Estimation of temperature setpoints and heat transfer coefficients among residential buildings in Denmark based on smart meter data

Thermal comfort preferences of occupants and their interactions with building systems are top influential factors of residential space heating demand. Consequently, housing stock models are sensitive to assumptions made on heating temperatures. This study proposes a heat balance approach, inspired by the classical degree-day method, applied to an extensive urban dataset.

The goal of this analysis is to determine heterogeneous characteristics, such as temperature setpoints of heating systems and thermal envelope characteristics from an overall population of residential buildings. Measured energy data are utilized for the purpose of the study from the city of Aarhus, Denmark, where the energy usage for heating of circa 14,000 households was monitored over time via smart meters.

These data are combined with actual weather data as well as data extracted by a national building database. Using linear regression and heat balance models, temperature setpoints for the whole dataset are determined with a median and average of 19 °C and 19.1 °C, respectively. Furthermore, building related characteristics such as thermal and ventilation losses per building and overall heat transfer coefficients are extracted at urban scale.

The reliability of the method over its complexity is discussed with regards to the big sample that has been applied to. In general, the overall performance of the approach is satisfactory achieving a coefficient of determination with an average of 0.8, and is found to be in line with previous findings, considering also the high uncertainty associated with building-related input parameters.

The extracted setpoint distribution should be transferrable across Scandinavia.