Classifying household water use into indoor and outdoor use from a rudimentary data set: a case study in Johannesburg, South Africa

Abstract

Distinguishing between indoor use and outdoor use is becoming increasingly important, especially in water-scarce regions, since outdoor use is typically targeted during water restrictions. Household water use is typically measured at a single water meter, and the resolution of the metered data is typically too coarse to employ on commercially available disaggregation software, such as flow trace analysis. This study is the first to classify end-use events from a rudimentary data set, into indoor use or outdoor use. This case study was conducted in Johannesburg, South Africa, and quantified the volume of water used indoors and outdoors at 63 residential properties over 217 days. A recently developed model for classifying water use events as either indoor or outdoor, based on rudimentary water meter data, was employed in this study. A total of 212,060 single end-use events were classified as being either indoor or outdoor. The indoor and outdoor consumptions were compared with survey results. It was found that 30% of all events were outdoor, based on the total volume. HIGHLIGHTS This case study was successful in classifying water use into indoor and outdoor water use events from coarse end-use data.; An average of 30% of the total water demand was classified as being outdoor use.; Classification tools implemented in this case study (PEET and WEAM) could be useful to monitor whether homes adhere to water restrictions, especially if outdoor use is limited or prohibited.;

Keywords

• end-use events
• low resolution data
• residential water demand
• water classification models