Going smart, staying confused: Perceptions and use of smart thermostats in British homes

With UK housing contributing almost one-third of national final energy consumption, energy research and policy are striving to understand how residents perceive and use domestic energy-saving technologies, including smart heating controls. Smart heating controls have been projected to generate significant energy savings, but there is little evidence to suggest that they are more usable than traditional thermostats, and that they cause residents to use their heating more efficiently. Our study aims to contribute to this evidence, by assessing how the perception and use of heating controls changed for a small group of residents in London, United Kingdom, after their dwellings were upgraded to smart heating controls. The residents were part of an interesting niche supplied by a district heat network (DHN) through slow-response underfloor heating systems, originally controlled by traditional programmable thermostats which residents used infrequently and perceived as complicated. The study focused on assessing one particular heating behaviour of residents: using the controls to program heating schedules that reflect their occupancy patterns, a behaviour of great interest to DHN operators due to its potential to help improve the accuracy of forecasting residential heating demand. The study also assessed residents’ perceptions of the usability of heating controls and of their heating systems’ matching of temperature adjustments. We found that after upgrading to smart thermostats, more residents programmed heating schedules to match their occupancy patterns, but they also increased the frequency of ad-hoc temperature and schedule adjustments, indicating that heating demand may vary significantly outside of pre-set schedules, and not conform as well to predictable patterns as expected. Perceived usability of the heating controls also increased after the upgrade, but perceptions of whether the system matched temperature adjustments showed that respondents still found it difficult to understand the intrinsic slow response of their underfloor heating system. Our findings provide insight into the perception and specific use of heating controls for a unique sample of residents, “twice removed” from the most intuitive methods of heating control and challenged by unique issues with their heating supply and system. However, the study is limited by the small sample size and lack of long-term monitoring of heating behaviour, and we recommend further research on larger samples and over longer timescales in order to investigate long-term change and whether this could improve demand forecasting accuracy and contribute to achieving higher efficiency, and thus reduced emissions, in DHN operation.