which is approximately at elevation 140.1. For! 
 the level desired in this reservoir this opening would Carry but a small quantity. . Th^'. 
 by-conduit was therefore left open, and because of this restriction the greater por- 
 tion of the 90,000,000 gallons (roughly estimated at 75,000,000 ga.llons) passed throiiglj, 
 the by-conduit and the remainder through the reservoir. This accounts for the largei 
 difference in elevations at the two ends of Georgetown reservoir and for the elevatiph'. 
 tieing several tenths lower than we expected, at McMillan Park reservoir. ' ! 
 
 Procedure during test. — Pumping at the high rate specified was kept up for a period! 
 of more than forty-eight hours, during which time observations were taken at varjougl 
 points at intervals of two hours. Between the hours of 10 a. m., Noveiiiber 27, aiid 
 10 a. m., November 28, there was practically no change in the established conditions 
 throughout the system. The elevations, meter re^ings, etc., at these times are 
 given below: 
 
 ELEVATIONS. 
 
 Great FaUs, M. H. No. 1 
 
 Dalecarlia reservoir 
 
 Oeorgetown: 
 
 Inlet 
 
 Outlet 
 
 MoMillaa Park: 
 
 Inlet 
 
 Outlet 
 
 10 a. 
 
 m. — 
 
 
 
 
 
 
 Raised. 
 
 NOV..27. 
 
 Nov. 28. . 
 
 
 150.5 
 142.6 
 
 150.5 
 142:7 
 
 if 
 
 140. S 
 
 140.9 
 140.2 
 
 
 140.1 
 
 G.l 
 
 137.4 
 137.0 
 
 137.4 
 137.0 
 
 
 
 
 METER READINGS. 
 [In millions of gallons.] 
 
 
 ■ 10 a. 
 
 m.— 
 
 Difler- 
 
 
 Nov. 27. 
 
 Nov. 28. 
 
 ence. 
 
 72-inch meter. 
 
 52,073.9 
 6,041.5 
 
 52,132.7 
 6,070.34 
 
 58:80 
 
 
 28.84- 
 
 
 
 
 
 
 87 64; 
 
 
 
 
 
 The differences in elevation at Georgetown inlet and outlet were explained above.. 
 Those at McMillan Park inlet and outlet were due to loss of head through the circu- 
 lating conduit and through the pumping station intake. 
 
 Capacity of aqueduct for established conditions. — For the levels established the aque- 
 duct system carried about 87,500,000 gallons per day. There was a tendency for 
 the upper sections to carry slightly more than this, as is indicated by the rise of O.L 
 foot in Dalecarlia and Georgetown reservoirs. This depth corresponds to between 
 2,000,000 and 3,000,000 gallons and it is probably reasonable to assume that the entire 
 system would have gradually adjusted itself to a rate of about 90,000,000 gallons. 
 
 Remits of increasing quantity pumped. — At the end of the period described above, 
 the pumps were speeded up so that the quantity discharged between 10 a. m. and 
 
72 
 
 WATEB SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 8 p. m., Maxch 28, was 40,580,000 gallons, or at the rate of 97,500,000 gallons per day. 
 Tne effect of this increase in rate -was to draw down McMillan Park reservoir to ele- 
 vation 136.6, while the elevation at Georgetown outlet remained unchanged at eleva- 
 tion 140.2, and of course Dalecarlia did not change. 
 
 Apparently this increase over a rate of 90,000,000 gallons per day was drawn 
 entirely from the storage in McMillan Park reservoir, and did not tend to steepen the 
 8lop"e of the hydraulic gradient above the outlet of Georgetown reservoir. Condi- 
 tions above that point remained just as they had been during the preceding period 
 of twenty-four hours during which the measured flow had been 88,000,000 to 90,000,000 
 gallons. 
 
 Conclusion. — It seems entirely reasonable, therefore, to conclude that the aqueduct 
 system as it now stands will discharge approximately 90,000,000 gallons per day, 
 but can not discharge much, if any, more than that. 
 
 Additional loss of head data. — ^As a result of the test just described, the following 
 revisions apply to the figures given above. The data from which were derived the 
 values for the coefficient C for the Dalecarlia-Georgetown section, include an error 
 due to the restriction in the middle of Georgetown reservoir. The elevations at 
 Georgetown used for determining the loss of head through that section were the ele- 
 vations at the outlet, as no regular record is kept of the elevation at Georgetown inlet. 
 In the absence of data it was assumed that the elevations at inlet and outlet would 
 not differ materially. They do differ, however, on account of the restricted area at 
 the opening in the division embankment, which is 15 feet wide and the sill at eleva- 
 tion 140.1 approximately. From computation, the loss of head through this opening 
 would seem to be as follows: 
 
 Elevation. 
 
 48 million 
 gallons. 
 
 60 million 
 gallons. 
 
 75 million 
 gallons. 
 
 142.1 
 143.1 
 144.1 
 
 0.52 
 .20 
 .11 
 
 0.96 
 .32 
 .18 
 
 
 0.54 
 .28 
 
 Correcting for this loss of head will increase the value of C in the Dalecarlia- 
 Georgetown section. This corrected value of is given in the table below, and 
 Seems more consistent than the lower value, c=92, previously deduced. 
 
 [Section, Dalecarlia to Georgetown, 10,150 feet.] 
 
 Period. 
 
 Quantity. 
 
 Observed 
 L. H. 
 
 Average 
 elevation 
 George- 
 town 
 outlet. 
 
 Corrected 
 L. H. 
 
 Slope per 
 1,000. 
 
 C 
 
 September 12 to October 15, 1906 
 
 November 11 to December 19, 1906 
 
 January 1 to March 3, 1907 
 
 70.8 
 64.5 
 70.9 
 63.4 
 64.5 
 
 1.60 
 1.18 
 1.44 
 1.28 
 1.24 
 
 143.2 
 145.4 
 145.0 
 145.4 
 145.6 
 
 1.14 
 1.08 
 1.30 
 
 0.112 
 .106 
 
 -12J! 
 
 106 
 100 
 99 
 
 May 29 to June 30, 1907 
 
 1.18 .116 
 1.14 .112 
 
 93 
 
 March 16 to 28, 1906 
 
 97 
 
 
 
 Mean value of C . . . 
 
 99 
 
 
 
 
 
 
 
 This value of C indicates a reduction of capacity of about 6 per cent, 3 per cent of 
 •which is an apparent reduction only, and due to the correction in the gauges. 
 
 The loss of head through the Georgetown reservoir by-conduit and through the 
 ■circulating conduit in McMillan Park reservoir are as follows: 
 
 Million gal- 
 lons per 
 day. 
 
 George- 
 town by- 
 conduit. 
 
 Circulating 
 conduit. 
 
 60 
 65 
 70 
 75 
 80 
 85 
 90 
 
 0.32 
 .38 
 .44 
 .50 
 .57 
 .64 
 .72 
 
 0.13 
 .15 
 .18 
 .20 
 .22 
 .25 
 .28 
 
WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 73 
 
 Summary of loss of head figures. — ^From the foregoing data the following summary 
 ia made up, giving the losses of head through the several principal sections of the 
 aqueduct ayBtem for the conditions stated: 
 
 
 
 Quantities (million gallons per day). 
 
 
 
 60. 
 
 65. 
 
 70. 
 
 75. 
 
 80. 
 
 85. 
 
 90. 
 
 95. 
 
 100. 
 
 Bentlon D, 7,201 leet, C=97: 
 
 Full, s<.isa 
 
 0.7 
 
 0.8 
 
 0.9 
 
 1.07 
 
 
 
 
 
 
 Mfudmum section, L. H.=il.OS for d<>82, 
 s>.i50 ... 
 
 
 
 1.33 
 
 1.47 
 
 1.63 
 
 Sections B and C, 40,245 feet, C=105: 
 Full, s<.150 
 
 3.4 
 
 3.9 
 
 4.5 
 
 5.1 
 
 5.8 
 
 
 
 Maxinium section, L. H.=6.05 for Q=88, 
 s> 150 
 
 
 6.55 
 
 
 
 Sections B, C and D combined, 47,446 feet, 
 C=103: 
 Full, s<.150 
 
 4.1 
 
 4.8 
 
 5.5 
 
 6.3 
 
 7.1 
 
 
 
 
 Uaxuaum section, L. H.=7.12 for Q=86, 
 s>.150 
 
 
 7.9 
 
 
 
 Secaon A, 10,150 feet, C=99: 
 
 Full,s<.160 ,.... 
 
