rial is dropped to the bottom of the receiver and discharged through a rotary lock into the screw conveyor, which distributes the material across the top of the storage bins. The exhaust air is passed through the filters for the removal of the dust in suspension.

The principal advantages of this system of handling chemicals are: (1) elimination of the most laborious and disagreeable job about a water softening plant; (2) elimination of dust, thereby making it easy to keep the plant clean; (3) saving effected by purchasing material in bulk.

Objection has been raised to this system for unloading lime, on the ground that the lime is brought into contact with a large volume of air and consequently there may be some carbonation. Carbonation of lime does not occur at all in perfectly dry air,* and it is doubtful whether the short exposure the lime has to air, even if a small amount of moisture were present, would result in much, if any, carbonation. Charles Spaulding† Chemist in Charge of the Springfield, Ill., plant, states that he finds no difference in the analysis of the lime at his plant on samples taken before and after unloading from cars by means of the pneumatic

a true solution, it could be more readily applied in uniform quantities than milk of lime and also the feasibility of thoroughly mixing raw water with milk of lime in large quantities was questioned. It was thought that the raw water and the milk of lime would not be adequately mixed, resulting in a marked loss in softening efficiency, owing to the fact that the suspended particles of lime (calcium oxide) would become coated with precipitated carbonate of calcium and magnesium hydrate, thereby becoming inactive. It was also customary to add lime first and then after a definite period of time to add soda ash. This was done on the theory that the lime would combine with the soda-ash to form sodium hydroxide, which would absorb free and half-bound carbon dioxide, resulting in a wastage of soda-ash for doing work that should be done by lime, a less expensive reagent.

This theory was wrong, because the absorption of carbon dioxide by the sodium hydroxide results in the formation of sodium carbonate, and it therefore apparently makes no difference whether the lime and soda-ash are added separately or together. In many plants using solution feeds, they are mixed together in the same tank and fed out as one solution.

Dry-feed machines for measuring the chemicals and these same machines equipped with continuous lime slakers have been developed, and their introduction permits the elimination of chemical solution tanks with all their attendant difficulties, familiar to all water-works operators.

Figures 2 and 3 show the method used for weighing and slaking lime at the water-softening plant at Piqua, Ohio. The chemicals drop from the bins through under-cut gates to the weigh hopper. From the weigh hopper, the material flows into a Browning feeder. Figure 3 shows the Browning feeder, which consists of a hopper, a cylinder provided with a piston, and means for adjusting the length of

stroke of the piston. The piston is operated by a crankpin on face of gear driven by the main drive-shaft of the lime slaker. This type of feeder has proved to be very accurate, but with the Piqua chemical feeding equip ment it is not necessary to depend entirely on the accuracy of the feeder, because the operator can see at a glance, by looking at the weigh-hopper scale, just how much lime has been fed during any interval of time. For instance at Piqua, approximately 250 pounds of lime are required per hour when softening at the rate of 5 million gallons in 24 hours. The operator reads the lime weigh scale each hour and if the feeder has not fed exactly 250

They are shown in Figure 4. Mechanical agitators are shown in Figures 5 and 6. Figure 5 shows the type of agitating device designed at the Sacramento, Calif., water-filtration plant under the direction of Charles G. Hyde. This type tank has been used at the water-softening plants at South Pittsburgh, Pa.; Miami, Fla.; Springfield, Ill.; Hinsdale, Ill., and Piqua, Ohio.

The paddles are swung from the paddle arms by means of plate clamps, the paddles thus being allowed to assume a more or less horizontal position when the agitating device is started, but the weight of the paddles is such that, when motion in the tanks is obtained,

pounds of lime, it is adjusted. This results in very accurate application of the chemicals to the water, which is one of the most important features of the whole process of water softening.

Mixing Tanks or Agitating Devices Two types of mixing or agitating devices. have been used at water-softening plants: (1) baffled tanks with various combinations; (2) mechanical agitators. The over-and-under type baffled tanks have been successfully used at the Columbus plant for the past twenty years.

the paddles assume a vertical position. This is to take the load off of the motor in starting up the agitators. The relation of the area of the paddles to the area of the central vertical section of the tank is 25 per cent. Even with this large paddle area, uniform velocities are not maintained.

There seem to be two zones-an inner and an outer one. The inner zone is from the center to about one-third out, and the outer zone from there to the walls of the tank. The velocity is too low in the inner zone, and scarcely any floc can be seen, whereas in the

outer zone very satisfactory floc formation is obtained. See Figure 7.

A modification of the type described is shown in Figure 6. Instead of discharging the water from the side of the tank, it is discharged through a take-off pipe from the center. It is believed that by this modification there will be no short-circuiting of the water through the tank and more uniform velocities will be obtained.

The baffled tanks as built at Columbus are satisfactory and none of the mechanical tanks already built give any better results, but mechanical tanks are less expensive to build and they remove the necessity of lifting the water the extra two or three feet necessary to pass it through baffled mixing tanks; and, although agitators require power to keep them running, the cost of operation, if proper drive is provided, should be less than the cost of the additional lift of the water. They also provide greater accessibility and flexibility.

