Classifying Household Water Use Events into Indoor and Outdoor Use: Improving the Benefits of Basic Smart Meter Data Sets

Abstract

This research investigated relationships between the most notable characteristics of end-use events, namely, event duration, volume, and intensity, in order to categorize water use as being indoor or outdoor. Three classification models were developed, calibrated, and compared using more than 200,000 household end-use events that were recorded independently in Australia and South Africa. The three methods were also compared to a practice-based limit classification scheme. The classification model presented in this paper correctly apportions ∼81% of the indoor end-use event volumes and ∼98% of the outdoor end-use event volumes, thus reinforcing the value of basic smart water meter data sets as a source of useful information for water demand management.