Occupant–building interactions significantly influence indoor environmental quality and energy use, yet they are still often represented in building energy simulation with simplified or schedule-based assumptions. This study quantifies how occupant-specific (heterogeneous) window-opening and window-closing behavior affects space heating energy use in a Swedish residential building under comparable conditions. Using four winters of monitored data from several single-occupant apartments, occupant-specific logistic regression models are developed for window opening and closing action. It is shown that common drivers (e.g., indoor air quality and time of day) coexist with substantial inter-occupant differences. These models are then integrated into a closed-loop co-simulation (IDA ICE–MATLAB) with a calibrated building model to compare occupant profiles under identical boundary conditions and against a consistent closed-window reference case. Impacts are reported as a normalized heating energy increase relative to the closed-window case, providing direct comparison across occupant profiles and different scenarios. The results show that window operation can lead to large and highly variable heating losses, with some occupant profiles increasing space heating energy use by up to threefold relative to the closed-window baseline. These findings demonstrate that representing inter-occupant heterogeneity is essential for reliable energy performance assessment and occupant-centric building control design. • Occupant-specific window opening/closing models learned from multi-winter data. • Twelve logistic-regression models capture heterogeneity across different occupants. • Closed-loop IDA ICE co-simulation quantifies occupant behavior impact on energy use. • Normalized energy-increase metric benchmarks against a closed-window baseline. • Window operation can raise residential space-heating energy use up to threefold.
Farjadnia et al. (Sun,) studied this question.