Abstract

Smart thermostats allow continuous learning, remote scheduling and control of indoor temperature. This paper empirically evaluates indoor environmental conditions, occupant experiences and prevalence of summertime overheating in three low-energy dwellings with smart thermostats and compares the results with three similar dwellings with standard programmable thermostats. The study uses building performance evaluation methods combining time-series data on temperature, relative humidity and window opening with survey data on occupant perception of thermal comfort and heating control over the period 2019–2020. While there was little difference observed in the measured and perceived indoor temperatures between dwellings with and without smart thermostats, the six dwellings were different in the way they heated their homes and controlled their indoor environment. A wide indoor temperature range of 16oC–22oC was observed in dwellings with smart thermostats during the heating season. The majority of dwellings also experienced summertime overheating with temperatures in bedrooms going up to 34oC. Individual heating preferences dominated the use of smart or standard thermostats ranging from Cool Conserver, On-off Switcher to On-demand Sizzler. It is vital that energy models consider a range of heating preferences to avoid a gap between expectation and reality. Practical application: Actual in-use performance of dwellings with smart thermostats is necessary for their large-scale deployment. A wide range of thermostat behaviours are documented; therefore, it is vital that energy models consider a range of heating preferences to minimise the gap between energy models and reality. As smart home appliances and controls become more commonplace, the findings demonstrate their need for resident training and trouble-shooting support to ensure smart thermostats deliver their expected benefits. Since most of the case study dwellings experienced summertime overheating, it is also vital that building design tackles overheating through passive measures.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call