Are hourly building energy simulations underestimating peak demand? A case for finer time resolutions.

Accurate peak electricity demand estimation is crucial for building energy simulations, yet the impact of time resolution remains under explored. This study quantifies how varying temporal resolutions (1-minute to 60- minute) influence peak demand estimates using real world data from 268 UK residential households.

Findings show that coarser resolutions (such as 30- or 60-minute) systematically underestimate peak demand, with errors increasing from 14.8% at 3-minute intervals to 61.2% at 60-minute intervals.

Normalised root mean square error (NRMSE) analysis further highlights the growing inaccuracies, with mean errors rising from 5.3% to 21.3% across resolutions. These results underscore the importance of selecting an appropriate resolution to balance computational efficiency and accuracy in building energy modelling, particularly for peak sensitive applications.