 0.9 
 
 1.1 
 
 1.3 
 
 1.4 
 
 1.6 
 
 1.8 
 
 
 
 Uaxunum section, 1.52 for Q— 87.5, s>.i50.. 
 
 1.95 
 
 2.6 
 2.8 
 
 
 
 Lydeoker tunnel, 20,700 feet, C=98: '■ 
 Full 
 
 1.2 
 ' 1.4 
 
 1.4 
 1.6 
 
 1.7 
 1.9 
 
 1.9 
 2.1 
 
 2.1 
 2.3 
 
 2.4 
 2.6 
 
 2.9 
 3.1 
 
 3.2 
 
 Add 0.2 foot 
 
 3.4 
 
 
 
 Discharge diagram. — From the loss of head figures so far given, a discharge di^am 
 has been constructed. In plotting a discharge diagram, the water level at Great 
 Falls (M. H. No. 1) may be taken for average conditions as elevation 151.0. This 
 elevation will not be proper to use in computing and jjlotting the limiting flow, how- 
 ever, for there are numerous periods of many consecutive days when the elevation of 
 water at M. H. No. 1 does not rise above 150.5. Some of these periods are more than 
 thirty days long. The elevation at this point never falls much below 150.5. More- 
 over, the same conditions that tend to produce the limiting flow in the aqueduct, that 
 is, long continued cold weather, tend also to keep the river at a low stage. In plotting 
 the diagram, therefore, elevation 151.0 has been assumed for average conditions and 
 150.5 for the limiting conditions of flow. 
 
 The data from which are plotted the normal discharge curves are given above. The 
 curves for limiting conditions of flow were determined by assuming the conduit to be 
 divided up into short sections and computing the slopes required in the successive 
 sections to discharge the stated quantity. 
 
 Effect of lowering the sill in the opening in the embankment in Georgetown reservoir. — 
 The removal of this sill would enable us to maintain a draft of 90,000,000 gallons per 
 day at McMillan Park reservoir without a loss of head of 0.7 foot at Georgetown reser- 
 voir; that is, with additional available storage equal to 0.7 foot in McMillan Park 
 reservoir and the east half of Georgetown reservoir, and with 0.7 foot less lift at the 
 pumping station. 
 
 The removal of this sill would not give any material increase in the ultimate capacity 
 of tihe aqueduct. The lowering of the hydraulic gradient 0.7 foot at this point it is 
 est mated would increase the discharge about 1 per cent. 
 
 Summary of study on aqueduct discharge. — ^The values of the coefficient C in the 
 Hazen-Wiiliams formula, as derived from the old and new series of observations are 
 as follows: 
 
 Section. 
 
 1896-7. 
 
 1908. 
 
 Great Falls to Dalecarlla, entire 
 
 119 
 105 
 
 103 
 
 Salecarlia to Georgetown '. 
 
 99 
 
 Georgetown to McMillan Park 
 
 98 
 
 
 
 
 The report setting forth the results of the earlier observations shows them to have 
 been conducted with great care. Admitting the approximate accuracy of both of 
 these studies there is indicated a reduction in capacity in the Great Falls-Dalecarlia 
 section since 1896-7 of about 13.5 per cent and in the Dalecarlia-Georgetown section 
 of about 6 per cent. After an inspection of the structures themselves, and examina- 
 tion into all the factors involved, this reduction seems explainable only by the de- 
 terioration of the wetted surfaces. As already stated, the earlier measurements were 
 
74 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 made designedly just after the only thorough cleaning the conduit has ever had. 
 Large areas of the interior surfaces especially on the bottom and sides, are now cov- 
 ered with a soft but rough layer of clay averaging perhaps one-fourth inch in thickneefe 
 The presence of this rough deposit seems to be the explanation for the reduction in 
 capacity in the last twelve years. Beductions in the capacities of other aqueducts 
 are on record. The capacity of a portion of the Croton Aqueduct is reported to have 
 diminished about 4 per cent; of other portions, larger but unstated percentages; of the 
 Wachusett; Aqueduct, Boston, 10 or 11 per cent; of the Cochituate and Sudbury 
 aqueducts, Boston, respectively, 11.75 per cent and 13.5 per cent. 
 
 The coefficients determined as a result of the recent studies are low in comparison: 
 witii those of the aqtueducts just referred to. They correspond more nearly to the 
 coefficients for good brick sewer work than for first-class' aqueduct construction. 
 
 The analysis of the aqueduct records of discharge, loss of head, etc., covering a: 
 threeryear period indicates that the maximum quantity of water that can be drawn 
 through the system at the present time, from Great Falls to McMillan Park reservoir.j 
 is 90,000,000 gallons per day. This conclusion from the computations was confirmeai| 
 by the actual test in November, 1908. t 
 
 Appendix E. 
 
 itemized estimates of cost. 
 
 Itemized estimate of cost of Great Falls gravity project, parallel location. 
 
 Land, 24.2 acres, at $150 $3, 630. Ofr 
 
 Land, 6.8 acres, at $17,500 , 119, 000. 00 
 
 Clearing and grubbing, 15 acres, at $160 - - r - - -.- - 2, 400. 00, 
 
 Great Falls gatehouse: 
 
 Excavation, earth, 1,003 cubic yards, at 60 cents 602, 00 
 
 Excavation, rock, 2,840 cubic yards, at $2.50 7, 100. 00, 
 
 Concrete, 2,047 cubic yards, at $8 16, 376. 00 
 
 Riprap, 140 cubic yards, at $2.25 ^... 315.00. 
 
 House, 996 square feet, at $4 3, 984. 00 
 
 Structural steel, 5,520 pounds, at 8 cents 442. OOi 
 
 Iron, guides, ladders, 6,950 pounds, at 10 cents 695. 00 
 
 Gates,. 3i by 9 feet, 2, at $900 1, 800. 00 
 
 Gate, 4-foot, circular 550. 00 
 
 Cast-iron pipe, 48-inch, 675 feet, at $14 9, 450.^00 
 
 Stop planks, 2,000 feet b. m., at $60 120.00 
 
 Reinforcing steel, 2,220 pounds, at 4 cents 89..iD0 
 
 Present aqueduct crossing 1, 000. 00 
 
 Railing, 24 linear feet, at $1.50 ---':,.- 36.|)0 
 
 Add for contingencies. , 6, 000.;to 
 
 , : 48,559.|)0, 
 
 i J 
 
 Tunnel, station to station 92 (10 per cent timbered): 
 
 9,200 feet, at $45 per foot 414, 000. 00 
 
 Crossidg near Anglers' Club (estimate covers 150 feet): 
 
 Excavation, earth, 4,160 cubic yards, at 50 cents 2, 080. 00 
 
 Embankment, ordioary, 3,918 cubic yards, at 40 cents 1, 567. 00 
 
 Top soil, 80 cubic vards, at $1.30 104. 00 
 
 Concrete, 390 cubic yards, at $8 3, 120. 00 
 
 Steel reenforcing, 6,820 pounds, at 4 cents ; 273. OOl 
 
 Cast-iron pipe, 12-inch, 60 feet, at f 2. , . ; , 120. 00' 
 
 Gate valve, 12-inch : ,. 75.00 
 
 Manhole top ■ , . 15. 00 
 
 Steel, structural, 17,280 pounds, at 8 cents 1, 382. 00 
 
 Hardware, 1,360 pounds, at 5 cents 68. 00. 
 
 Lumber, 30,000 feet b. m., at$60 ,..,.. 1,800.00'. 
 
 Add for contingencies 2,500.00.' 
 
 Total cost of crossing '....,........., 13il04.0rf 
 
 Cost per foot. ....'....'.!.:.... 87. 40' 
 
WATEB SUPPLY OP THE DISTRICT OF COLUMBIA. 
 
