For example, let it be desired to determine the dimensions of two continuous-flow sedimentation basins as shown in Fig. 157, in which the period of retention in each is to be 2 hours, the velocity of flow is not to exceed one foot per second, and the sludge accumulated will be 3 cubic yards per million gallons of sewage treated. The quantity of sewage to be treated is 18,000,000 gallons per day. The shortest time between cleanings will be 2 weeks.

The capacity of each basin must be 2
24 of 18,000,000 gallons, or 200,000 cubic feet in order to allow a period of retention of 2 hours. To this volume should be added sufficient capacity to allow for the 2 weeks of sludge storage between cleanings. When a basin is being cleaned the load must be put on the remaining basins. Then if Q represents the rate of accumulation of sludge per day, n represents the number of days between cleanings, m represents the number of basins, and S the sludge capacity of one basin, then

The sludge storage capacity for the example given will be approximately 11,000 cubic feet.

In expressing the total cost of the basins let

It is now necessary to express the three variables b, l, and h, in terms of one of them. From the relation Q = 2blh it is possible to rewrite the expression for the total cost as:

Holding h constant and differentiating with regard to b in the first expression and with regard to l in the second expression, equating to zero and solving we get:

The economical relation between b and l is therefore

regardless of the value of h.

Substituting these values of l and b in the original expression for the total cost, it becomes

Differentiating with regard to h, equating to zero, and solving

In the example given if q = 0.2 and p = 1.0 then

Since these are reasonable dimensions and in accord with good practice they should be used, unless other conditions are unsuitable or the velocity of flow is too great. A width of channel of 120 feet as compared to a length of 160 feet is conducive to a poor distribution of velocity across the basin. A ratio of width to length of about 1:4 is desirable. In this case, by the use of three baffles parallel to the length of the basin, thus dividing it into channels 40 feet wide and 11.6 feet deep, the ratio of width to length is changed to 1:4 and the velocity will be increased only to 0.06 foot per second or 3.6 feet per minute, which is a reasonable velocity. It could be reduced by increasing the spacing of the baffles or the depth of the chamber.

Complicated baffling is undesirable. Two or three overflow baffles may be used to permit quiescent sedimentation in the space thus formed, and hanging baffles may be placed before the inlet and outlet to break up surface currents, or to prevent the movement of scum. The hanging baffles should not extend more than 12 to 18 inches below the water surface. The inlet and outlet are sometimes arranged to permit the reversal of flow, and the connecting channels between basins to allow the operation of any number of basins in series or in parallel, although such arrangements are more important in water purification. Sewage should enter and leave at the top of the basin.

Cleaning is facilitated by the location of a central gutter in the bottom of the basin with the slope of the bottom of the basin towards the gutter from 1:25 to 1:80 or steeper. A pipe, 2 inches or larger in diameter, containing water under pressure with connections for hose placed at frequent intervals is a useful adjunct in flushing the sludge from the sedimentation basins. For equal capacity, deep vertical flow tanks are more expensive and difficult to construct than the shallower rectangular type. Deep tanks are advantageous, however, in that sludge can sometimes be removed by gravity or by pumping without stopping the operation of the tank. They will also operate successfully with shorter periods of detention and higher velocities. The upward velocity should not be greater than the velocity of sedimentation of the smallest particle to be removed. The efficiency of sedimentation in them will be increased by the sedimentation of the larger particles which drag some of the smaller particles down with them. The Dortmund tank shown in Fig. 158 is an example of this type.

Ordinarily it is not necessary to roof sedimentation basins as the odors created are not strong, and difficulties with ice are seldom serious.

Chemical Precipitation

241. The Process.—Chemical precipitation consists in adding to the sewage such chemicals as will, by reaction with each other and the constituents of the sewage, produce a flocculent precipitate and thus hasten sedimentation. The advantages of this process over plain sedimentation are a more rapid and thorough removal of suspended matter. Its disadvantages include the accumulation of a large amount of sludge, the necessity for skilled attendance, and the expense of chemicals. The process is not in extensive use as the conditions under which the advantages outweigh the disadvantages are unusual. Sewage containing large quantities of substances which will react with a small amount of an added chemical to produce the required precipitate are the most favorable for this method of treatment.

Chemical precipitation accomplishes the same result as plain sedimentation, although the effluent from the chemically precipitated sewage may be of better quality than that from a plain sedimentation basin.

242. Chemicals.—Lime is practically the only chemical used for the precipitation of the solid matter in sewage. Commercial lime used for precipitation consists of calcium oxide (CaO), with large quantities of impurities. It should be stored in a dry place and protected from undue exposure to the air to prevent the formation of calcium carbonate (CaCO₃), the formation of which is commonly known as air slacking. The active work in the formation of the precipitate is performed by the lime (CaO) or calcium hydroxide (Ca(OH)₂). The lime should therefore be purchased on the basis of available CaO, which may be as low as 10 to 15 per cent in some commercial products. The amount of lime necessary depends on the quality of the sewage, the period of retention in the sedimentation basin, the method of application, the required results, and other less easily measured factors. Full scale tests for the amount of lime needed to produce certain results are the most satisfactory. In practice the amount of lime necessary when lime alone is used as a precipitant has been found to be about 15 grains per gallon. This may be markedly different, dependent on the quality of the sewage. For acid sewages, lime alone is not suitable as a precipitant since it is necessary to add sufficient lime to neutralize the sewage before the calcium carbonate will be precipitated.

