ST. LOUIS WATER SUPPLY.
Excerpts From the Majority Report of Engineers Benezette Williams and G. Y. Wisner.—The Meramec Scheme Indorsed.—Some Historical Facts as to the Growth of the Waterworks System.
On July 25, 1901, a commission of hydraulic engineers was appointed to examine and investigate the existing water plant of the city and to report as to the most feasible manner of providing an adequate supply, both present and prospective, of clear, wholesome water; the aggregate cost of the commission to be $25,000. The commission was composed of Benezette Williams, George Y. Wisner and Allen Hazen. The last named engineer formed the minority.
The report of the majority is very full of details, and covers the field dealing with the present system and future sources of supply. Following are some of the principal points brought out by the majority committee:
The city of St. Louis, with a present population of about 600,000 people, and increasing at the rate of 14,000 per year, has requirements in regard to the present and future water supply of the city which practically eliminate all the projects which have been considered except those of filtering the waters of the Missouri and Mississippi rivers, and of obtaining a supply by gravity flow from the watershed of the upper Meramec river. The examinations and investigations have been made, therefore, with the view of determining the relative merits of these two projects with reference to the cost, capacity of supply and quality of the respective waters for domestic and manufacturing purposes.
Complete plans and estimates for a filter plant for filtering the water from the Mississippi river have been prepared, and a careful study made of the present system of pumping, and of the changes which will be needed in the pumping plants and distribution mains, in order to insure a satisfactory service in the future for all parts of the city.
Careful studies have been made of all the available data to be obtained bearing upon the chemical and bacteriological characteristics of the different waters from the two sources of supply, of the chemical treatment necessary to render the Mississippi river water clear and wholesome, and to the effect of such treatment on the effluent to be obtained from the filters.
In this connection the members of the commission visited and observed the operation of the sedimentation and filter plants at East St. Louis, Kansas City, St. Joseph, Omaha, Davenport, Rock Island and Quincy. While the water at all of these plants is greatly improved by the treatment it receives, an absolutely clear and wholesome effluent is not generally obtained. In fact, aside from a few experimental plants, operated, on a small scale by experts, there are no filter plants in the United States where a perfectly clear and wholesome effluent is continually produced from water similar to that of the Mississippi river at St. Louis.
Very little is known of the chemical and bacteriological constituents of the effluents at these different plants, for the reason that regular tests are made at only a few of the stations.
As appears from the St. Louis waterworks reports, the original plant in use, prior to organisation of the second board of water commissioners, March, 1867. consisted of two pumping engines, having an aggregate capacity of about 12,000,000 gallons per day. These engines were installed, one in 1853, the other in 1855, at what was known as the Bates street station .At that time there were also in use two reservoirs having an aggregate capacity of 52,000.000 gallons. There were also about eighty-one miles of distribution pipe, valued at the time the new water board was organised at $1,349,589.
The task of constructing and establishing new works was immediately entered upon, Bissell’s point being selected as the site for the pumping works and subsiding basins, and Compton Hill as the site for a new reservoir. The capacity and condition of the old works being insufficient to supplv the city with water during the construction of the new works, temporary pumps were installed, and a supplementary reservoir was built on Gamble and Dayton streets, known as the Gamble street reservoir. The works at Bissell’s Point consisted of two river pumping engines, or low-service engines, as they are called, having a capacity of 16,000,000 gallons each, installed in 1870. Another one of 18.000.000 gallons was installed in 1874, making a total capacity for the river station of 50,000,000 gallons. These engines pumped into four settling basins, 600 feet by 280 feet in size, having an average depth of about sixteen feet of water, planned to draw twelve feet, with an available capacity of 15,000,000 gallons each. From thence the water went to the pumps that supplied the city. Of these there were installed in one pumping station, two, with 18,000,000 gallons capacity each, in 1870, and one in 1874 of 20,000,000 gallons capacity.
Subsequently pumping station No. 2 was built at Bissell’s Point, in which three 18,ooo,ooo-gallon pumping engines were installed; one in 1884, one in 1888, and one in 1893—the total pumping capacity in the two stations supplying the city, at Bissell’s Point, being 110.000,000 gallons.
