REPORT ON IMPROVEMENT OF LOWELL WATER SUPPLY
James H. Carmichael, Commissioner of Water and Fire Protection, of Lowell, Mass., has forwarded to the city council, under date of June 18th, a report by F. A. Barbour, C. E., describing an investigation conducted by him of the present condition of the water supply and the conclusions reached as to the best method of providing the city with satisfactory water. The city has two sources of supply, the mirth hank of the Merrimac River and the valley of the River Meadow Brook. Both are ground water supplies, drawn through wells. That on the Merrimac—known as the Boulevard system, furnishes the greater part of the water used; the other, which includes two gangs of driven wells—the Cook and the Hydraulic—is only brought into service when inadequacy of the Boulevard supply makes this necessary, for the reason that in 1899 the State Board of Health condemned the Cook Wells because of the action of this water on lead pipe, and the consequent danger of poisoning. With continued use, and to a constantly increasing degree, the Boulevard supply has grown unsatisfactory, owing to the amount of iron and manganese in this water, which renders it turbid and altogether disagreeable. Three courses of improving the supply have been considered: treatment of the present supplies to render them safe and satisfactory, filtration of the Merrimac River and connection with the Metropolitan system. It was concluded that the city should further develop the Boulevard system and should construct works too for the removal of the iron and manganese, which alone prevent that supply from being satisfactory. The city council has since voted $325,000 to build the filtration plant and Commissioner Carmichael recently announced shat woik on the plant is to commence as soon as the money is available.
In the course of his report of the experiments and the conclusions arrived at Mr. Barbour said: To determine whether the iron and manganese could be satisfactorily and economically Temoved from the Boulevard water an experimental plant has been operated during the past eight months. Different w’aters, containing these metals, require quite different treatments. In some cases, simple aeration and filtration through sand is sufficient, in others pre-treatment through beds of coke—operated either as contact beds or tricklers—is necessary, and in still others, chemicals must be added for successful results. The object of the experiments was, therefore, not only to determine whether the iron and manganese could be removed, but by what method and at what rates this could most economically be done. In other words, the experiments were intended to determine, on a practical scale, the basis of the design of an actual plant, and to make certain, before undertaking the construction of such a plant, that it would work successfully. The experimental plant consisted of a prefilter of gravel or coke and a final sand filter. The experimental work proved that the iron and manganese can be successfully removed from the Boulevard water at reasonable cost. The necessary method was found to require a coke prefilter not less than 8 feet deep, operated as a’contact bed at a rate of between GO and 70 million gallons per acre daily, and a sand filter, operated at a rate of 10 million gallons daily. By this process, the iron and manganese were each reduced to less than .12 part per million in the final effluent. It was also proved that a lesser depth of the prefilter, or too much aeration, or the operation of the prefilter as a trickier, was not adapted to the treatment of the local water. On the basis of these experimental results, a plant has been designed, with a net average daily capacity of 7,500,000 to 8.500,000 gallons. The estimated cost, including the low lift pumping apparatus and the necessary addition to the pumping station, is $180,000. The average operating expense, including interest, depreciation and maintenance of the plant, and the cost of extra pumping, is estimated to be $7.65 per million gallons. This is equivalent to about 14 cents per year per person, assuming a per capita consumption of 50 gallons daily, and at this cost, the Boulevard supply can be made in every way acceptable. The estimated cost of the extension of the well system downstream, immediately necessary, is $30,000, including land, wells and suction pipes. Turning to the consideration of the feasibility of filtering the Merrimac River water, it is estimated that the first cost of a plant at the Boulevard for the purification of this water would approximate $250,000, and that the average cost of filtering this water, including interest, depreciation and operating charges, would be $10.27 per million gallons, or $2.62 more than the cost of treating the ground water. When, however, the expense of maintaining the well system:—estimated at $4.40 per million gallons—is added to the cost of the ground water supply, it follows that the total is greater than that of filtering the Merrimac River by $1.78 per million gallons. Because of the greater