Effects of building form on energy use for buildings in cold climate regions

Abstract

Building form has significant effects on energy use in buildings, especially in cold climate regions. This research is focused on exploring the influences of parameters relevant to building forms on energy use for office buildings in Harbin, China. The input parameters include building orientation, aspect ratios, window-wall ratio, number of floors, and overall scales. The results show that the number of floors is the only dominant variable that affects annual heating energy use intensity, while the overall building scale is the most critical factor influencing both cooling and electricity use per unit of floor area. The comparison of results derived from machine learning methods indicates that the bagging MARS (Multivariate Adaptive Regression Splines), MARS, RF (random forest) are better models in predicting annual heating use. By contrast, the GP (Gaussian process) and bagging MARS are two most effective models for estimating both cooling and electricity use. The prediction for cooling and electricity intensities is more difficult than heating energy use in this case.