Conditional Effectiveness of Night Ventilation During Heatwaves: Nonlinear Interactions Among Event Morphology, Thermal Mass and Operating Strategies

Abstract

Consecutive warm nights during heatwaves erode the cooling potential of night ventilation (NV), making its effectiveness strongly dependent on both the event and the building conditions. This study quantified the nonlinear effects of heatwave morphology, building thermal mass (TM), and NV strategies on multidimensional thermal responses and identified the operating conditions associated with ventilation performance. Using a 1991–2020 climatological baseline, 36 heatwaves were characterized by eight morphology indicators. A total of 18,000 EnergyPlus simulations examined combinations of three TM levels, two ventilation configurations, window-to-wall ratios (WWRs) ranging from 0.10 to 0.70, openable fraction, and opening schedules. Six indicators captured daytime peaks, persistent overheating, peak temperature lag, nighttime overheating, thermal recovery, and nighttime cooling rate (NCR). Machine learning models were used to identify the governing mechanisms and delineate ventilation strategy regimes. All models achieved R2 values above 0.96 under case-level testing, while event-level performance varied across responses. Increasing TM lowered the median daytime peak temperature by 4.5 °C under cross-ventilation but did not improve every response. Predicted daytime peak risk increased sharply when the mean daily minimum temperature exceeded about 27.0 °C. Ventilation configuration also reversed the effects of opening time, duration, and WWR, precluding a single optimum that is independent of heatwave and building conditions. Effective cooling, operationally defined as positive predicted NCR, was generally associated with earlier opening and longer ventilation. Under single-sided ventilation, achieving stable positive predicted NCR generally required ventilation durations longer than approximately 4 h for buildings with low TM and longer than 6 h for those with medium TM. Under cross-ventilation, the corresponding effective domain in the strategy space contracted markedly, and no stable region with positive predicted NCR emerged for buildings with high TM. These findings support a shift from fixed window opening rules to control strategies tailored to heatwave characteristics and building thermophysical properties. The framework provides a basis for natural ventilation operation informed by forecasts, passive cooling retrofits, and thermally resilient building design.