safety of the ground water, the lesser color, the lower temperature in summer, and the popular prejudice against the use of a polluted river water, however well filtered, it is concluded that this small difference in the cost is not of moment, and that, so long as Lowell can continue to obtain an adequate supply from the ground at such reasonable cost as has been shown to be possible, this should be done. A study has also been made of the possibility of connecting with the Metropolitan supply at the Wachusett dam, and it has been found that, on the basis of the cheapest reasonable development, this connection would involve a first expenditure of not less than $1,750,000, and a final cost of not less than $60 per million gallons of water taken, or practically double tile cost of maintaining and purifying the present ground water supply, or of filtering the Merrimac River water. It is, therefore, finally concluded that the city should further develop the Boulevard system of ground water, and should immediately undertake the construction of works for the removal of the iron and manganese, which alone prevent this supply from being entirely satisfactory. The total estimated cost of the purification plant, and the extension of the well system, now necessary, is $210,000, which, ii at once made available, will permit the improved supply to be furnished in the early months of the coming year. The Cook Wells —so-called—located in the valley of River Meadow Brook, were driven in 1893. The obtainable supply proved to be less than 3.000.000 gallons daily during the summer months, and in 1894. additional wells, known as the Hydraulic W’ells, were driven at a point about one mile further up-stream. the combined yield proving to be somewhat less than 5,000.000 gallons daily. As a result of the condemnation of the Cook Wells, the Water Board further extended the Boulevard system. During the experimental work, the Hydraulic Wells were out of service, and the investigation has been confined to a study of the supply from the Cook system. The results, therefore, apply only to the treatment of the Cook Well water, in which the important factor is the high carbonic acid content, which is responsible for the corrosive action of this water on metals. The iron in this water is low—less than .10 parts per million—-but in the Hydraulic Well water, from .30 to .60 parts of iron per million were found during the years when this supply was in service. The treatment of the Hydraulic Well water necessary to render it satisfactory would therefore involve a plant for the removal of iron, whereas, in the case of the Cook Well supply, the removal of the carbonic acid is all that is necessary to render it acceptable. As this latter supply is capable of furnishing from 2,500,000 to 4,000,000 gallons daily, dependent on the time of year, it has an obvious value as an auxiliary to the Boulevard system. The investigation was planned to make possible a comparison of the lead found in the untreated Cook Well water after passing through 50 feet of ⅛-inch lead service pipe, with that found in the water after passing through similar coils of pipe when the carbonic acid had first been neutralized to various degrees by lime or removed by aeration. The 50-foot coils of lead pipe were used—one for the raw water, another for the lime-treated water, and the third for the aerated water. The effect of the period of contact on the amount of lead taken up by the water was determined by taking samples after a period of four hours’ standing in or passing through the pipe, and also after a period of 15 hours. The apparatus for the addition of lime consisted of a funnel-shaped saturator, 24 inches high and 16 inches in diameter at the top. The lime was slaked in a small receptacle at the side of the saturator and introduced as milk of lime by a pipe which led to the bottom of the funnel. Water was introduced from a calibrated nozzle into a pipe which extended vertically through the center of the funnel to the bottom, the supply of water then rising slowly through the milk of lime and overflowing at the top from a circular weir. The capacity of the saturator was such that when treating 5,000 gallons per day of water with one grain of lime per gallon, the time of passage of the lime water through the saturator was about one hour. The lime water varied in strength more or less, but on the average contained 1,120 parts per million of calcium oxide, or 65.5 grains per gallon of this oxide, these figures being based on determinations of the caustic alkalinity of the lime water as it issued from the saturator. The flow of lime water was maintained at as nearly a constant quantity as possible, and the amount of raw water was varied when it became necessary to change the quantity of lime added per gallon. The lime water was applied to the raw water without exposing the latter to the air, duplicating, as far as possible, the conditions under which, if this method of removing carbonic acid were adopted, the lime water would be discharged into the suction pipe of