Author(s)

R. Magini, I. Pallavicini & R. Guercio

Abstract

The conventional modelling of distribution networks normally makes reference to a deterministic approach, not merely with regard to the supplies, but also with respect to the geometrical description and the hydraulic characteristics of the system. A different approach for the modelling of water supply systems derives from the explicit consideration of the random nature of the processes analysed, and by the reciprocal interaction and attenuation operated by the hydraulic system. The main objective of this paper is to achieve the implementation of a stochastic simulation model for the instantaneous demand able to preserve in space and time the characteristics of the physical peculiarities of the elementary demand. The results obtained from the processing of progressive spatial-temporal aggregations of the various elementary users are presented in order to highlight the non-linearity characteristics which distinguish the phenomena examined and facilitate the adoption of optimum observation and simulation scales. Keywords: residential water demand, stochastic modelling, spatial scaling. 1 Overview In the stochastic modelling of water demand, firstly proposed by Buchberger et al [1] [2], two phases can be distinguished: the first step consists of the estimation of the parameters of the stochastic process from the recorded demand series the second step is the use of the model as a demand simulator, as a stochastic process, for the types of users considered. In both cases is essential to understand if spatial scaling effects exist that can affect the correct modelling of phenomena.

Keywords

residential water demand, stochastic modelling, spatial scaling.