 75 
 
 Masonry aqueduct, station 92 to station 235 (excluding cross-over) 
 (full masonry sectioa): 
 
 ' ■ Excavation, earth, 85,480 cubic yards, at 50 cents $42, 740. 00 
 
 Embankment) ordinary, 95,276 cubic yards, at 40 cents 38, 110. 00 
 
 Embankment, rolled,- 2,054 cubic yards, at 55 cents 1, 130. 00 
 
 Top soil, 22,093 cubic yards, at $1.30 28, 721. 00 
 
 Concrete, 25,797 cubic yards, at $8 , 206,376.00 
 
 Length, 14,150 feet 317,077.00 
 
 CoBtperfoot 22.40 
 
 Blow-oft, station 172: 
 
 ' ■ Excavation, earth, 134 cubic yards, at 50 cents 67. 00 
 
 Embankment, ordinary, 106 cubic yards, at 40 cents 43. 00 
 
 Concrete, 17 cubic yards, at $8. 136. 00 
 
 Steel reenforcing, 140 pounds, at 4 cents 6. 00 
 
 ' Sluice gate, 48-inch.. 550. 00 
 
 Cast-iron pipe, 48-ineh, 60 feet, at $14 840. 00 
 
 1, 642. 00 
 
 Masonry aqueduct, station 235_to station 267, except 200 feet at bridge. 
 No. 3 siphon (reduced masonry section): 
 
 Excavation, earth, 10,135 cubic yards, at 50 cents 5,068.00 
 
 Excavation, rock, 8,901 cubic yards, at $2.50 22, 253. 00 
 
 Embankment, ordinary, 15,97* cubic yards, at 40 cents 6, 391. 00 
 
 Embankment, rolled, 1,655 cubic yards, at 55 cents 911. 00 
 
 Top soil, 3,928 cubic yards, at $1.30 5, 107. Otf 
 
 Concrete, 4,620 cubic yards, at $8 36, 960. 00 
 
 Length, 3,000 feet 76,690.00 
 
 Cost per foot. 25. 56 
 
 Siphon and bridge, station 257 to station 259: 
 
 Excavation, earth, 2,698 cubic yards, at 60 cents. , 1, 619. 00 
 
 Embankment, ordina,ry , 854 cubic yards, at 40 cents 338. 00 
 
 Top soil, 330 cubic yards, at $1.30 429. 00 
 
 Concrete, plain, 686 cubic yards, at $8 5, 488. 00 
 
 Steel reenforcing, 624 pounds, at 4 cents 25. 00 
 
 Cast-iron pipe, 12-inch, 70 feet, at $2 140. 00 
 
 Gate valves, 12-iBch, 2, at $40 80. 00 
 
 Iron doors, 2, at $50 100. 00 
 
 Steel pipe, 7 feet diameter, 372 feet, at $18.25 6, 790. 00 
 
 Stop plank, 4,320 feet b. m., at $60 260. 00 
 
 Stop plank grooves, 3,152 pounds, at 10 cents.. 315.00 
 
 Checker plate, steel, 140 square feet, at 75 cents 105. 00 
 
 Length, 200 feet 15, 689. 00 
 
 Cost per foot 78. 45 
 
 Masonry aqueduct, station 267 to station 3214-42 (full masonry section) : 
 
 Excavation, earth, 43,874 cubic yards, at 50 cents 21, 937. 00 
 
 Embankment, ordinary, 37,299 cubic yards, at 40 cents ' 14,920.00 
 
 Embankment, rolled, 26 cubic yards, at 55 cents 15. 00 
 
 Top soil, 8,638 cubic yards, at $1.30 11,230.00' 
 
 Concrete, 9,600 cubic yards, at $8 76, 800. 00 
 
 Length, 5,442 feet >. 124,902.00 
 
 Costperfoot ....:,....- 22.93' 
 
76 WATEB SUPPLY OP THE DISTKICX OP COLUMBIA. 
 
 Siphon and bridge, Cabin John Creek, station 321+42 to station 327: 
 
 Bridge — 
 
 Excavation, earth, 1,230 cubic yards, at 60 cents $738. 00 
 
 Concrete, plain, 1,710 cubic yards, at $8 13, 680. 00 
 
 Concrete, face stones, 290 cubic yards, at $10 2, 900. 00 
 
 Blow-off— 
 
 Caat-iron pipe, 24-inch, 50 feet, at $6 300.00 
 
 Gate valves, 24-inch, 2, at $175 350.00 
 
 Bridge and blow-off 17, 968. 00 
 
 Siphon — 
 
 Excavation, earth, 6,766 cubic yards, at 50 cents 3, 383. 00 
 
 Embankment, ordinary, 4,300 cubic yards, at 40 cents 1, 720. 00 
 
 Top soil, 150 cubic yards, at $1.30 195.00 
 
 Concrete, 1,096 cubic yards, at $8 8,768.00 
 
 Steel pipe, l,180feet, at $18.25 21,525.00 
 
 Steel reenf orcing, 624 pounds, at 4 cents -. 25. 00 
 
 Stop plank, 4,320 feet b. m., at $60 260.00 
 
 Checker plate, steel, 140 square feet, at 75 cents 105. 00 
 
 Stop plank grooves, 3, 152 pounds, at 10 cents 315. 00 
 
 Cost of siphon, length 558 feet 36, 296. 00 
 
 Cost per foot, siphon alone 65. 00 
 
 Total cost with bridge 54,264.00 
 
 Cost per foot, including bridge. . . , 97. 25 
 
 Masonry aqueduct, station 327 to station 416 (full masonry section) : 
 
 Excavation, earth, 71,670 cubic yards, at 50 cents 35, 835. 00 
 
 Embankment, ordinary, 65,369 cubic yards, at 40 cents 26, 148. 00 
 
 Embankment, rolled, 2,902 cubic yards, at 55 cents 1, 596. 00 
 
 Top soil, 13,700 cubic yardSj at $1.30 17, 810.00 
 
 Concrete, 16,266 cubic yards, at $8 130,128.00 
 
 Length, 8,900 feet ...; 211,517.00 
 
 Cost per foot 23. 76 
 
 Masonry aqueduct, station 416 to station 451 (reduced masonry sec- 
 tion) : 
 
 Excavation, earth, 10,367 cubic yards, at -50 cents 5, 184. 00 
 
 Excavation, rock, 4,388 cubic yards, at $2.50 10, 970. 00 
 
 Embankment, ordinary, 23,926 cubic yards, at 40 cents 9, 571. 00 
 
 Embankment, rolled, 1,239 cubic yards, at 55 cents 682. 00 
 
 Top soil, 5,898 cubic yards, at $1 .30 7, 668. 00 
 
 Concrete, 5,439 cubic yards, at $8 ; 43,512.00 
 
 Length, 3,500 feet '77,587.00 
 
 Cost per foot 22.17 
 
 Masonry aqueduct, station 451 to station 465 (full masonry section) : 
 
 Excavation, earth, 15,386 cubic yards, at 50 cenla 7, 693. 00 
 
 Embankment, ordinary, 9,332 cubic yards, at 40 cents 3, 733. 00 
 
 Top soil, 1,851 cubic yards, at $1.30 2,407.00 
 
 Concrete, 2,408 cubic yards, at $8 19, 264.00 
 
 Length, 1,400 feet 33,097.00 
 
 Cost per foot 23. 64 
 
 Tunnel, station 465 to station 473-1-70, 870 feet (25 per cent timbered): 
 
 870 feet, at $45 39, 150. 00 
 
WATEB SUPPLY OP THE DISTEICT OF COIiTJMBIA. 
 
