Planning rainwater harvesting systems in a context of uncertainty: from the building to the urban scale

Rainwater harvesting (RWH) can be a useful way of obtaining multiple benefits for the sustainable management of urban water. Among such benefits, the reduction of the consumption of potable water has been widely analyzed in the scientific literature, but the effects of different sources of uncertainty have often been neglected. The present work analyzes the water-saving efficiency of RWH for indoor use at the building and urban scales by assessing the variability of the performance estimations considering uncertainties on both the demand and supply sides. We evaluated the variability of water consumption on the demand side, with reference to both daily and multiyear scales, through Monte Carlo simulations that considered the statistical distribution of the indoor, nonpotable water demand. Diversion of the first flush, which is required to ensure acceptable water quality standards, is an important source of variability that has here been analyzed considering both the role of different setups of the first flush device and the volume of the first flush that is diverted. We evaluated the influence of the variability of rainfall patterns, considering different lengths and periods of the used rainfall series, to establish the length of the minimum series that can guarantee reliable results. The city of Turin (Italy) was considered as a case study for the proposed approach. The obtained results indicate which buildings are more sensitive to the considered variabilities and that the factor that impacts RWH efficiency the most is the average, long-range, nonpotable water demand, while daily variabilities showed no effects. Furthermore, the length and period of the rainfall series were found to drive the variability of the RWH efficiency on the supply side but, when the series is longer than 15 years, the variability related to the first flush amount and setup has an even greater impact.