The use of copperas (FeSO₄) together with lime, leads to economy in the use of chemicals as the flocculent precipitate of ferrous hydroxide (Fe(OH)₂) is more voluminous than the precipitate of calcium carbonate. This is commonly known as the lime and iron process. The presence of iron in certain trade wastes may reduce the cost of chemical precipitation, as the necessary amount of copperas is reduced. Where 15 grains of lime alone will be needed per gallon of sewage, the total amount of chemicals used will be reduced to 8 to 10 grains per gallon with the use of lime and iron. This combination is less expensive than the use of lime alone, and is even cheaper where the iron is already present in the sewage. Such a condition is well illustrated by the sewage at Worcester, Mass., where the oldest and best known chemical precipitation plant in the United States is located. The amount of lime used at this plant has varied between 6 and 10 grains per gallon of sewage, the normal amount being about 7 grains. No iron is added because of the amount already in solution.

The results of a series of experiments on the chemical precipitation of sewage by Allen Hazen, are given in the 1890 Report of the Massachusetts State Board of Health, on p. 737 of the volume on the Purification of Water and Sewage. Hazen concludes as the result of his experiments: concerning lime,

There is a certain definite amount of lime ... which gives as good or better results than either more or less. This amount is that which exactly suffices to form normal carbonates with all the carbonic acid of the sewage. This amount can be determined in a few minutes by simple titration.

Concerning lime and iron (copperas) he states:

Ordinary house sewage is not sufficiently alkaline to precipitate copperas, and a small amount of lime must be added to obtain good results. The quantity of lime required depends both upon the composition of the sewage and the amount of copperas used, and can be calculated from titration of the sewage. Very imperfect results are obtained from too little lime, and, when too much is used, the excess is wasted, the result being the same as with a smaller quantity.

In precipitation by ferric sulphate and crude alum, the addition of lime was found unnecessary, as ordinary sewage contains enough alkali to decompose these salts. Within reasonable limits the more of these precipitants used the better is the result, but with very large quantities the improvement does not compare with the increased cost.

Using equal values of different precipitants, applied under the most favorable conditions for each, upon the same sewage, the best results were obtained from ferric sulphate. Nearly as good results were obtained from copperas and lime used together, while lime and alum each gave somewhat inferior effluents.... When lime is used there is always so much lime left in solution that it is doubtful if its use would ever be found satisfactory except in case of an acid sewage.

It is quite impossible to obtain effluents by chemical precipitation which will compare in organic purity with those obtained by intermittent filtration through sand.

It is possible to remove from one-half to two-thirds of the organic matter by precipitation ... and it seems probable that ... a result may be obtained which will effectually prevent a public nuisance.

243. Preparation and Addition of Chemicals.—Lime is not readily soluble in water. Therefore, it is not best to add the lime as a powder to the sewage, but to form a milk of lime, that is, a supersaturated solution containing from 2,000 to 4,000 grains per gallon, although dry slaked lime has sometimes been applied directly. The solution is prepared in tanks in a quantity sufficient for some part of the day’s run, commonly sufficient to last through one shift of 8 or 10 hours. The lime is prepared by placing the amount necessary to fill one storage tank into a slaking tank containing some cold water. Sufficient water is added to keep the solution just at the boiling point, or steam may be added to make it boil. After slaking, it is run into the milk-of-lime solution tank and sufficient water added to bring to the proper strength. The milk of lime is added in measured quantities, being controlled by a variable head on a fixed orifice or weir, so that it may be varied with the amount of sewage flowing through the plant. The amount of lime to be added is determined by titration with phenolphthalein, experience indicating the color to be obtained when the proper amount of lime has been added.

The use of either copperas or alum has been so rare, for the precipitation of sewage, that a description of the methods of handling these chemicals as a sewage precipitant is not warranted. An excellent description of the methods of handling these chemicals in water purification will be found in “Water Purification” by Ellms.

244. Results.—The results of Hazen’s experiments indicate that a greater amount of suspended matter can be removed in the same time by chemical precipitation than by plain sedimentation. The percentage of removal of suspended matter may be as high as 80 to 90 per cent with a period of retention of 6 to 8 hours and the addition of a proper amount of chemical. That the method is not always a success is shown by the results of some tests at Canton, Ohio.^[150] The report states:

... lime treatment removes about 50 per cent of the suspended matter, and in the main about 50 per cent of the organic matter.... These data are instructive as indicating that the addition of lime to the Canton sewage in quantities as previously stated does not materially improve the character of the resulting effluent over and above that which could be produced by plain sedimentation alone.

The plant at Worcester, Mass., is the largest in the United States and information from it is of value. A summary of the results at Worcester for 1900, 1910, and 1920 are shown in Table 81.