 77 
 
 Intake gatehouse, Dalecarlia reservoir: 
 
 Excavation, earth, 3,450 cubic yards, at 60 cents 
 
 Concrete, 2,173 cubic yards, at $8 
 
 House, 3,550 square feet, at $4 
 
 Iron, guides, ladders, 19, 100 pounds, at 10 cents 
 
 Cast-iron pipe, 36-inch, 300 feet, at $9 
 
 Gates, 3i by 9-foot sluice, 12 at $900 
 
 Gate, 3 by 3-foot sluice 
 
 Stop planks, 6,000 feet b. m., at $60 
 
 Reemorcing steel, 14,230. pounds, at 4 cents 
 
 Railing, 160 linear feet, at $1.50 
 
 Checker plate, steel, 150 square feet, at 75 cents 
 
 Same gatehouse, redesigned for requirements of coagulation at 
 Stubblefield site 
 
 Tunnel, station 508-f 50 to station 642-1-50, 13,400 feet (75 per cent 
 timbered): 
 13,400 feet, at $45 
 
 $2, 070. 00 
 
 17, 384. 00 
 
 14, 200. 00 
 
 1, 910. 00 
 
 2, 700. 00 
 
 10, 800. 00 
 
 350. 00 
 
 360. 00 
 
 570. 00 
 
 240. 00 
 
 113. 00 
 
 50, 697. 00 
 72, 000. 00 
 
 Masonry aqueduct, station 474-|-20 to station 508-|-50, except 1,450 feet 
 in tunnel, under-water section, and gatehouse (reduced masonry 
 section) : 
 
 Excavation, earth, 9,439 cubic yards, at 50 cents 4, 720. 00 
 
 Excavation, rock, 18,458 cubic yards, at $2.50 46, 145. 00 
 
 Embankment, ordinary, 21,949 cubic yards, at 40 cents 8, 780. 00 
 
 Top soil, 2,370 cubic yards, at $1.30 3, 080. 00 
 
 Concrete, 2,333 cubic yards, at $8 18,664.00 
 
 Steel reenforcing, 70,775 pounds, at 4 cents 2, 832. 00 
 
 Length, 1,980 feet net 84, 221. 00 
 
 Cost per foot 42. 50 
 
 ^Tunnel, station 481 to station 490, 900 feet (50 per cent timbered): 
 
 900 feet, at $45 40, 500. 00 
 
 By-conduit, Dalecarlia reservoir, under-water section:' 
 
 Excavation, earth, 2,610 cubic yards, at 60 cents 1, 566. 00 
 
 Masonry, reenforced concrete, 916 cubic yards, at $12 10, 992. 00 
 
 Reenforcing steel, 39,390 pounds, at 4 cents 1, 576. 00 
 
 Add for contingencies 5, 000. 00 
 
 Length, 460 feet 19,134.00 
 
 Cost per foot 41. 60 
 
 Gatehouse, south connection, Dalecarlia reservoir: 
 
 Excavation, earth, 748 cubic yards, at 60 cents 449.00 
 
 Excavation, rock, 900 cubic yards, at $2.50 2,250.00 
 
 Masonry, concrete, 1,107 cubic yards, at $8 8,856.00 
 
 House, 1,540 square feet, at $4 6, 160. 00 
 
 Iron, guides, ladders, 9,650 pounds, at 10 cents 965.00 
 
 Gates, 3i by 9 feet, 6 at $900 5,400.00 
 
 Stop planks, 3,000 feet b. m., at $60 .• 180.00 
 
 Reenforcing steel, 4,590 pounds, at 4 cents 184.00 
 
 Railing, 84 linear feet, at $1.50 126.00 
 
 Checker plate, steel, 100 square feet, at 75 cents 75. 00 
 
 24, 645. 00 
 
 602, 800. 00 
 
78 
 
 WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 Maeoniy aqueduct, station 642+50 to station 723+50, except 1,850 
 feet in siphon at Eock Creek and tunnel at Sixteenth street (reduced 
 mafionry section): 
 
 Excavation, earth, 21,665 cubic yards, at 50 cents 
 
 Excavation, rock, 23,743 cubic yards, at $2.50 »►. . - 
 
 Embankment, ordinary, 35,550 cubic yards, at 40 cents 
 
 Embankment, rolled, 4,709 cubic yards, at 55 cents 
 
 Top soil, 9,155 cubic yards, at $1.30 , 
 
 Concrete, 9,821 cubic yards, at $8 
 
 Length, 6,250 feet. 
 Cost per foot 
 
 Rock Creek siphon, including bridge, station 670+50 to station 686: 
 Rock Creek siphon bridge — 
 
 Excavation, earth, 5,100 cubic yards, at 60 centa 
 
 Excavation, rock, 3,890 cubic yards, at $2.50 
 
 Masonry, 10,620 cubic yards, at $8 
 
 Reenforcing steel, 45,000 pounds, at 4 centa 
 
 Blow-oft house — 
 
 Excavation, 408 cubic yards, at 50 cents 
 
 Concrete, 90 cubic yards, at $8 
 
 Reenforcing steel, 1,160 pounds, at 4 cents 
 
 Steel ladder, 200 pounds, at 10 centa 
 
 Manhole 
 
 Cast-iron pipe, 24-inch, 144 feet, at $6 
 
 Cast-iron pipe, 12-inch, 164 feet, at $2 
 
 Gate valve, 24-inch 
 
 Gate valve, 12-inch 
 
 Bridge and blow-off (not including.roadway) 
 
 Rock Creek siphon; total length, 1,550 feet — 
 
 Excavation, earth, 5,616 cubic yards, at 60 centa 
 
 Excavation, rock, 8,750 cubic yards, at $2.50 
 
 Embankment, ordinary, 12,105 cubic yards, at 40 centa. 
 
 Top soil^ 1,400 cubic yards, at $1.30 
 
 Concrete, 2,163 cubic yards, at $8 
 
 Steel pipe,. 7 fee.t. diameter, 3,108 feet, at $18,25. 
 
 Steel reenforcing, 624 pounds, at 4 cents 
 
 Stop plank, 4,.320 feet b.m.., at $60 per M , 
 
 Stop-pJank grooves, 3,152 pounds, at lO.cents. 
 
 Steel checker plate, 140 square feet, at 75 cents 
 
 Total for siphon alone 
 
 Cost per foot for siphon alone. 
 Total, including bridge, etc... 
 Total cost per foot'. 
 
 Tunnel, station 705+50 to station 708+50, 300 feet (all earth): 
 300 feet, at $45 
 
 Tunnels, statioii 723+50 to station 781+20, 5,770 feet (all earth): 
 
 5,770 feet, at $45....'.........:.. 
 
 Add for contingencies, street repairs, etc 
 
 Intake gatehouse, McMillan Park reservoir: 
 
 Excavation, earth, 2,180 cubic yards, at 60 c^nts. 
 
 Masonry, concrete, 1,166 cubic yards, at $8 
 
 House, 966. square fpeti at $4....... ■ 
 
 Centrifugal pump and motor 
 
 Iron, guides, ladders, 3,350 pounds, at 10 cents... 
 
 Gates, 3iby 9 feet, 2 at $900 
 
 Stop planks, 1,000 feet b. m., at $60 1 
 
 Reenforcing steel, 3,170 pounds, at 4 cents 
 
 Railing, 20 linear feet, at $1.50 , 
 
 Checker plate, steel, 70 square feet, at 75 cents. . . 
 
 $10,833.00 
 59, 358. 00 
 14, 220. 00 
 2,590.00 
 11,902.00 
 78,568.00 
 
 177, 471. 00 
 28.39 
 
 3,060.00 
 
 9,725.00 
 
 84, 960. 00 
 
 1, 800, 00 
 
 204.00 
 
 720.00 
 
 47.00 
 
 20.00 
 
 25.00 
 
 864.00 
 
 328.00 
 
 175.00 
 
 40.00 
 
 101,968.00 
 
 3,370.00 
 
 21,875.00 
 
 4,842.00 
 
 1,820.00 
 
 17,304.00 
 
 56,750.00 
 
 25.00 
 
 260.00 
 
 315.00 
 
 105.00 
 
 106,666.00 
 
 68.85 
 
 208,634.00 
 
 134. 60 
 
 13,500.00 
 
 259, 700. 00 
 20,000.00 
 
 279,700.00 
 
 1, 308.00 
 
 9, 328; 00 
 
 3,864.00 
 
 833.00 
 
 335.00 
 
 1,800.00 
 
 60.00 
 
 127.00 
 
 30.00 
 
 52.00. 
 
 17,737.00 
 
WATJiB STJI^PLy OF THE DISTRICT OF COLiUMBJA. 
 
