Investigating the impact of temporal and spatial lumping on rainwater tank system modelling

Due to the need to consider rainwater tanks as an option within urban water resources policy development and strategic planning within Australia, modelling tools are used to predict their yield and volumetric reliability. The accuracy of the estimate of these performance measures is important as they form the basis of the predicted potable supply reduction level. This paper investigates how temporal and spatial lumping impacts on the estimated yield and volumetric reliability of suburban rainwater tank systems. This question is of interest because the overall behaviour of a group of household rainwater tank systems, comprising a diverse mix of roof catchment areas, tanks storage capacities, and end use demand characteristics, is not identical to the behaviour of a single household rainwater tank system which has the mean characteristics of the group of rainwater tank systems. The analysis presented in this paper suggests that the spatial averaging of multiple rainwater tank characteristics systematically predicts greater yield and volumetric reliability. The magnitude of this variation is significant enough to warrant further attention. Using a daily time step (temporal lumping) and spill before yield, in combination with a depression storage depth of >0, produced larger estimates of yield and volumetric reliability relative to using a 6 minute time step. In comparison, using a daily time step, in combination with a depression storage depth of 0, produced insignificant differences in yield and volumetric reliability relative to using a 6 minute time step, except in the case of the very small tank system tested.