Smart Meter Analytics to Pinpoint Opportunities for Reducing Household Water Use

Abstract

Knowledge of when, how, and by whom water is being used is crucial for planning ways to conserve drinking water. The goal of this paper is to identify groups of similar households (whom) based on their regular high-magnitude behaviors (RHMBs) of water consumption (when and how). RHMBs are frequent recurrences of high water use with regular timing. Household RHMBs are promising targets for behavior change. A two-stage data analytics approach is proposed. First, smart meter data is analyzed to identify RHMBs automatically. Second, salient features of the RHMBs are used to group households with similar behaviors. The approach is evaluated on two contrasting towns from low-rainfall regions of Australia. RHMBs accounted for 2 to 10 times more water than the traditional water efficiency target of continuous flows. For one group of 220 households, 60% of peak-hour demand was RHMBs. This paper demonstrates how RHMBs can be used to pinpoint opportunities for tailored demand management. Targets for substantial reductions in water consumption and supply costs are identified.