or can be foreseen what kind of manufacturing will be undertaken on a given territory, provision for its waste water can be made in the sewers. Otherwise, possibly the only practicable plan is to provide such capacity as would be called for by residence occupancy; then, if it develops in the future. that this capacity does not suffice for manufacturing wastes, either the industries must themselves be required to provide for their own waste water, or a supplementary sewer or system of sewers can then be provided.

ART. 8. AMOUNT OF SEEPAGE

If the soil is continuously or occasionally wet at the depth to which the sewer is to be laid (and this includes a larger proportion of localities than most persons realize, since, in most towns, previous to constructing the sewer system, few excavations of any kind have been carried to this depth), an allowance must be made for ground water entering the sewer through its joints or walls. Sewers can be made absolutely tight by using iron pipe, and practically so by the use of special jointing material and special pains in the laying of vitrified or cement pipe; but the majority of sewers are laid with the ordinary Portland cement joint; which, as ordinarily made, is more or less porous. The use of iron pipe or special jointing material adds to the cost, but it is coming to be appreciated that, where ground water is anticipated, their use results in ultimate economy. Where Portland cement joints are used in wet ground the infiltration may with care be kept down to 1 cubic foot per second for each 50 to 100 miles of sewer; but under extreme conditions the seepage has been known to more than equal the entire capacity of the sewer. (The average leakage of 137 miles of 8-inch to 36-inch sewer in Boston was found to be .06 cubic foot per second per mile, there being double this seepage in the spring.) Some engineers have estimated seepage as a certain percentage of the domestic sewage, but there seems to be no logical excuse for this, but the seepage would seem naturally to vary with the number of joints and to a certain

extent with the circumference of the pipe. For this reason, 3-foot lengths of pipe should give but two-thirds as much seepage as 2-foot lengths (assuming sufficient ground water to supply all that the leaks will take). In some cases large amounts of ground water have entered the sewer through branches left for house connections which were not sealed against the entrance of ground water.

In general, the allowance to be made for seepage should be governed by decision as to the methods and material decided upon for making the sewer joints and the probability of the ground water plane being above the level of the sewer.

ART. 9. MAXIMUM TOTAL RATES

The total amount of sewage flowing in a separate sewer is the sum of the domestic, commercial, and industrial sewage and ground water seepage. The last is continuous and ordinarily varies slowly with the rise and fall of the ground water. Domestic sewage reaches a minimum between twelve and three o'clock at night (depending upon the local habits of the people), and remains very small, except for the water escaping through leaking house fixtures, until about five or six o'clock in the morning, when it rises rapidly and continues high for about twelve hours, then falls at a more or less uniform rate until a minimum is again reached. Commercial and industrial sewage is ordinarily zero from closing-down time in the evening until the beginning of work the next morning, except for leaking fixtures and unless night work is carried on. During the day the amount of this sewage flowing in the trunk sewers may be nearly uniform or may be subject to sudden considerable fluctuations at one or two or more periods, depending upon the nature of the manufacturing process.

The sewer must, of course, be of ample capacity to carry the maximum rate of flow due to any or all of these four classes of sewage that reach it. Concerning commercial and domestic sewage, it is a common plan to assume that the maximum rate during any one day is 75 per cent greater than the average.

As to the manufacturing waste waters, the same allowance should be sufficient for them also, and if it is found that discharge of an entire daily contribution in one or two flushes is practiced and is objectionable, it may be made compulsory that this discharge be continued throughout the day by providing storage tanks or otherwise.

We would therefore have the maximum capacity of sewer to be provided equal to 175 per cent of the maximum daily rate of combined domestic and industrial sewage, together with the average daily rate of ground water infiltration. As a further safeguard, a factor of safety is ordinarily employed of 2 for the smallest sewers and of 1 for the larger sewers; while for the trunk sewers the factor may be reduced to 11. That is, the actual capacity when running full is made from 2 to 1 times the amount calculated by the above assumptions and estimates. This allows for errors in assumptions as to future growth in population, use made of given districts, etc. Estimates of small districts are more liable to error than is that of the large district that embraces all of them, since in the latter the errors may balance, or that of one small district will be a smaller percentage of the total sewage flow than of its own flow only.

CHAPTER III

AMOUNT OF STORM SEWAGE

ART. 10. RATES OF RAINFALL

THE amount of storm-water reaching a given sewer depends upon the rate of rainfall, the time during which this rate is continued, the proportion of the rainfall which flows off, and the time taken by a raindrop after falling to reach the point in the sewer system under consideration. This last depends on the shape, extent, and nature of the surface over which this raindrop must flow and the length and grade of the sewer through which it must pass after having entered it. As in the case of separate sewers, there may be more or less infiltration of ground water into storm sewers, but this is generally so small in amount compared to the maximum amount of storm-water to be handled that it may be disregarded.

It is apparent that the rate at which water falling as rain reaches the sewer depends to a greater or less degree on the rates of rainfall from minute to minute and not upon the total amount that flows during a day or even an hour. Records giving the amount of rainfall per day are therefore valueless to the sewerage engineer, but he needs to know the maximum rate for shorter intervals, such as five or ten minutes. Gauges are in use which automatically register the rate of rainfall continuously or at intervals of about five minutes, and such gauges are now found in most of the large cities, operated either by a city department or by the United States Weather Bureau.

Intelligent designing of a sewerage system for a given city requires a knowledge of the maximum rainfall rates to be expected in the city in question for intervals of not more than

35

five or ten minutes. As will be explained later, what is needed is not simply the maximum rate for any five-minute period, but a record of a considerable number of consecutive fiveminute periods both before and after the occurrence of a maximum or very high rate. The engineer should obtain all available records of his own locality from the United States Weather Bureau or other reliable source; and in addition should study all similar records available that apply to the section of country in which lies the city in question and where the conditions affecting intensity of rainfall are similar. Even then there may occur at any time in the future a rainfall greater than any to be found recorded; but the courts have held and common sense affirms that neither the engineer nor the city can be held responsible either legally or morally for not anticipating such an event.

A common method of utilizing information collected as suggested above is to plot on cross-section paper the several maximum rates recorded for periods of 5, 10, 15, etc., minutes up to two or three hours; the rates being used as ordinates and the lengths of periods as abscissas; then draw a more or less regular curve which shall as nearly as possible pass through the maximum points. It may then be assumed that this curve will represent the maximum intensity of rainfall for which the engineer is justified in designing the sewer.

A city may not be justified in spending the money required for constructing a sewer capable of carrying the run-off from the greatest storm of which there is record, as a considerable percentage of this sum could be saved by providing the less capacity which would suffice for all storms except those occasional cloudbursts which may occur at intervals of ten or twenty years. It has been calculated in some instances that the amount of money so saved, if placed at interest, would provide sufficient funds to pay for all damage done by the very infrequent storms of excessive intensity which such a sewer would not remove promptly. On the other hand, the loss resulting from flooding of streets and cellars by water rendered more or less foul by the dirt washed from the street