building energy

energy consumption

energy demand

green building

heating energy

Building Simulation

Urban building energy modeling has become an efficient way to understand urban building energy use and explore energy conservation and emission reduction potential. This paper introduced a method to identify archetype buildings and generate urban building energy models for city-scale buildings where public building information was unavailable. A case study was conducted for 68,966 buildings in Changsha city, China. First, clustering and random forest methods were used to determine the building type of each building footprint based on different GIS datasets. Then, the convolutional neural network was employed to infer the year built of commercial buildings based on historical satellite images from multiple years. The year built of residential buildings was collected from the housing website. Moreover, twenty-two building types and three vintages were selected as archetype buildings to represent 59,332 buildings, covering 87.4% of the total floor area. Ruby scripts leveraging on OpenStudio-Standards were developed to generate building energy models for the archetype buildings. Finally, monthly and annual electricity and natural gas energy use were simulated for the blocks and the entire city by EnergyPlus. The total electricity and natural gas use for the 59,332 buildings was 13,864 GWh and 23.6×10 6 GJ. Three energy conservation measures were evaluated to demonstrate urban energy saving potential. The proposed methods can be easily applied to other cities in China.

Archetype identification and urban building energy modeling for city-scale buildings based on GIS datasets