**1. Introduction**

By 2050, 66% of the world's population will live in urban areas [1], making urbanization one of the critical themes and challenges in this century. This is the case especially for some Asian countries, such as China, where city boundaries are expanding with numerous new constructions every year. China has contributed to approximately 50% of the world's new constructions since 2010 [2]. Rapid global urbanization has resulted in significant increases in energy consumption, greenhouse gas emissions, pollutant emissions, and widespread environmental degradation. Urban areas account for 67–76% of global energy use and 71–76% of CO2 emissions [3]. Cities around the world are searching for strategies to reduce energy consumption and to become green and low-carbon cities, and enhance their resilience in a changing climate.

Building energy consumption accounts for 36% of the global final energy use in 2017, and this number is much higher in urban areas [4,5]. In the U.S., national level building energy consumption databases have been developed and regularly updated to represent actual building energy usage levels. For example, the Residential Energy Consumption Survey (RECS) and Commercial Buildings

Energy Consumption Survey (CBECS) collect energy-related building characteristics and energy usage information [6,7]. However, this kind of open source national building energy consumption database is not available in China.

To better understand building energy consumption in urban areas, besides survey and measurement, urban datasets and urban-scale building energy consumption platforms have been developed based on urban-scale building energy simulations. Urban-scale building energy simulation can play an essential role in sustainable urbanization, allowing planners and policy makers to develop planning strategies using the lens of energy performance.

A research group from the college of Architecture at Georgia Institute of Technology developed a GIS-based urban building energy modeling system, called Urban-EPC. It includes four main models: the Data Preparation Model, the Pre-Simulation Model, the Main Simulation Model and the Visualization and Analysis Model. This Urban-EPC tool also uses physics models and calculates the hourly heat balance of the whole building. It contains three categories of building vintage (based on the construction year), each of which includes 16 building types representing most of the commercial buildings across 16 US climate zones. The development team also conducted a case study for Manhattan. They obtained the building footprint data from New York city planning database with references to Google Earth 3D building [8].

The sustainable design lab at Massachusetts Institute of Technology (MIT) also developed an Urban Building Energy Model (UBEM) for Boston to estimate citywide hourly energy demands at the building level. In this project, the geometric input for Boston was also extracted from GIS shapefiles into the Rhinoceros 3D V5 CAD environment, and a total of 76 di fferent building archetypes were then assigned to individual buildings based on land use and building age. Bayesian calibration was applied to update the probability distributions of uncertain parameters in archetype descriptions using monthly and annual measured energy usage data. EnergyPlus was used to simulate the energy consumption results of individual building models. The urban energy use pattern of di fferent times of the day is visualized and overlaid with the Boston map. The tool can help local communities to evaluate energy related decisions and building retrofit strategies to reduce building energy use. They also predicated future scenarios, including solar photovoltaic (PV) penetration, and demand response strategy implantation [9,10].

Lawrence Berkeley National Laboratory (LBNL) developed and released a web-based urbanscale building stock simulation platform, called City Building Energy Saver (CityBES). It is designed to support building retrofit analysis. CityBES uses an open standard, CityGML, to represent the 3D city models, and then it categorizes buildings into di fferent types, including small/medium/large o ffices, hotels, schools, and hospitals. For each type of these buildings, CityBES generates baseline EnergyPlus simulation models based on the cities' building datasets and user-selected energy conservation measures (ECMs). There are three main layers: the data layer, the simulation algorithms and software tools layer, and the use-cases layer. The neighborhood buildings in CityBES are modeled as shading surfaces in EnergyPlus to consider the shading interactions between buildings. Simulation results, such as energy use intensity (EUI), can be color-coded and mapped to the 3D buildings with the GIS database. A case study using CityBES for San Francisco shows a potential retrofit site energy saving of 23–38% per building [11].

In addition, the Oak Ridge National Laboratory and National Renewable Energy Laboratory have also developed urban scale simulation tools, called AutoBEM and URBANopt, respectively [12,13]. They used similar approaches: generate baseline building energy models for each building type as a template, categorize buildings in the area of interest into corresponding archetype and link to the template results, map the simulation results to a GIS platform for visualization. This method can provide quick design support for large scale energy decision making based on archetype data, without running detailed building energy simulation.

However, the above case studies are based mainly on simulation results. It is important to validate the numerical simulations using ground truth building energy survey data and consider occupants' energy usage behavior. Furthermore, the case studies are for large cities in the US, where rapid urbanization has almost been completed. Due to rapid urbanization, China has a large percentage of new constructions. Meanwhile, old buildings with different years of building exist in the same urban region. The building age variation could be as high as several decades. As they were subject to different building design standards/codes, the same type of building, if built in different years, could show very different building consumption profiles. Therefore, building vintage is a key parameter to consider. However, open-source building energy models for typical archetypes have not been well developed in China. It is important to develop an updated urban-scale building energy consumption platform for China, to understand both the spatial and temporal urban energy system.

This paper shows our efforts on building archetype development for the urban energy simulation platform. Three main archetype buildings (residential building, small office building, and large office building) are created and demonstrated in this paper with the following innovations.

