Environmental Engineering Reference
In-Depth Information
The latter aspect is important from the perspective of water quality in distribution networks,
which may deteriorate due to stagnation of water. In case of reliable and affordable electricity
supply, the practice of water distribution in some West European countries suggest that the
(large) reservoirs are not economic due to expensive cleaning and/or expensive land use
where they are to be constructed. That logic maybe completely different in developing
countries, where the supply of electricity is much less reliable and the problem of reservoir
maintenance can be mitigated by cheap labour. Still, the (extreme) topographic conditions
will be prevailing while making the final decisions.
As far the pipe capacity, much of spare capacity in urban water distribution networks is often
resulting from (sometimes overdesigned) hydrant capacity. That in itself produces more
reliable networks but also the situation where more frequent pipe cleaning is necessary. The
network of Amsterdam North discussed in Chapter 5 is a typical case. Last but not least,
network designers easily add comfortable safety factors; a slight (linear) increase in pipe
costs will result in large (exponential) increase of conveying capacity.
8.6
HYDRAULIC SIMULATIONS AND RELIABILITY ASSESSMENT
To find out how the parameters described in the previous paragraphs interact with network
reliability, a series of hydraulic and reliability simulations has been conducted on two simple
networks: one with direct pumping supply (A), and the other where pumping is combined
with balancing storage (B), which actually is an adaptation of the A-layout. Both networks
are shown in Figure 8.6. Depending on the terrain configuration, the tank connected in the
figure to node N10 , can also be connected to nodes N4 or N6 .
Figure 8.6 Network layouts A (left) and B (right)
The basic information about the network geometry and hydraulic performance has been given
in Tables 8.2 and 8.3, respectively. The total network lengths are similar: 8.16 km ( A ) and
8.31 km ( B ), and all pipes have the uniform k-value of 0.5 mm. As the nodes are not densely
connected, increased network resistance has been generated by selecting relatively large pipe
lengths. Furthermore, the EPANET pump curve has been defined by single pair of duty flow
and duty head points: 300 l/s at 85 mwc, in case of network A , and 255 l/s by 63 mwc, in case
of network B . In both cases, the efficiency curve has been calculated for the maximum
efficiency of 75% (achieved at the duty flow).
Search WWH ::




Custom Search