Sensitivity analysis for identifying key parameters affecting energy consumption in early-stage building design

Residential and commercial buildings account for 40 % of the nation’s energy consumption and are responsible for 35 % of total carbon emissions. To reduce energy consumption, a comprehensive sensitivity analysis of the design parameters that affect the energy performance of building is essential. For commercial buildings, the U.S. Department of Energy (DOE) prototype office buildings serve as standardized benchmark models, offering a consistent baseline for comparative analysis. This study investigates how baseline energy performance varies with factors such as floor area, window and shade size, building height, and aspect ratio etc., highlighting the necessity for detailed evaluations. The aim of this study is to evaluate the impact of 17 early-stage design parameters using Latin Hypercube Sampling (LHS) and parametric simulations. This study accounts for climate-specific variations in heating and cooling energy use intensity (EUI) by developing multiple linear regression (MLR) models for three representative climate zones: 2A (hot), 4A (mixed), and 6A (cold). The MLR models, incorporating both main and interaction effects among parameters, exhibit strong predictive performance, with goodness of fit, R2, values exceeding 95 % for both heating and cooling EUI. Key findings indicate that larger floor areas reduce heating and cooling EUI, while higher ceiling heights increase them. Additionally, for a fixed floor area, a higher length-to-width aspect ratio impacts heating EUI but has negligible effects on cooling EUI. We have also qualitatively explored the underlying cause-and-effect relationships among the influential form parameters with energy use intensity. The study finds that energy use intensity (EUI) varies even among similar buildings due to differences in area-to-volume ratio, window size and orientation, shading length, and aspect ratio. By offering a climate-responsive framework for analyzing design parameters without extensive energy simulations, this research provides novel insights that enhance practical decision-making in energy-efficient building design.