Integration of Behavioural Change Into Demand Forecasting and Water Efficiency Practices

This project was funded to build upon previous UKWIR research into customer behavior and demand. The research objective was to better understand how attitudes and behaviors and household water use are related, and whether customer behavior could be incorporated into demand forecasts. The research was conducted through an in-depth survey to approximately 1,000 UK residents, followed by metered household end-use using micro-component logging at 62 residences. The study found that stated customer attitudes could be categorized into five behavioral typologies from “disengaged” to “conscious consumers,” but that these typologies were not well correlated with metered consumption data. This finding is in line with published literature that self-reported attitudes and behaviors do not appear to be strongly related to household water use.

Using the survey and consumption data, the authors were able to explain about 50% of the variation in household demand using household factors including occupancy, property type, garden size, dishwasher ownership, etc. Predictive models were able to explain up to an additional 30% of the variation through frequency of use information from the micro-component measurements, but that the models were potentially over-fitted.

Further analysis of the consumption data at the sub-daily level showed that three different patterns of water use through the day could be defined by the time of the morning peak use: “early risers,” “late risers,” and “transitionals.” The three groups could be used to explain different levels of household or personal use, as long as hourly customer use data is available. This report includes a framework for including behavioral factors in demand forecasts. Published in 2016.

The Appendices are posted below, under Project Papers.