In the case of storm sewers, the only change of conditions affecting the volume of sewage which is likely to occur is in the imperviousness of the contributing area. If this is taken at the maximum, as for a business district, no allowance need be made. In any case, the allowance for change can best be made in the selection of the factor of imperviousness and the sewer built of corresponding capacity. It is probable that no condition of size or character of tributary area will in actual practice call for a storm sewer of a diameter less than 10 or 12 inches, and 15 or 18 inches is used as a minimum by many engineers.

A 10-inch sewer flowing full at a velocity of 3 feet per second will carry 100 cubic feet per minute, or the run-off from a rainfall at a rate of 2 inches per hour on a totally impervious area of 0.8 acre, or a residential area of double that, or say 200 by 350 feet. A 12-inch sewer with the same velocity of flow would carry the same run-off from an area 50 per cent larger, or from the same area at a 3-inch rate. Consequently these sizes are no more than sufficient for inlet connections or a sewer fed through but one inlet. Moreover, the water, when it enters the sewer, quite generally has a velocity of less than 3 feet per second, acquiring this velocity only after flowing some distance. Methods of overcoming this will be discussed in the next article.

A circular or egg-shaped sewer is sometimes limited in size by the amount of covering necessary and the distance below the street-surface of its invert, where this is fixed by the elevation of the outlet and the necessary grade from that to the point in question. If the whole sewer at this point be lowered, the grade and velocity become less and the size of the sewer must be increased, thus raising the crown. The size can be reduced only by increasing the grade, which means raising the sewer. Under these conditions, the sewer can be built as an "inverted siphon" to flow under a head (Art. 22), two or more parallel smaller sewers can be substituted for the one, or the shape can be modified. In adopting the last alternative engineers have devised many forms which can be generally clas

sified as those flattened on the bottom and those flattened at

the top.

ART. 15. SHAPES OF SEWERS

Of all possible shapes of sewers of equal area of cross-section, the circular gives the greatest velocity when flowing full or half full and, having the shortest perimeter, contains the least material. Also, being devoid of angles, it offers little opportunity for deposits. For sewers intended to always flow at least half full it is, therefore, the most desirable shape from this point of view. This is not true, however, of a combined sewer—that is, one which carries both house sewage and storm water-for, as we have seen (Art. 13), the house sewage may occupy only of the capacity of the sewer and have a velocity only about as great if a circular sewer be used. If the sewer, considered as a storm sewer, be given a grade adapted to a velocity of 4 feet per second when flowing full or half full, the velocity of the house sewage would be about of a foot per second. If on the other hand the grade be so increased (which is seldom possible) as to give the minimum house sewage flow a velocity of 1 feet per second, the depth of this flow would be only about .02 of the sewer diameter. Neither of these conditions is permissible in a good sewerage system.

The result of adopting too flat a grade is shown by the illustration (Fig. 11) of obstructions in the old London sewers which came to be known as "sewers of deposit." These required frequent cleaning, since almost the entire sewage matter was deposited in them, and became very dangerous to the health of the city. The question thus forced upon the attention of engineers was first solved by building in the bottoms of the old sewers channels of much shorter radius of curvature. These, by increasing R and consequently V, as well as the depth of flow relative to the invert radius, had the same effect upon the flow as the use of smaller sewers, which they in fact were, and answered the purpose, practically the same design being employed in recent years in Washington, D. C., and other

American cities. It will be noticed, however, that there is considerable useless material in this design; also that the bench on either side of the small channel offers opportunity

FIG. 11. LONDon "Sewer of Deposit," and SAME WITH MODIFIED INVERT. for the deposit of material, which may putrefy there. To meet these objections the egg-shaped sewer was designed and is used extensively for combined, and sometimes for storm-water, sewers. Several proportions have been suggested and used,

FIG. 12.-WASHINGTON, D. C., EARLY STANDARDS. a-Egg-shaped. b-same with modified invert. c-pipe bedded in concrete, with concrete collar at joint.

but that most frequently found in modern American practice is represented here. The diameter of a circular sewer having an equal area is 1.209D. In this sewer

Reference to Table No. 14 shows that a flow of 16 of the full capacity of this sewer would have a velocity about 0.3 as great as if the sewer flowed full, or 85 per cent greater than the same amount in a circular sewer of equal total capacity; also the depth would be about o.1D, or 0.4r". If the velocity of the house sewage in the

above be 2 feet per second (as it should be), that when the sewer were full would be 8 feet or more per second. This form does not, therefore, quite meet the requirements of a combined sewer, intended to carry a run-off of 3 inches from the area drained, as to either depth or velocity of house sewage. As we shall see later, this requirement applies to lateral combined sewers only,

and this design is suitable for most combined sewer mains, whose maximum flow is only 1 or 2 inches run-off from the drainagearea. In laterals or other sewers, however, where the proportion of house to storm sewage will be too small, or for some other reason sufficient velocity and depth for the house sewage cannot be thus obtained, the adoption of an egg-shaped sewer with r" =D or D, or a form similar to that shown in Fig. 12, is recommended; the purpose being, whatever the form adopted, to get a satisfactorily high value for R for the house sewage flow. Whatever the radius of invert, the grade must. not be less than that which would be required by a circular house sewer having a radius=r". The radius r" should be so chosen, also, that the depth of house sewage will never be r". less than A flat bottom should never be used for house or combined sewers unless the sewage will always be sufficient

2

to cover it at least 6 inches deep.

Angles in the section are

In storm sewers it is ad

to be avoided as favoring deposits. visable that the shape be such as to give good velocity to small amounts of storm-water, but the penalty of not following this rule is not so serious as in the case of house sewers.

Economy of construction will frequently require the use of other designs for sewers, as when built in tunnel, on special foundations, etc. These will be considered under the head of "Designing."

When entering a sewer from a storm-water inlet or a houseconnection, the sewage too commonly does not have a smooth flow parallel to the axis of the sewer and of the desired velocity, but must flow for some distance in the sewer before this is attained. Much can be done to improve this condition by proper designing of intakes and junctions. The latter should be so designed that the sewage enters from lateral or connection to main sewer in a direction as nearly parallel to the axis of the main as possible. Inlet connections can be greatly increased in capacity if the upper or intake ends be given a flaring or bell-mouth shape, and if the inlet itself be so designed as to guide the street run-off into it with the least possible amount of agitation. Among other things this means the avoidance of all angles in the structure. In general, it may be said that angles should be eliminated altogether in all parts of a sewer system reached by flowing sewage unless it is desired to retard velocity.