Spectral analysis of residential water demand.

Residential water demand is one of the most uncertain variables within hydraulic analysis of water distribution networks. It is stochastic in nature and several methodologies have been previously developed to represent this variable (Buchberger and Wu, 1995; A/cocer et al, 2007). However, in these methodologies it is not possible to define accurately two situations: 1) the degree of information lost while sampling the water demand with a certain recording interval and the possible presence of aliasing, and 2) the range in time up to which time aggregation and disaggregation processes can be applied. It is shown in this paper that it is possible to apply Fourier's analysis as a time disaggregation tool and obtain the optimal recording interval. Using such methodology it is shown that the application range of the schema is between 240 seconds and 1 second. In other words, it is possible to measure with a recording interval of up to 240 seconds and retrieve information corresponding to a recording interval of 1 second. The method was validated with field measurements of residential water demand in a Mexican city.