Dorr Clarifiers

The disposal of the large quantities of sludge

formed by the softening reactions was a problem in the past, especially when a water-softening plant was built along a small stream.

It was the custom, and still is at a few of the older plants, to allow the sludge to accumulate in the settling-basins and to discharge it from the basins intermittently. At the Columbus plant the sludge is held for a period of thirty days, and one of the disadvantages of doing this is that on each successive day the capacity of the tank decreases. Another disadvantage is that when the sludge is discharged in such large quantities into a stream, especially one as small as the Scioto River, it causes the stream to become very unsightly.

Apparatus has been developed which may be installed in the settling-basins or clarifiers, making it possible to remove the sludge produced from the water-softening reactions continuously. Practically all the municipal watersoftening plants built within the last five years have been provided with Dorr clarifiers for removing sludge continuously. Figure 8 shows the square type clarifier, and Figure 9 shows a round clarifier. There is a difference of

opinion as to the relative merits of the round and the square tank, and there is also a great deal of discussion as to the best method of flowing the mixed water from the mixing-tanks into the clarifier.

Methods for Further Reducing Hardness

In the early days of water softening, limited results were obtained in reduction of hardness on account of the formation of soluble complex basic carbonates of magnesium, which may be represented by the following formulae:

Mg(OH)2(CO3)3

MgCO3. NaHCO3. 4H2O MgCO2. Na2CO3

These soluble salts cause the alkalinity of the softened water to be high, especially when soda-ash is used with the lime to reduce noncarbonate hardness, and consequently water that is high in non-carbonate hardness is more difficult to soften than water with an equivalent amount of hardness but low in non-carbonate hardness.

Figure 10 shows, by curves, the change of the alkalinity and non-carbonate hardness of a magnesium water when treated by the lime soda-ash method. The non-carbonate hardness

On

should decrease along the broken diagonal theoretical line; instead, the non-carbonate hardness decreases along the circled line which remains above the broken theoretical line, thus showing that the theoretical calculated amount of non-carbonate hardness is not removed. The alkalinity should remain constant and follow the broken horizontal theoretical line. the addition of soda-ash, instead, the alkalinity rises along the circled line above the theoretical line. This behavior is characteristic of magnesium waters, regardless of whether the magnesium is in the form of sulfate, chloride or other radical. It is believed that the portion of soda-ash not going to remove the noncarbonate reacts with the magnesium salts in the water, forming complex basic salts.

The limitations in the reduction of hardness have been largely overcome by (1) the hot process, (2) excess lime treatment, (3) split treatment, (4) excess lime followed by carbonation, (5) the use of compounds of alumina, and (6) substitution of zeolite for soda ash to remove non-carbonate hardness. These processes are explained in detail by the author in Industrial and Engineering Chemistry, May, 1927, Vol. 19, No. 5, and only the last one will be discussed in this paper.

$150,000. The interest on this investment at 5 per cent would amount to $7,500 the first year, but would be less each succeeding year. Retirement charges for five-year bonds would amount to $30,000 per year, and it is estimated that the operation cost would be $75,000, making a total cost of $112,500 per year. This amounts to less than is now being spent annually for soda-ash, and if the zeolite plant is installed, it should pay for itself in five years. These cost figures are tentative esti

CURVES SHOWING THE RESIDUAL HARDNESS IN SCIOTO RIVER WATER AT COLUMBUS, OHO WHEN SOFTENED, IN BOTTLE EXPERIMENTS, BY THE LIME SODA ASH PROCESS AND WITH A RESIDUAL MAGNESIUM CONTENT OF 26 PARTS PER MILLION

FIGURE 9. ROUND TYPE DORR CLARIFIER FOR REMOVING PRECIPITATED SLUDGE FROM SOFTENED WATER CONTINUOUSLY

Substitution of Zeolite for Soda-Ash to Remove Non-Carbonate Hardness

Where salt may be obtained at a fair cost. it is cheaper to remove non-carbonate hardness by means of zeolite than by means of sodaash. At Columbus, the discontinuance of the use of soda-ash and the substitution of zeolite equipment is being seriously considered. If it is adopted, the entire supply of water treated will be softened with lime, coagulated with alum, settled, and carbonated; then a sufficient quantity of it will be softened to zero hardness by zeolite, so that by mixing the zero water with the lime-softened water the resultant mixture will have the desired degree of hardness. At the Columbus plant $115,000 is spent annually for soda-ash. It is believed that a plant for zeolite treatment can be built for

FIGURE 10. CURVES SHOWING THE RESIDUAL HARDNESS IN SCIOTO RIVER WATER AT COLUMBUS, OHIO, WHEN SOFENED IN BOTTLE EXPERIMENTS AS INDICATED

mates and must be carefully checked before a decision to change the method of treatment is made.

Lime-softened Water Not Stable Lime-softened water is not stable, because it is super-saturated with normal carbonates of calcium and magnesium, and distribution sys