On the choice of the demand and hydraulic modeling approach to WDN real‐time simulation

This paper aims to analyse two demand modelling approaches, i.e. top-down deterministic (TDA) and bottom-up stochastic (BUA), with particular reference to their impact on the hydraulic modelling of water distribution networks (WDNs). In the applications, the hydraulic modelling is carried out through the extended period simulation (EPS) and unsteady flow modelling (UFM). Taking as benchmark the modelling conditions that are closest to the WDN's real operation (UFM+BUA), the analysis showed that the traditional use of EPS+TDA produces large pressure head and water discharge errors, which can be attenuated only when large temporal steps (up to 1 hour in the case-study) are used inside EPS. The use of EPS+BUA always yields better results. Indeed, EPS+BUA already gives a good approximation of the WDN's real operation when intermediate temporal steps (larger than 2 min in the case-study) are used for the simulation. The trade-off between consistency of results and computational burden makes EPS+BUA the most suitable tool for real-time WDN simulation, while benefitting from data acquired through smart meters for the parameterization of demand generation models.