Generating Water Demand Scenarios Using Scaling Laws

Abstract

This paper addresses uncertainty inherent to water demand and proposes an approach to generate demand scenarios and calculate their probability of occurrence. Nodal water demands are modelled as correlated stochastic variables. The parameters which characterize demand vary with spatial and temporal aggregation levels. Scaling laws allow the definition of these parameters for different users and sampling rates. Different scenarios are generated by considering different combinations of demands at each node of the network. A multivariate normal distribution is used to obtain the probability of each demand scenario. Correlation between demands is found to significantly affect the scenarios probabilities.