Experimental analysis of residential water demand data: probabilistic estimation of peak coefficients at small time scales

Residential water demand is the predominant fraction of the total amount of water distributed by municipal aqueducts, reaching and overcoming in general 70% of the total. The accurate description of demand, both in terms of mean values and spatial and temporal variability is then fundamental in a realistic modelling of water distribution networks. An experimental analysis of residential water demand carried out on 82 homogeneous households in the city of Latina (Italy) is presented. A first study is aimed at the characterisation of the mean features of the consumption and of its empirical variability. A second approach is devoted to the probabilistic estimation of the peak coefficients, in relation to different time scales and to different spatial aggregations. General expressions of the peak coefficients are obtained introducing suitable scaling laws for the variance of consumption.