During the early construction of the Bissell’s Point works the Compton Hill reservoir was built, with a capacity of 60.000.000 gallons, and standpipes Nos. 1 and 2. situated near the Bissell’s Point station.
The Bissell’s Point pumping station having been outgrown, new river works were established and put into operation at the Chain of Rocks in 1895. These works comprise a masonry inlet tower, an inlet tunnel built in rock from the works to the tower, and pumping works, in which engines have been installed as follows: Two ,20.000,ooo-gallon pumping engines in 1894, two 30,000,000-gal Ion pumping engines in 1895, two 30,000,000-gallon pumping engines in 1898; making a total pumping capacity at this station of 160,000,000 gallons. Connected with these works are six settling basins of about 29,000,000 gallons canacity each; each basin being 660×400 feet in dimensions.
Water from the Chain of Rocks is conducted to Bissell’s Point through a conduit of 130,000,000 gallons canacity.
A third high-service pumping station has been established at Baden, on the line of the conduit leading from the Chain of Rocks to Bissell’s Point. At this place two 10,000,000-gallon pumps and two 15.000,000-gallon pumps were installed in 1898, and there are now in process of erection at this sta tion two 15,000,000-gallon pumps, one of which has been put into operation, and the other one of which will soon be ready for use. When these are completed. the total pumping capacity at this station will he 80.000,000 gallons.
Three 20,000,000-gallon pumns are in course of errection. and under contract to take the place of the first three pumps built at Bissell’s Point. These, it is expected, will be completed and ready for service within a year, which will make the total pumping capacity at Bissell’s Point 114,000,000 gallons. When these are complete the pumping machinery in the St. Louis waterworks will be as follows;
River pumping works at Bissell’s Point, three engines, total. 50,000,000 gallons capacity. These works are held only as a reserve in case of failure at the Chain of Rocks.
Six pumps at Bissell’s Point, supplying the lower district of the city, having a total canacity of 114,000,000 gallons.
Six pumps at Baden, supplying the higher district of the city, with a capacity of 80,000.000 gallons.
Six river pumping engines at Chain of Rocks, pumping into the settling reservoirs, having a total capacity of 160,000,000 gallons.
There has also been built, in connection with the Baden numping works, a third standpipe at Compton Hill, provided with overflows into the Compton Hill reservoir.
The distribution system, as it existed on April t, 1901. was made up of 638 miles of distribution pipe. 7,325 fire hydrants, 67,243 service taps and 4.331 meters.
The population of the city, by decades, from United States Census, has been as follows; 1820. 10.049; 1830, 14,125; T840. 16.469; 1850. 77,860: 1860, 160,773: 1870, 310,864; T88O, 350,518: 1890, 451,770: 7000. 575.238.
The total amount of water annually supplied to the city, at the time the second board of water commissioners was organised, was 2.432 million gallons, and the total revenue collected therefor, during the first vear of the board’s existence, was $288,910,000. Total supplied for the year ending April 1st, TOOT, was 22,994 million gallons, and the total revenue collected therefor was $1,599,960.
The cost of supplying water for thirty-four years ending April T. 1901. was as follows:
Total cost of supply works to April I. 1901, including the pumping works, inlets, settling basins, conduits, standpipes, Compton Hill reservoir, and ten per cent, of the force mains.$10,297,634
Total expenditures incurred in operating the supply works, including repairs, . and maintenance of pumping works, settling basins, standpipes and the Compton Hill reservoir. 6,427,371
The total amount of water furnished during these thirty-four years was 384,116 millions of gallons, which gives a cost of each million of gallons of $43.54. In this cost there is nothing included for interest or discount on bonds, nor for a renewal fund. If these were known and added, it would bring the cost of supplying water for the thirtyfour years to at least $50 per million gallons, which represents the cost of delivering water into the distribution mains The cost per million gallons for maintaining the distribution system, collecting water rents and for general expenses, for the thirty-four years, averaged $8.43. For the last twelve years the average has been about $7 per million gallons.
