Environmental Engineering Reference
In-Depth Information
Two main outputs of the study have been used to support the DST: (a) a set of measures for
quick assessment of the network reliability using the performance indicators under regular
operation, and (b) an algorithm that incorporates the developed method for reliability
assessment and full diagnostics of single or multiple network layouts. The computer
programme has been developed with the help of EPANET Toolkit Functions available with
the main programme, also distributed by US Environmental Protection Agency (EPA).
Owing to its robustness, simplicity and free distribution via the Internet, EPANET software
has become very popular amongst the researchers and practitioners worldwide.
The first phase of the study upon the completion of the literature survey has focused to the
development of emitter based pressure-demand driven algorithm needed for calculation of the
available demand fraction. Furthermore, a choice has been made to analyse the patterns by
drawing statistical correlations on larger samples of networks. A network generation tool has
therefore been necessary in the second step, in order to create consistent samples, yet
sufficiently variable to arrive at solid conclusions about the patterns. This tool, based on the
principles of graph theory, has been developed for arbitrary spatial distribution of (demand)
nodes, using their coordinates in the generation process.
To provide reference layouts in the pattern analysis, the network generation tool has been
further expanded by GA-optimiser, building in the feature of the least-cost diameter
optimisation. In parallel, a number of indices has been developed that should quantify
network shape, its buffer and resilience. All generated networks, optimised or not, have been
further processed using the developed network diagnostics tool, which includes all proposed
reliability measures combined with a few others from the literature. Next to the reliability
assessment, the tool has been intended for assessment of network geometry, hydraulic
performance and annual costs of the assets, including operation and maintenance. To be able
to do it, the diagnostics tool has been designed to allow gradual or random increase/reduction
of the most typical network parameters to be able to monitor their possible correlation with
the reliability measures. With this, sufficient volume of the results of the network diagnostics
tool could have been statistically correlated.
In the final stage, the DST has been completed by adding the network filtering and
initialisation modules, allowing selection of networks that fit pre-selected routes of the pipes.
Hence, an integrated tool has been created that can generate, filter, optimise, and diagnose
large number of layouts in relatively short period of time.
Nearly 20,000 network layouts have been processed using the tool, varying between three
and 5044 pipes. The largest network analysed has been a real-case network supplying part of
Amsterdam. Additional simulation runs have been conducted on several samples, by altering
particular parameters of some layouts, namely the nodal demand multiplier and pipe
diameters.
Search WWH ::
Custom Search