 79 
 
 Conduit from gatehouse to reservoir; 
 
 Excavation, eartii, 2,410 cubic yards, at 60 cents |1, 446. 00 
 
 Concrete, 137 cubic yards, at $8 .■ 1, 096. 00 
 
 Embankment, earth, 1,912 cubic yards, at 40 cents 765. 00 
 
 Loam, 118 cubic yards, at $1.30 154.00 
 
 3, 461. 00 
 
 Miscellaneous items: 
 
 Repairing present by-conduit at Dalecarlia reservoir 4, 000. 00 
 
 Repairing damages to Conduit road 24, 500. 00 
 
 Tunnel rights * 3,570.00 
 
 Manholes, 48, at $25 1, 200. 00 
 
 33, 270. 00 
 
 Swnmary: 
 
 Land 122, 630. 00 
 
 Clearing and grubbing 2, 400. 00 
 
 Intake and gate house at Great Falls 48, 559. 00 
 
 Tunnel, station to 92 " 414,000.00 
 
 Crossing, station 94+78 to station 96+28 13, 104. 00 
 
 Masonry aqueduct, station 92 to station 235 317, 077. 00 
 
 Blow-off station 172 1, 642. 00 
 
 Masonry aqueduct, station 235 to station 267 » 76, 690. 00 
 
 Siphon, station 257 to station 259 15, 689. 00 
 
 Masonry aqueduct, station 267 to station 321+42 124, 902. 00 
 
 Siphon, station 321+42 to station .327 54,264.00 
 
 Masonry aqueduct, station 327 to station 416 211, 517. 00 
 
 Masonry aqueduct, station 416 to station 451 77,. 587.1 00 
 
 Masonry aqueduct, station 451 to station 465 33, 097. 00 
 
 Tunnel, station 465 to station 473+70 39, 150:00 
 
 Intake gatehouse, Dalecarlia reservoir ' 72, 000. 00 
 
 Masonry aqueduct, station 474+20 to station 508+50o 84,22L00 
 
 ' Tunnel, station 481 to station 490. 40, 500. 00 
 
 Under- water section, by-conduit, Dalecarlia 19, 134. 00 
 
 Outlet gatehouse, Dalecarliai 24; 645. 00 
 
 Tunnel, station 508+50 to station 642+50 602, 800. 00 
 
 Masonry aqueduct, station 642+50 to station 723+50 <• 177, 471. 00 
 
 Siphon and bridge, Rock Creek 208, 634. 00 
 
 Tunnel, station 705+50 to station 708+50 (Sixteenth street) 13, 500. 00 
 
 Tunnel, station 723+50 to station 781-1-20 279, 700. 00 
 
 Intake gatehouses, McMillan Park reservoir 17, 739. 00 
 
 Connection McMillan^Park gatehouse to reservoir 3, 461. 00 
 
 Coagulating eqiiipment for future installation 300, 000. 00 
 
 Miscellaneous.... : 33,270.00 
 
 Engineering and contingencies 450, 000. 00 
 
 Total estimated cost . . . : 3, 879, 383. 00 
 
 Condensed summary: 
 
 Aqueduct, standard section, 44,622 feet 1, 102, 562. 00 
 
 Tuimel, standard section, 30,440 feet , 1, 389, 650. 00 
 
 Crossings and sections of unusual type, 2,918 feet 310, 825. 00 
 
 Gatehouses, other atructures, land, etc 626, 346. 00 
 
 Engineering and contingencies 450, 000. 00 
 
 Total estimated cost..... 3,879,383.00 
 
 Estimated cost of construction per linear foot: 
 
 Aqueduct, standard section 24. 71 
 
 Tunnel, standard section 45. 65 
 
 Crossings and sections of unusual type 106.50 
 
 Estimated total cost of entire system per linear foot 49. 60 
 
 "Excepting certain portions in siphons, tunnels, etc., which are estimated sepa- 
 rately. 
 
80 WATER SUPPLY OF THE DISTBICT OF COLUMBIA. 
 
 Itemized estimate of cost of Great Falls gravity project, tunnel location. 
 
 Land, 21.5 acres, at $150 $3, 225. 00 
 
 Land, 6.8 acres, at $17,500 : 119, 000. 00 
 
 , Clearing and grubbing, 21 acres, at $160 3, 360. 00 
 
 Great Falls gatehouse (identical with that given in estimate for paral- 
 lel location; 48, 559. 00 
 
 Tunnel, station to station Z90, rock {25 per cent timbered): ' 
 
 29,000 feet, at $45 1, 305, 000. 00 
 
 Masonry aqueduct, station 29& to station 298+84 (reduced masonry sec- 
 tion): 
 
 Excavation, earth, 3,210 cubic yards, at 50 cents : 1, 605. 00 
 
 Excavation, rock, 2,799 cubic yards, at $2.50 6, 997. 00 
 
 Embankment, ordinary, 3,426 cubic yards, at 40 cents 1, 371. 00 
 
 Top soil, 888 cubic yards, at $1.30 1,155.00 
 
 Concrete, 1,380 cubic yards, at $8 11, 040. 00 
 
 Length, 864 feet 22, 168. 00 
 
 Cost per foot 25. 65 
 
 Cabin John Bridge and siphon, station 298+64 to station S05+04: « ===== 
 
 Bridge (complete as in parallel location) 17, 968. 00 
 
 Siphon, 640 feet, at $65 » 41, 600. 00 
 
 59,568.00 
 
 Masonry aqueduct and all other structures from station 305+04 on 
 tunnel location (station 327 parallel location) to McMillan Park 
 reservoir, identical with those given in the estimate for parallel 
 
 location 2,205,156.00 
 
 Miscellaneous: 
 
 Bepairing present by-conduit at Dalecarlia reservoir 4, 000. 00 
 
 Repairing damages to Conduit road 10, 500. 00 
 
 Tunnel rights 2, 550. 00 
 
 Manholes, 24 at $25 600. 00 
 
 17, 650. 00 
 
 Summary of cost of proposed aqueduct, tunnel location: 
 
 Land 122, 225.00 
 
 Clearing and grubbing 3, 360. 00 
 
 Intake and gatehouse, Great Falls 48,559.00 
 
 Tunnel, station to station 290 1,305,000.00 
 
 Masonry aqueduct, station 290 to station 298+64 22, 168. 00 
 
 Siphon and bridge, station 298+64 to station 305+04 59, 568. 00 
 
 Aqueduct, tunnel and other structures, from this point to McMillan 
 
 Park reservoir 2,205,156.00 
 
 Miscellaneous 17,650.00 
 
 Engineering and contingencies 550, 000. 00 
 
 4,333,686.00 
 
 Condensed summary: 
 
 Aqueduct, standard section, 22,894 feet 606, 061. 00 
 
 Tunnel, standard section, 50,240 feet 2,280,650.00 
 
 Croasings and sectiofls of unusual type, 2,650 feet 287, 336. OO 
 
 Gatehouses, other structures, land, etc 609, 639. 00 
 
 Engineering and contingencies 550, 000. 00 
 
 Total estimated cost 4,333,686.00 
 
 Estimated cost of construction per linear foot: =^=== 
 
 Aqueduct, standard section 26. 46 
 
 Tunnel, standard section 45. 40 
 
 Crossings and section of unusual type 108.40 
 
 Estimated cost of entire system per linear foot 57.20 
 
 a Station 305+04 of tunnel location is station 327 of parallel location. 
 b This is cost per foot of siphon alone for parallel location. 
 
WATER SUPPLY OF THE DISTRICT OF COLUMBIA. 81 
 
 Itemized estimate of cost of project for additional storage at the Stubblejield site (adaptable 
 only to the parallel location). 
 