the pumps. Apparatus for the determination of hardness, alkalinity, dissolve oxygen, carbonic acid and iron was provided, and these analyses were made on the ground by Mr. Clifton L. Rice, chemist. The determinations of the lead were made by the State Board of Health, which has co-operated throughout in this investigation. In starting the investigation such portions of the raw. lime-treated and aerated water were passed through the lead pipes during the day that four hours were required between the time of entering and leaving these pipes. At a later date, as always in the case of the 15-hour exposure, the fourhour samples were collected from water which had been allowed to stand in.the pipes for this length of time. In the interim between the periods during which the water stood in the pipes, the water was passed through at a rate of about one gallon in five minutes, the surplus of the treated water being wasted through a bypass. In general, the scheme was intended to represent, as nearly as possible, the actual conditions of a house service pipe. The tests indicate that by a pressure of five pounds on the nozzle, the carbonic acid can be reduced from 45 parts per million in the raw water to about 3.3 parts per million in the aerated water, and that with this reduction, the effect of the water on lead pipes is so far lessened as to render the supply safe. In stating this conclusion, the fact that, in the 15-hour tests, the lead taken up somewhat exceeded .5 parts per million is not overlooked, because it is believed that the retention of the water for 15 hours in the pipe is too severe a test, and that the fourhour period of contact more nearly represents the conditions of actual use. The experimental work has proved that the Cook Well water can be made suitable for use by aeration at relatively small cost. When thus rendered safe, it can be, in the future, utilized, if required, as an auxiliary to the Boulevard system in providing for the summer peak in the consumption curve. The Boulevard supply has always been hygienically safe—a fact shown by the low typhoid death rate of the city. To make it entirely acceptable, it is only neessary to remove the iron and manganese, and to determine how this could most efficiently and economically be done, the experimental work, which is now to be discussed, was undertaken. A small test plant was erected at the Boulevard station. The required chemical analyses were made by Mr. Rice, chemist in charge, and also by the State Board of Health, which has co-operated throughout the work. The experiments continued with such regularity as the operation of the pump at the Boulevard station permitted from September, 1913, to April of the present year. Water was supplied by a small pump connected with the force main to the overflow box, which maintained a constant head on the orifice, supplying the prefilter. After passing through the prefilter, the water passed directly to a sand filter, equipped with effluent rate controller. By meters and by daily calibration, the rate of filtration per acre was maintained as desired. The experiments indicated that the iron and manganese in the Boulevard water cannot be removed by aeration and direct application to sand filters, nor by aeration, sedimentation and sand filters. That excessive aeration and the operation of the prefilter as a trickier is not adapted to the treatment of the local water. That the iron and manganese can be successfully and economically removed by limited aeration, passage through a coke prefilter not less than 8 feet in depth, operated as a contact bed at a rate of 67.5 million gallons per acre daily, and subsequent filtration through sand at a rate of 10 million gallons per acre daily. On these conclusions, plans of a purification plant have been prepared, based on the use of coke prefilters 10 feet deep, an intermediate sedimentation basin of 1 hour’s capacity, and sand filters 3.0 feet deep. The depths of both filters are greater than in the experimental apparatus, and the addition of the in’ermediate sedimentation basin will serve to reduce the work to be done by the sand filter. Coincidently, therefore, with the experimental work in the treatment of this water, an investigation of the condition of the present wells and of the possibility of developing additional well fields in the Merrimac valley was undertaken. The Boulevard supply is dependent in a large measure on infiltration from the river. The amount entering the well field from the river at any point depends — other conditions being equal — on the difference in level between the water in the river and the water tablet at such point. The character of the water which is thus drawn from the river—its deoxidation by accumulated organic matter, and its consequent tendency to throw into solution the iron and manganese deposited in the soils—is largely determined by the rate at which the water passes through the material intervening between the river and the wells. Observations of th.c level of the water table in the well field have shown that only in