|
|
|
|
|
|
|
flow conditions. Most of these are based on European or North American conditions but at least one (UNCHS, 1986) is based on research in Brazil. This argues for a minimum gradient of 1 in 167, based on the assumption of a flushing rate of 2.2 1/s. The latter is high for a pour-flush system and ignores the effect of flow attenuation. An important practical point which has emerged in several of the projects reviewed in this chapter is that house connections often enter manholes at some distance above the invert level of the sewer. This is a departure from conventional practice which requires that branches enter the main sewer with their soffits at or slightly above that of the main sewer, if necessary following a backdrop pipe, so that benching can be provided to guide flows and conserve momentum. It is likely to invalidate some of the assumptions which are made in the various theoretical attempts to arrive at a rational design philosophy. |
|
|
|
|
|
|
|
|
The gradients adopted in North-east Lahore were related to the number of houses connections and have been reported previously (Tayler, 1990). The aim was generally to achieve a gradient of about 1 in 120 at the head of a sewer with progressively flatter gradients allowed as the number of house connections increased. The minimum gradients adopted for tertiary sewers, receiving flows from up to about 100 houses, were about 1 in 180. In a more recent publication (Tayler and Cotton, 1993), slightly flatter gradients were suggested in accordance with Table 4.3. These figures are to some extent empirical. They start from the assumption that a gradient of 1 in 150 is acceptable for a sewer receiving flow from 10 houses and extrapolate between this assumption and a conventional approach to the design of a 230 mm diameter sewer running full (in the latter case, the assump- |
|
|
|
|
| | No. of houses | | | 10 | | | 20 | | | 40 | | | 60 | | | 100 | | | 200 | |
|
|
|