 Land, 181 acres, at $200 $36,200.00 
 
 Clearing and grubbing, 36.2 acres, at $160 5, 790. 00 
 
 Stubblefield reservoir, embankment, etc.: 
 
 Excavation, earth, 164,724 cubic yards, at 50 cents 82, 362. 00 
 
 Embankment, rolled, 823,988 cubic yards, at 55 cents 453, 194. 00 
 
 Concrete, core wall, 51,780 cubic yards, at $8 414, 240. 00 
 
 Slope paving, 43,640 square yards, at $1.80 78, 600. 00 
 
 1, 028, 396. 00 
 
 Gatehouses and sluice tower: 
 
 Inlet and outlet gatehouses, just alike and exact duplicates of 
 
 Dalecarlia inlet gatehouse, cost of 2 gatehouses 100, 050. 00 
 
 Blow-off tower — 
 
 Excavation, earth, 356 cubic yards, at 60 cents 214. 00 
 
 Concrete, 2,611 cubic yards, at $8 20, 888. 00 
 
 Cast-iron pipe, 48-inch, 600 feet, at$14 8,400.00 
 
 Sluice gate, 4-foot 550. 00 
 
 Iron, ladders, etc., 1,000 pounds, at 10 cents 100. 00 
 
 30, 152. 00 
 
 Summary: 
 
 Land ^.... 36, 200. 00 
 
 Clearing and grubbing 1 . . . 5, 790. 00 
 
 Embankment, paving, etc .■ 1,028,396.00 
 
 Inlet gatehouse 50, 025. 00 
 
 Outlet gatehouse 50, 025. 00 
 
 Sluice tower, etc 30, 152. 00 
 
 Engineering and contingencies 180, 000. 00 
 
 Estimated total cost of construction 1, 380, 588. 00 
 
 MR. ALLEN HAZEN'S REVIEW OF MR. LONGLEY's REPORT. 
 
 June 28, 1909. 
 
 Major: In accordance with your instructions I have examined Mr. Longley's work 
 and report in reference to securing an additional supply of water for Washington; and 
 I beg to report as follows: 
 
 Mr. Longley's conclusions in general axe well founded, and I agree with him entirely 
 in regard to all his main positions. There are some matters, however, that should be 
 investigated further before the question is finally settled; and this must therefore be 
 considered as a progress report and not as a final one. 
 
 It may be as well that the matter must be left in this way, because it now appears 
 that the present works will be suflBcient to serve the city for some years, and it is not 
 necessary to immediately build additional works; and itwill be better to defer decision 
 as to many matters concerning these new works until it becomes necessary to build 
 them. 
 
 Capacity of present works. — Before the construction of the filtration plant the present 
 works, according to the best information available, had a capacity of delivering water 
 at the rate of about 76,000,000 gallons per day in the city. The tests upon which this 
 rating was made were carried out after the most thorough cleaning that the aqueduct 
 has ever had, and (iey therefore represented conditions considerably more favorable 
 than average conditions. 
 
 The constructipn of the filtration plant increased the capacity of the aqueduct to a 
 notable extent. This was brought about by the pumping plant installed in connec- 
 tion with the filters. The use of this pumping plant allowed more head to be used up 
 in friction in the aqueduct than had previoudy been available, and therefore allowed 
 more water to be brought through it. On the basis of the old gaugings the calculated 
 
 H. p<ic. 347, 61-2—^ 
 
82 WATEB SUPPLY OF THE DISTEIGT OF COLUMBIA. 
 
 increase in capacity was about 20,000,000 gallons per day, or in round figures the total 
 capacity was raised to 95,000,000 gallons per day. An actual test, arranged and carried 
 out by Mr. Longley, under existing conditions, demonstrates that the capacity of the 
 structures as they actually stand is about 90,000,000 gallons per day. 
 
 The falling off from the 95,000,000 gallons computed from the earlier gauginga to 
 the 90,000,000 gallons actually found is clearly due to the fact that the structures at 
 the present time are not as clean and otherwise in good condition as they were when 
 the previous gau^ings were made. 
 
 The accumulation of sediment in the aqueduct and the growth of sponges and other 
 organisms and the deterioration of the inner surface of the aqueduct may be expected 
 to continue, and if not adequately cleaned the discharging capacity of the aqueduct 
 may be expected to be further decreased during a term of years. The difficulty of 
 cleaning the aqueduct thoroughly, so as to keep the sides entirely smooth, will beccme 
 more difficult as the consumption approaches the carrying capacity of the aqueduct. 
 It will therefore be prudent,, in considering the time during -which the present works 
 will serve, to rate them at something less than EO,000,OCO gallons, their present capac- 
 ity; and without attempting further calculation 1 shall take this limiting capacity for 
 the purpose of estimate at 85,000,000 gallons per day. 
 
 Mr. Longley has investigated the question of peak load, or the excess in the rate of 
 consumption at times over the annual average consumption. He has found that 
 under present conditions the winter consumption in periods of extreme cold weather 
 may rise to 45 per cent above the annual average rate. He is right in assuming that 
 this excess winter consumption will largely be cut off with the gradual installation 
 of meters on the services; but when this is done there will be left a summer peak 
 load, representing the extra use of water in watering streets and lawns, and for other 
 uses in continued hot and dry weather. 
 
 There is no evidence to determine just how great this summer peak will be. I have 
 examined the record of the daily use of water since the filter plant was started, and 
 also general experience in other cities; and I do not think it safe to count on this 
 summer peak being less than 30 per cent above the annual average rate. 
 
 On this basis the extreme capacity of the aqueduct will be required at time of 
 maxiinum consumption in summer, when the average annual consumption is about 
 65,000,000 gallons per -day. In other words, w^hen the average annual consumption 
 reaches 65,000,000 gallons per day; th& use during hot, dry weather may be reasonably 
 expected at times to reach 30 per cent more than this, or 85,000,000 gallons per day, 
 the capacity which I take as a safe one for the present works through a reasonable 
 period of the future. 
 
 During the last five years the consumption has reached and exceede.d at times these 
 figures. If no new conditions had been introduced, there would seem to be urgent 
 need for an additional supply. This was thecondition when Mr. Longley's studies 
 were undertaken. 
 
 Use and waste of water. — During the past year there has been a great decrease in the 
 consimiption of water. This has been brought about largely by the efforts of Mr. 
 W. A. McFarland, superintendent of the distribution system, which is managed by 
 the District of Columbia. Mr. McFarland's work in finding and stopping leakage 
 can not be too highly commended. The reduction in consumption has been brought 
 about in part by the partial metering of the services and in part by a study of the 
 flow of water in the pipes in the streets and a detection and stoppage of leakages in 
 the mains and in service pipes. 
 
 The results already accomplished have so far reduced the consumption that it is 
 clear that no additional works for the supply of water will be required for several years. 
 The continuation of the work of stopping leaks by Mr. McFarland is certain to result in 
 further saving of water and it will still further extend the period during which the 
 present works will suffice. 
 
 The postponement of new works is iiot to be regarded as indefinite. Washington 
 is growing rapidly and there is every indication that the growth will continue. A 
 liberal use of water is desirable and necessary. Mr. McFarland's effort is to cut off 
 needless waste but not to restrict any legitimate use. With the waste cut off there 
 will still remain a large use of water and a larger supply than the present one will be 
 Required at a period that is not very remote. 
 
 It maj; be that cutting off the waste will extend the period during which the present 
 works will suffice, by as much as from ten to twenty years. The exact period can 
 Iiot be determined, as it depends upon the rate of growth of the city,- the effectiveness 
 of the measures taken for suppressing waste, and upon the actual use of water that will 
 remain; and none of these matters can be determined with certainty. Whether the 
 iteriod proves to be longer or shorter, the works for additional supply have ceased 
 
WATEK SUPPLY OF THE DISTRICT OF COLUMBIA. 83 
 
 to be necessary at the present time, and deliberate attention can be given to selecting 
 the best means of getting more water when the time for it arrives. 
 
