This study extends previous residential analysis to office buildings and aims to quantitatively evaluate the effectiveness of passive design strategies. As in the prior study, annual energy consumption and Spatial Thermal Autonomy (sTA) are used as performance metrics. Because internal heat gains in offices are larger and more variable, lighting, equipment, and occupant density are systematically varied as analysis parameters. To capture nonlinearities and interactions among design factors, SHAP-based feature importance is introduced for global sensitivity analysis. Results show that dominant factors differ between energy and sTA and that lower internal loads increase the relative importance of passive design.
Suga et al. (Sun,) studied this question.