Mathematical modelling of water quality in distribution systems

In the last few years many mathematical model have been developed for the evaluation of space and time variabilities of water quality parameter in water distribution systems. The differences among these models stem mainly from the flow conditions and phenomena consideredf and also the nature of the substances carried by the flow. Most of these models refer to a steady state condition and the transportation of substances in the pipes is usually considered to be due only to advection while molecular and eddy diffusion tend to be ignored. Furthermore, the substances contained in the water are divided into two classes; conservative and non conservative. For the former, variations in concentration are due exclusively to the mixing, in every node, with the water flows deriving from the pipes flowing into the node itself, whereas the latter are subject to reactions causing their transformation, to decay phenomena and to volatisation processes in the tanks located along the system. The present paper describes a model previously proposed and used by the authors (Puabe et al., 1995). This model differs from already existing ones because it is more complete, has shorter processing times and makes it possible to adopt physically-based reasoning criterua in order to follow the development of each single process within the network. In order to verify the results of the model, its application to an experimental system is presented and the results of numerical processing are compared with experimental measurements of the concentrations detected at the system nodes.