A clustering-based climatic zoning method for office buildings in China

Abstract Since the influence of building types has not taken into consideration, the current climatic zoning scheme of China may not fully reveal the real energy requirements, especially the annual energy consumption variation of different types of buildings. This study proposes a novel zoning method based on the clustering of building thermal load data of a typical office building. Thermal load simulations were performed under 274 cities’ meteorological conditions across China. Two widely used clustering algorithms : K-Means and Agglomerative Hierarchical Clustering were adopted and compared. Based on the inter-city clustering results, a new zoning map with five distinct zones for building thermal design was developed. To demonstrate the thermal load patterns, especially its dynamic characteristics, three representative cities from the three most densely populated zones were selected for intra-city clustering. The results demonstrate that K-Means method performed better in building thermal design zoning and thermal load pattern analysis, and the optimal thermal zone number is 5. The present zoning method could be applied to other building types and promotes an improvement of the existing zoning strategy. It benefits not only architects, but also building operators and policy makers for energy efficiency design and even green urban design.