Exploring drivers of residential water demand in regional Western Australia to inform water savings programs

ABSTRACT

Targeted water demand management interventions can be a cost-effective method for reducing consumption and deferring expensive alternative supply options. Determining where to focus efforts first requires an understanding of expected water demand in different regions. This article uses elastic net regression to examine the drivers of residential water demand in 68 towns in regional Western Australia. The model explained 84% of the variance within the training data, with climate and demographic variables having the strongest contribution. The model residuals, when given new data, can be considered the savings potential of that town. Key towns to target specific demand management interventions can then be evaluated by ranking residuals based on percentage and absolute metrics.