.The total estimated cost of furnishing water into the distribution mains for thirty-nine years—1868 to 1906 inclusive—when it is assumed that new works will go into operation, is as follows:
Total expended to April t, 1901, as per preceding statement $16,725,005
Estimated cost of operating supply works from April 1. 1901, to April 1, 1906, 126,361 million gallons at $15. 1,895,415
Estimated cost of completing present contracts for pumps, etc. 748,399
Total expended to April Ist, 1906 $19,368,819
The amount of water which will have been supplied by April I, 1906, is estimated at 510,477 million gallons. Average cost per million gallons, $37.94. In this cost nothing is included for interest, discount on bonds, or for renewal fund. If these were included, the cost would probably be near $43 per million gallons. If the maintenance of the distributing system and cost of collection be added, it would bring the cost per million of galIons, not including the cost of constructing the distribution system, to over $50 per million gallons.
From investigations made upon water systems of cities where the ouantity of water used is measured, it appears that in large cities the average volumes needed for domestic, public and manufac luring purposes are respectively, about thirty gallons, five gallons and twenty-five gallons per day for each inhabitant, and that anv sunply in excess of sixty gallons per day per capita goes to underground leakage or unnecessary waste.
Heretofore the gross consumption in St. Louis has been kept below no gallons per capita daily, and with proper management there is no reason why this amount should be exceeded in the future, especially when, with a pure water supply, the necessity of flushing of mains and fixtures will be largely eliminated.
If the Mississippi river is to continue as the source from which water is to be drawn for the city of St. Louis, there can be but one opinion as to the desirability of the installation of a filtering ap paratus as an adjunct of the present pumping plant.
But, as previously stated, we have an alternative in a gravity supply of water from the upper Meramec watershed. Without this it would be a question merely between different plans for filtration, and the perfection of the plan adopted. With it there is required a critical inquiry as to which source promises the nearest approach to an ideal water supplv.
The study of water and water supplies, and the experiments instituted bv the State board of health of Massachusetts to determine the best methods of water purification, and the invention of the mechanical filter, has led to wider investigations directed to the clarification and purification of the very turbid waters of the Missouri and Mississippi rivers. as well as to the comparatively clear waters of New England.
The European practice of filtration, which Mr. Kirkwood imported into America, is known as slow sand filtration. It consists of passing the water slowly through beds of sand, from the surface of which the deposits are subseouently scraped. While this method is more successful than any other for water carrying a small amount of suspended solids, it has not met with success with very turbid waters.
The application of a coagulant to the raw water in filtration, bv the inventors of what are known as mechanical, or American filters, opened UP a previously unoccupied field in the water purification. This class of filters consists of a comparatively small tank containing sand, through which water previously treated with a coagulant is rapidly passed downward, until the surface of the sand becomes clogged. A current of water under pressure is then forced upward in reverse course, washing the sand, thus preparing it for another supply of the influent water,
Growing out of the experiments at Louisville and Cincinnati the Louisville Water company is now engaged in the construction of a rapid filtration plant, and it is understood that Cincinnati will do likewise. It may be said that there is no large plant in existence adapted to the treatment of such water as the Mississippi river furnishes at St. Louis, and no small one which furnishes much aid in such a work as your commission has in hand. Indeed, it may be said that no city has entered upon the solution of such a problem as is presented at St. Louis, because, though the conditions existing at Louisville and Cincinnati, while analogous in some respects, differ materialy from those at St. Louis.
The plan proposed for a filter plant involves baffle walls being placed in the present settling basins, of which there are six; each being 670 feet by 400 feet, with depth ranging from twelve feet to twenty feet and averaging 14.4 feet. The capacity of each basin is 29,000,000 gallons. The plan provides for five baffle walls going nearly across each basin, dividing it into six (6) divisions; the baffle walls to consist of A shaped frames built of steel angles, fastened together in pairs, so that each set would stand bv itself. These frames are to carry a covering of planks, or slabs of concrete, upon expanded metal.
After leaving the plain sedimentation basins, the water passes through a coagulating house where the coagulant is applied, thence through three coagulating basins, each being 500 feet by 260 feet, with a depth of from seventeen and one-half to twenty feet, total capacity of basins to bottom being 19,000,000 gallons each. From the coagulating basins the coagulated water passes through a conduit to the filters Between the basins and the filters is to be placed a coagulating chamber where additional Coagulant can be added to the effluent, if desired.