a limited area, dependent on the location of these wells which are the newest and least clogged, was the ground water much depressed. In other w’ords, the difference in level or the head inducing infiltration from the river was effective only in a relatively small portion of the field, and without question, a greater quantity can be obtained by so maintaining and operating the wells as to depress the water tabic more uniformly over the entire well field. The present Boulevard system comprises some 450 wells, connected to a number of suction lines, which converse with the main pipe leading to the air chamber at the station. In connection with the question of future available quantity of ground water, reference may be made at this point to the possible utilization of the known supplies in the valley of River Meadow Brook. From the Cook and Hydraulic Wells, together, about four million gallons per day can be obtained. As already stated, aeration alone is sufficient to render the first-mentioned supply safe, but treatment to remove iron is necessary in the case of the hydraulic supply. Studies have shown that at the present time it is cheaper to develop an additional quantity of ground water at the Boulevard than to attempt the purification of the combined River Meadow Brook supplies, either by an independent plant, or by pumping this water to the Boulevard works. The Cook and hydraulic wells, however, constitute a valuable reserve for future use. The proposed purification plant is located on the north side of the Boulevard, immediately opposite tlse present lower pumping station. It consists of 0 coke prefilters, 10 feet in depth and twofifths of an acre in total area; a settling basin, divided into 2 units, with a total capacity of 500,000 gallons; 6 sand filters, with a total area of 1 acre; and a filtered water reservoir of 1,000,00o gallons’ capacity. Over the inner walls of the prefilters, a superstructure is to be built, so as to provide an operating gallery under cover. Immediately in front of this building, and connected thereto, a headhouse and laboratory is located, and practically all the operations involved in the control of the plant are concentrated in this central structure, which contains the main valves and the recording apparatus. The present Holly pump will be used to draw water from the wells and lift it to the prefilters. After passing through the coke, the water will flow’ through orifice boxes, by which the prefilter rate will be determined and recorded, thence through the settling basins to the sand filters, front which it will return through effluent pipes and Venturi meters to the headhouse, and finally flow’ through a conduit in the central wall of the settling basins to the filtered wafer reservoir, from which it will be lifted by new steam turbine driven centrifugal pumps, located in a small extension of the present station, into the force main leading to the West Sixth street station. At a rate of 75 million gallons per acre per day through the prefilters, and a 10 million gallon rate through the sand filters, the areas provided are equal to a 10 million gallon daily output. Allowing for cleaning and for the possible desirability of a lower rate through the coke, the plant is believed to be ample for an average daily supply of 7,500,000 to 8,500,000 gallons, or—if the past growth of the population in Lowell holds in the future—sufficient for the needs of the city until 1935. In estimating the capacity of the proposed plant on the basis of the experimental results, it is to be noted that the depth of coke is 40 per cent, greater and that of the sand 50 per cent, greater than the corresponding dimensions of the testing apparatus, and that there is interposed between the prefilters and sand filters a settling basin which will materially reduce the load on the final filters. Provision has been made for great elasticity in the plant, so that the method of operation can be adapted to any variable conditions which may develop in the water. All structures are designed to he of the most permanent character; the sand filters, settling basins and filtered water reservoir will be roofed by concrete groined arches and covered with soil and grass. The superstructures will be of the same material as that of which the lower Boulevard station is built. The appearance of the finished plant is indicated by the sketch included in this report. The intent has been to show the lowest probable cost of obtaining water from the Metropolitan supply, and on that basis, it appears that the cost will not be less than $00.00 per million gallons, or practically double the expense of maintaining and purifying the present ground water supplies. Aside from the question of relative expense, the Metropolitan water would not be so acceptable as the present ground waters after removal of iron and manganese, because of greater color and high temperature in summer. It does not, therefore, appear that there is any reason for further consideration of the possibility of obtaining water from the Metropolitan supply.