 Metering government buildings. — I recommend most strongly the installation of meters 
 on all services for public buildings. Even though the water is not paid for by meter 
 measurement, the meters on the services will allow accurate knowledge to be had of 
 what water is drawn and at what times, and comparison can be made between the 
 actual drafts and the actual needs and uses of water. Carrying out this policjr is 
 certain to disclose leaks in services and fittings in government buildings now taking 
 place, the stoppage of which will represent a saving in water equal to an added capacity 
 of the present works reaching in value many times beyond the cost of the meters 
 required to demonstrate the co4;iditions and allow the leaks and wastes to be stopped, 
 
 Flashboards on Great Falls dam. — ^The capacity of the conduit is figured for extreme 
 low water conditions in the Potomac Eiver. Under all other conditions more water 
 will flow through it. The difference is not very great, perhaps only a matter of 
 3,000,000 to 5,000,000 gallons a day. I have not been through the calculation needed 
 to determine the amount exactly, so the figure mentioned must be taken as only a 
 roiigh approximation. Mr. Longley believes that it would be unsafe and unwise to 
 raise the Great Falls dam on its present base, and I fully agree with this conclusion; 
 but iron pins can be put in holes drilled in the top of the stone dam and flashboards 
 placed against them, raising the water by from 1 to 2 feet without very great expense. 
 The pins should be calculated so that they will bend over and allow the boards to 
 escape before the flood flow reaches such a height that there would be any increased 
 strain on the dam beyond that represented by ordinary floods without flashboards. 
 With this done it is perfectly safe to use flashboards in this way. I regard this simply 
 as a temporary measure that can be taken in the years when the consumption approaches 
 dangerously near the capacity of the conduit. 
 
 Avxiliary pump at Dalecarlia. — ^Mr. Longley inentions in his report the installation - 
 of a pumping station at Dalecarlia to lift the water as it comes from the aqueduct 
 to the reservoir and thereby allow the hydraulic slope of both ends of the aqueduct 
 to be increased. This would increase slightly the carrying capacity of the aqueduct. 
 I am disposed to think that a very much simpler and cheaper pumping station than 
 that proposed by Mr. Longley can be used for this service. I would suggest a single 
 large screw pump driven by an inexpensive steam engine, supplied by common 
 fire tube boilers, without any of the appliances for securing economy which were 
 included in Mr. Longley's estimate. The Duilding to contain the pumping machinery 
 also might be temporary and of an inexpensive character. This pumping station 
 should be regarded as only for temporary use when the consumptioti approaches near 
 the carrying capacity of the aqueduct, and regarded in this way the few million gallons 
 additional daUy capacity, would be very cheaply obtained. All needed apparatus 
 can be installed without interfering with the normal operation of the aqueduct, and 
 whenever a new aqueduct is built the pumps may be removed, as it would not be 
 worth while to keep them with a wide margin of capacitjr.otherwise available. 
 
 Stability of present aqueduct.— The question has been raised as to the stability of 
 some of the structures in the present aqueduct, and it has been suggested that as a 
 matter of insurance a second aqueduct ought to be provided so that in case of failure 
 of any of the structures the supply to the city would be maintained. 
 
 This is certainly an important matter. Mr. Longley has investigated the condition 
 of the structures of the aqueduct. I have not investigated them personally, but I 
 have talked over with him his investigations and have questioned him concerning 
 many matters. 
 
 In general, the structures of the present aqueduct are of a stable character, and there 
 does not seem to be much reason to apprehend accidents which will shut oS the supply 
 of water. Even in case of the actual collapse of a section of the aqueduct there are 
 several days' supply of water always held in the reservoirs; and with the facilities of the 
 Engineer Corps for putting men promptly on the ground for repairs in an emergency 
 it would ssem possible to construct timber trestles and flumes to bridge over in a tem- 
 porary way any section of aqueduct that might be so destroyed, before the water in the 
 reservoirs was exfiausted, and so maintain the supply. 
 
 Every reasonable effort should be made to ascertain the condition of all the struc- 
 tures of the aqueduct, and to strengthen any points that may be found weak; and with 
 this done there does not seem to be any risk in the use of the present structures which 
 justifies at this time the construction of a new aqueduct not otherwise required. 
 
 Best arrangement for additional supply. — I have examined Mr. Longley's survey for 
 an additional aqueduct from Great FaUs to McMillan Park reservoir, and it seems to 
 me that the route selected by him is well chosen and the structures proposed by him 
 are in general suitable and adequate. 
 
84 WATEB SUPPLY OF THE DISTRICT OF COLUMBIA. 
 
 A number of thoughts concerning possible improvements in details and matters of 
 secondary importance have occurred to me, and these I have suggested to Mr. Longley. 
 It seems to ine unnecessary to take up a discussion of these points in detail, as under 
 present conditions the immediate construction of an additional agueduct is not neces- 
 sary; and it will be time enough to take up its details when it is necessary. 
 
 Cost of a second aqueduct from Great Foils.— I have been over Mr. Longley's estimates 
 of cost, and in the light of general experience it is my judgment that a new aguedurt, 
 iequal in capacity to the present one, with the auxiliary works and connections that 
 would naturally go with it is likely to cost fully $4,000,000. 
 
 This estimate is for an aqueduct delivering water at the present level of McMillan 
 Park reservoir, from which level it is pumped to the filters. It is not possible, from 
 Great Falls, to deliver water at McMillan Park by gravity at a sufficient elevation so 
 that this pumping can be dispensed with. 
 
 Cost of pumping. — The cost of pumping to the filters at McMillan Park, including 
 the capital charges on the machinery, is probably not less than $2 per million gallons. 
 For 65,000,000 gallons per day average annual use to be ultimately expected from an 
 aqueduct of this size, this amounts to $47,500 per annum, which is 4 per cent interest on 
 $1,190,000. 
 
 Possibility of a gravity supply from a higher level. — If water could be obtained at a 
 higher level it might be possible to eliminate the cost of pumping the new water at 
 McMillan Park, and also if the elevation of the water at the source were great enough, 
 it would be possible to build the new aqueduct on a steeper slope; and if this could be 
 done water would flow in it more rapidly and a smaller aqueduct would bring the same 
 quantity of water, and the cost of the aqueduct per foot would therefore be reduced. 
 
 The fall from Great Palls to McMillan Park reservoir is less than 1 foot per mile, an 
 extremely flat slope for the economical construction of an aqueduct. A source higher 
 than the Potomac at Great Falls would be more desirable in these respects. 
 
 Other sources of supply. — If the point of intake were moved from Great Falls to a 
 more distant point on the Potomac, the water could be secured at a higher elevation. 
 There are also various tributaries of the Potomac which would be available at higher 
 elevations. Mr. Longley has studied one or two such possible arrangements in a pre- 
 liminary way, but the distances involved to secure an increase in elevation that is 
 worth going after are so great that the projects are not flnancially attractive. There 
 does not seem to be any reason for considering the Potomac further up or the upper 
 tributaries at a considerable distance. 
 
 Rock Creek. — Mr. Longley has made surveys for the development of Rock Creek as 
 a source of water supply. The water could be taken within the District of Columbia 
 and carried by a steel pipe about 5 miles long to McMillan Park reservoir and de- 
 livered at an elevation great enough so that it would not require pumping to reach 
 the fllters. A large dam and reservoir would be required. My calculations indicate 
 that with 12 inches of storage some 40,000,000 gallons per day would be available for 
 supply while leaving a certain amount of water flowing in the stream for park pur- 
 poses. 
 
 f This development in many respects is an attractive one, although the dam and 
 reservoir would have to be built in a region being built up and developed to such an 
 extent as to raise a question as to its use for water-supply purposes. The development 
 would consist of the construction of a reservoir with a flow line at about elevation 227. 
 The dam would hold water to a maximum depth of about 77 feet and would be about 
 700 feet long. The reservoir would have an area of some 1,700 acres and would contain 
 14,000,000,000 gallons of water. Such a reservoir would be filled by the winter and 
 spring floods from a tributary area of over 60 square miles, and the water contained in 
 it, without further help from the catchment area, would serve to maintain the supply 
 at the present rate of draft through seven months of extremely dry weather. If it were 
 hot for the developments .on Rock Creek this would seem to be clearly the most ad- 
 vantageous source lor an additional supply to the limit of the capacity as above stated. 
 ■Wrai the developments of roads, railroads, parks, buildings and other structures as 
 they are, it may be seriously questioned whether it will be advantageous to attempt it. 
 
 Patuxevt River. — This source of supply is not mentioned in Mr. Longley's report, 
 because it was only considered near the end of the inveptigation, when there was 
 neither time nor money available for Mr. Longley to investigate it adequately. The 
 information concerning it is from the government topographical maps, supplemented 
 by a rapid preliminary inspection in an automobile and on foot, of some of the points 
 involved. 
 