At Bissell’s Point, where the samples were taken from the basins in 1885 and 1886. the average time of subsidence was about thirty-six hours, the minimum being twenty-four hours. Under usual conditions of operation these times were lengthened a few hours,
During the years 1887 and 1888, when the samples were taken from the clear well, the time of subsidence did not differ materially from that of the two preceding years, though, owing to a small increase in the water consumed, the time of subsidence was shortened somewhat.
At the Chain of Rooks, when all the six subsiding basins were in use. the average time for settlement was sixty hours, and the minimum time forty hours. If one of the basins was out of use the average time was fifty-one hours, and the minimum thirty-four hours.
It is one of the objects of filtration to remove the necessity of special precautions against the propagation of disease through the water supply. It can be said that in those places where water has been taken from a polluted source, and a good filter plant has been put in. there has almost always been a reduction in the typhoid fever death rate: the exceptions thereto being caused by conditions having no relation to the water supply. It seems to be conceded that slow’ sand filters give the best results in the removal of bacteria, but that the rapid, or American filter, goes a long wav in that direction under the best conditions. The efficiency of filters has usually been expressed in percentages of the total number of bacteria in the raw water, but. as has been frequently pointed out. such manner of expressing the thoroughness of bacterial removal may be very misleading.
If the water contains 270,000 bacteria per cubic centimeter, as the Mississippi river at St. Louis does at times, ninety-nine per cent, removal would leave 2.700 bacteria in the water, which would be much below the standards generally recognised as essential.
There are practically no obtainable data showing bacteria efficiencies of rapid mechanical filters in use. the most, if not all. of the knowledge upon this subject being contained in the reports recording the determinations from experimental plants.
In the Louisville report, pages 232 and 234. is a Mintmarv of results obtained by the Warren and Jewell filters by periods. Deductions made from these two tables, by grouping those periods when the average sediments carried did not exceed too parts per million in any one period, and those in which the sediments averaged from 180 to 450 parts for the several periods, gives the following results:
On page 240 of the same report it is stated that the grand total averages for the entire investigation were as follows;
Warren filter.96.7 per cent, efficiency
Jewell filter.96 “ “ “
In the Cincinnati report, pages 32 to 39, is given the numbers of bacteria contained in the raw water at that place per cubic centimeter. These vary from a minimum of 1,000 to a maximum of 144.000. On pages 316 to 333 the percentages of efficiencies obtained, based upon the bacteria contained in the water subsided for seventy-two hours are given; the bacteria contained in the subsided water being considerably less than that contained in the raw water. These efficiencies vary greatly, though usually average from 95 to 99.8 per cent. The amount of sulphate of alumina used at Cincinnati averaged two grains per gallon.
At Pittsburgh the combined average results of the Warren and Jewell filters, as deduced from tables found in the Pittsburgh report, pages 172 to 179 and 194 to 201, are as follows:
Average amount of coagulant used, grains per gallon. 1.15; bacteria in raw water, per cubic centimeter, 12,028; bacteria in effluent, per cubic centimeter, 561; efficiency, 95.4 per cent.; omitting the month of January the average efficiency was, 96.8 per cent.
Experiments made on the Potomac river at Washington, with rapid mechanical filters, gave efficiencies averaging from 80 to 99.9 per cent.: the bacteria in the raw water being from 135 to 51,000 per cubic centimeter.
The number of bacteria contained in the Mississippi river at St. Louis greatly exceeds that in any of the waters experimented on in the foregoing references.
The plan of gravity supply upon which the estimates are based involves one reservoir of 200,000,000 gallons capacity, located about two miles west of Dwyer on the Clayton road, with a flow line elevation above sea level of 660 feet; another reservoir on the Clayton road, about 2,000 feet west of the city limits, and near to Forest park, with a capacity of 150,000,000 gallons, and a flow line elevation of 600 feet; also the improvement and use of the Compton Hill reservoir, with a flow line elevation of 589 feet and a capacity of 60,000,000 gallons.
