Evaluating the Summertime Overheating Signature of Domestic Buildings Using Synthetic Temperature Data

Overheating occurs when the indoor thermal environment presents conditions in excess of those acceptable for human thermal comfort or those that may adversely affect human health. Summertime overheating of homes without active cooling has been demonstrated across diverse locations, such as the UK, USA, and New Zealand. Climate change is predicted to cause hotter summers in many countries with more frequent and intense heatwaves. There is, therefore, a need to understand the likely overheating risk of homes in these future summers. Simple physics-based models are very limited in their ability to produce valid assessments of overheating. More complex modeling using Dynamic Thermal Simulation (DTS) software can simulate internal temperatures when the modeled building is subjected to future weather files. There are, however, acknowledged uncertainties attached to the overheating determined from these simulations. Data-driven models can use temperature monitored in existing buildings to predict future overheating risk. This paper presents the idea of ‘overheating signatures’, simple mathematical models which relate the internal temperature in spaces to the external conditions and occupant behavior. Synthetic data from a single-zone building were used to derive such models and evaluate their ability to ‘predict’ overheating for different UK weather conditions. Analysis of the data revealed that there was a strong correlation between number of hours overheated and the warm period average outdoor air temperature (R2 above 0.94). Applying the regression model to two different UK locations showed high correlation between overheating results predicted by the mathematical model and those from dynamic thermal simulation (R2, 0.94 to 0.98). Based on these findings, we conclude that data-driven models have an important role to play in evaluating overheating risk. Future work is, however, needed to refine the mathematical models with data on a daily timescale and to test them on real-world buildings. Although this research has a focus on the UK dwellings, it is likely of interest to other countries with a temperate climate.