 The stream could be taken almost directly north of McMillan Park reservoir, at a 
 distance of about 16 miles, or at about the same distance as the present intake at Great 
 Falls. The present water level at a point about a mile above the Columbia turnpike 
 
WATER SUPPLY OP THE DISTRICT OP COLUMBIA. 85 
 
 troasing, indicated by the Geological Survey maps and checked by aneroid barometer, 
 is 250 feet above tide, or 85 feet above the water on the filters at McMillan Park. 
 
 "The catchment area of the streain at this point, measiired from the government maps, 
 is about 130 square miles. For some distance up and down stream both sides are ol 
 ledge, and the ledge frequently approaches the stream, so that I estimate that a dam 
 70 or 80 feet high would not be more than 600 or 800 feet long. Ledge is also exposed 
 here and there in the bed of the stream. The Geological Survey maps, which must 
 be taken as only very rough approximations, indicate that a dam 90 feet in height 
 would flow the water back 9 miles, forming a long and generally narrow reservoir, but 
 with a number of considerable branches on tributary streams. The land in the bottom 
 and sides of the valley is not especially valuable, and only a few small and inexpensive 
 buildings would be flooded. 
 
 A reservoir formed by a dam at some site to be selected, probably within a mile of 
 the Columbia turnpike, either up or down, holding 12 inches run-off from the catch- 
 ment area, would make available in a dry year 100,000,000 gallons of water per day* 
 The water would be available at a sufficient elevation so that it could be taken to the 
 filters at McMillan Park by gravity, and the cost of pumping would be saved. 
 
 A tunnel 3 miles long, more or less, through a flat ridge would take the water into a 
 valley, from which a direct and apparently easy route for a steel pipe line would lead 
 to the filters. The head would be sufficient so that a 6-foot steel pipe would carry as 
 much water as could be brought from Great Palls by a 9-foot aqueduct. 
 
 The catchment area is mostly good farm land, under general cultivation. It is rolling, 
 the high points being comparatively flat, with the valleys deep cut between. All of 
 the villages and the best farm buildings are on the summits. The villages are small 
 and are scattered along the rim of the catchment area, in general about one-half of 
 each village draining toward the stream. There are no considerable villages on the 
 catchment area, and no lines of steam railroad or- trolleys running through it. I 
 estimate the population very roughly at 30 per square mile, and probably not in- 
 
 The indications are that a stfeel pipe line and tunnel from this source could be built 
 for very much less than the cost of an aqueduct from Great Palls. On the other hand, 
 it would be necessary to build a dam and storage reservoir, and to acquire the property 
 to be flowed, and to raise various roads and build bridges for them, and large and 
 undetermined amounts of money would be required for these purposes. 
 
 The information now at hand is not suflScient to determine whether or not such a 
 supply could be more advantageously obtained than to build an additional aqueduct 
 from Great Palls. The project is sufiiciently promising, so that it would not seem 
 wise to determine upon the construction of an additional aqueduct from Great Falls 
 until the possibilities of the Patuxent River are determined by preliminary surveys, 
 covering the sites for a dam and reservoir and a tunnel and pipe line, and until esti- 
 mates of the cost of the required work are made. 
 
 A comparison of the quality of the water to be obtained with the quality of the 
 Potomac water should also be made. From a sanitary standpoint, the use of water 
 from an impounding reservoir fed by a sparsely settled catchment area would have 
 advantages over the use of Potomac water. The questions of the amount of turbidity 
 in the water collected by such a reservoir, and the rate at which it would settle out, 
 and of what preliminary treatment, if any, would be required before taking the water 
 to the filters at McMillan Park, present problems that can certainly be solved, but 
 which will require considerable study to insure solving them in the most advantageous 
 ways. 
 
 Capacity of the present filters. — The filters were designed to purify the water that 
 could be secured through the present aqueduct when filtering at the rate of about 
 3,000,000 gallons per acre daily. Mr. Longley has shown the need of preliminary, 
 chemical treatment of Potomac water at times. With preliminary chemical treat- 
 ment adopted as part of the system, there is no question but what the results of filtra- 
 tion would be as good, and in some respects better, with the filters operated at a six- 
 million rate than as they now are, with the filters operated at one-half that rate. If 
 the new source should be an impounding reservoir upon the Patuxent, there is also 
 little question that the water could be filtered at a six-million rate witii satisfactory 
 results. Whatever the source of the new supply, it therefore seems possible, with 
 suitable preliminary treatment, to count on increasing the rate of filtration as may be 
 necessary, so that the whole of the supply to the extent indicated may be filtered at 
 the present works. 
 
 Increasing the rate of filtration in this way will no doubt involve some changes in 
 the piping and other arrangements at the filter plant. The present piping and arrange- 
 ments are sufficient for the treatment of much more water than can be delivered 
 through the present aqueduct, and it would only be necessary to make these changes 
 
86 WATER SUPPLY OF THE DISTRICT OF COLUMBIA, 
 
 when there was need for doing it. This question is therefore one of a future so remote 
 that there is no occasion for going into it further at the present time. 
 
 Other measures suggested. — There are a f^w things which, it seems to me, can be 
 done to advantage to the system, and although perhaps not strictly covered by your 
 instructions, it may be well to mention them. 
 
 (1) Reconstruction of the Georgetown reservoir.— This has been considered at various 
 times, and preliminary plans have been made for doin^ it. Reconstructing it will 
 make it more efficient as a sedimentation basin, especially in connection with the 
 preliminary chemical treatment of Potomac water, which has been recommended. 
 
 (2) Preliminary chemical treatment. — This has been considered as a part of the sys- 
 tem of purification from the start. It was a part of the process as originally recom- 
 mended by Messrs. Hering, Fuller, and the writer. The plant has never been com- 
 pleted in this respect. Completing the plant at this time involves the expenditure 
 of only a small amount of money, and would have the great advantage that it would 
 allow bright, clear water to be delivered at all times. At present such water is deliv- 
 ered for the greater part of the year, but at times of turbid water in the Potomac the 
 filtered water as delivered in the city is more or less milky and unsatisfactory to people 
 accustomed to the use of perfectly clear, bright water. 
 
 (3) Covering the highrservice distributing reservoirs. — Filtered - water deteriorates 
 when stored in the sunlight. For this reason the distributing reservoir for the low 
 service, built in connection with and as a part of the filtration plant, is covered. The 
 older distributing reservoirs connected with the several high-service systems are not 
 covered. When raw Potomac water was used in the city there was no advantage in 
 having the distributing reservoirs covered. Under present conditions there are at 
 times growths of microscopical organisms in all of these high-service reservoirs which 
 impart to the water which passes through them tastes and odors that are disagreeable 
 and unsatisfactory to the consumers. This condition could be entirely corrected 
 by covering these reservoirs, and, in my judgment, this should be done. These reser- 
 voirs belong to and are operated by the District of Columbia and not by your office. 
 
 (4) Inspection of present aqueduct. — A careful inspection of all the structures on the 
 present line of the aqueduct, with special lookout for any signs of weakness, and fol- 
 lowed up by such repairs and strengthening aa may be indicated at any time. 
 
 (5) Meters. — A continuation of the meter system until all services are metered. 
 This is the only adequate way of stopping unnecessary and constant waste through 
 defective plumbing and fixtures. 
 
 (6) Leakages. — ^A continuation of the present work of Mr. McFarland in discovering 
 and stopping leaks from mains and service pipes. 
 
 Some of the matters above enumerated are under way and are being continued, 
 while others have not yet been taken up. In my judgment all of them should be 
 taken up and followed to completion, together with such additional surveys as are 
 needed to show whether or not the Patuxent River or any other source may be a better 
 source for additional supply than the Potomac River. 
 
 With all these measures under way there is no reason for undertaking at the present 
 time the construction of an additional aqueduct. The money that would otherwise 
 be required to pay the interest alone on the cost of a new aqueduct will pay many 
 times over the cost of all that is herein suggested. 
 
 Respectfully, Allen Hazen. 
 
 Maj. J. J. Morrow, 
 
 Corps of Engineers, Washington, D. C. 
 
 o