The elevation of the water surface of the Meramec spring is about 781 feet above sea level, 370 feet above that of the Mississippi river at low stage at St. Louis, and 180 feet above the highest points in the city at which water will have to he supplied. Bv giving suitable dimensions to a conduit from the spring to the city, and using a reservoir with 700 feet elevation at Stratman, water may he supplied to the city mains with a head of eighty feet for the higher locations, and through distributing reservoirs in or near the city to the lower portions of the city at any head best adapted to the service.
From Meramec spring to the Compton Hill reservoir is about ninety miles in a direct line. The difference of elevation between the flow line of the reservoir and the spring is about 192 feet, which is all available to produce flow in conduit, and pines, for the water that goes direct to this reservoir. The difference in elevation between the spring and the proposed Forest Park reservoir is 121 feet, and between the spring and the proposed Dwyer reservoir eighty-one feet.
The water from the spring unites with the Water Fork branch of the river, about 4,000 feet below where the soring rises. Above the junction of the two streams the bed of the river rises above seven feet per mile to the dam at Wilson’s mill—a distance of about twelve miles. In the vicinity of Wilson’s mill there is a fall of about twenty feet in one mile, making that point a very favorable location for the upper end of a storage reservoir.
The following estimates are based noon the construction of a collecting reservoir of 41,500,000 gallons capacity in the valley of the Water Fork, a small dam and reservoir at the Meramec soring, a conduit, tunnel and nipe line from the collecting reservoirs to the distributing reservoirs at city, with a capacity of 225,000,000 gallons per day, a 200,000.000-gallon distributing reservoir at Dwyer, a 150.000,000-gallon distributing reservoir near Forest park, and a 60,000,00-gal Ion reservoir at Compton Hill, with connecting mains to the distribution system of the city. The distributing reservoirs are to be covered to prevent the growth of algae unless it is found from experiments during the construction of the works that covers for the reservoirs are unnecessary. The gross cost of constructing the Meramec system is $31,000,000, which may be reduced to $27,000,000 by credits from water receipts.
If the works should be ready for service in 1906, the amount of water required at that time would be only about 26,400,000 gallons per year, making the cost per million gallons about $52, whereas in 1946, with an annual consumption of 54,600,000,000 gallons, the cost per million gallons would be about $25.
A general estimate of the cost of supplying water by the pumping plan, based upon the actual outlay during fifty-five years for plant, renewals and operation, the latter being taken at $20 per million gallons. is $29.46 per million gallons. This includes nothing for interest, it being assumed that all expenditures will be met when they occur.
For the gravity plan the cost per million gallons is based on the total outlay for a period of fifty-four years for construction of works, interest on $27,000,000 of bonds, redemption fund, renewals, and 00eration ; it being assumed that at the end of the period the plant will be paid for and left without debt. This amounts to $23.30 per million gallons.
If the works are bonded for $31,000,000, the full amount of their estimated cost, less credits from sale of old works, the average cost per million gallons will he $26.
Relative to the adoption of either of the two projects presented, there are certain general considerations which may be appropriately mentioned.
The time necessary for the construction of works has been estimated in this report at three years for the pumping and filtration project, and at four years for the gravity supply plan.
From the results obtained by experiments at Louisville, Cincinnati, Pittsburgh, Washington and New Orleans, together with the radical differences which have been shown to exist between the water at these cities and that at St. Louis, it seems evident that it would be verv unsafe to design plans for the filtration of the Mississippi river water at St. Louis, without first making an exhaustive experimental investigation relative to the cost and requirements necessary to obtain a clear, wholesome supply of water by filtration.
Such investigations would likelv require from one to two years, making the time before a supply could he obtained by such a plan at least as long as required to build works for a gravity supply.
A gravity supply of water from such a source as the upper Meramec watershed has always been regarded as the nearest possible approach to the ideal, and granting all that can he claimed for artificial methods of purification, there is no reason to accord first place to any water, which either from natural or artificial causes has once become ladened with silt and organic pollution.
The commission, therefore. is convinced that in recommending the adoption of the Meramec supply, it is following a course dictated by considerations which cannot be ignored.
In the next isue of FTRE AND WATER will be given excerpts from the minority report.