**2. Reference Building Models**

Currently, there are no open sourced well-developed reference building models in China. Residential and office building are selected for prototype development in this paper because these two types of buildings are the top two largest building stocks in China, with a percentage of 73% and 9%, respectively [14].

Working with local project partners, the most popular configurations and geometries for residential buildings as well as office buildings were collected through a survey. Figure 1 shows the geometry of a typical residential building. It is a 10-story apartment building with a total building area of 7836 m2. According to the different orientation, each floor is divided into nine thermal zones: eight apartment units, and one corridor. The floor area of each apartment is about 88 m2.

**Figure 1.** Geometry of the residential reference building.

Figure 2 shows the most popular geometries of a typical small office building (left) and a large office building (right). The small office building has three floors and the large office has eighteen floors. Each floor has four external zones and one core zone. The total building areas are 8176 m<sup>2</sup> for the small office and 26,142 m<sup>2</sup> for the large office, respectively.

**Figure 2.** Geometries of reference office buildings (left: small office, right: large office).

Key building design parameters, such as the building envelope's thermal properties, lighting power density, equipment load density, HVAC system and schedules, were defined based on the corresponding residential and commercial building design standards [15,16].

Based on the survey, the HVAC systems for these three building types are different. Residential and small office buildings use ductless mini-split heat pumps for heating and cooling, while large office buildings use chillers and cooling towers for cooling and boilers for heating.

To capture the average energy usage level, the China Residential Energy Consumption Survey (CRECS) data were used to determine the heating and cooling schedules for residential and small office building. The CRECS was conducted by Renming University in 2012. The CRECS2012 includes residential appliance usage and electricity consumption data from 1450 residential buildings across 26 provinces in China [17]. Valid instances numbering 218 from the hot summer and cold winter climate zone (where Wuhan is located) were used to calibrate the baseline residential model. Figure 3 shows the number of heating days in winter and the number of cooling days in summer, respectively. It can be observed that most people only use heating for less than one month in winter, and use cooling for one to two months in summer. Compared with cooling, the residents seem to be more tolerant of heating. Figure 4 shows the daily distribution of heating and cooling hours. It shows that most people use heating or cooling for less than 5 h per day. The heating and cooling schedules (days/year and hours/day) as well as the temperature setpoints of the reference buildings were adjusted to be the average values according to the survey data.

**Figure 3.** Distribution of heating/cooling days per year in winter/summer.

**Figure 4.** Distribution of heating/cooling hours per day in winter/summer.

The baseline reference buildings were developed using EnergyPlus software. EnergyPlus is an open source simulation engine for whole building energy consumption [18]. It was developed and is supported by the U.S. Department of Energy. EnergyPlus has been widely used and validated by researchers and designers. It is a console-based program, not a user interface. Some graphical interfaces for EnergyPlus, such as DesignBuilder and OpenStudio, are also available. Since the inputs and outputs for EnergyPlus are all text-based, users can easily edit the information to develop a customized system and run parametric simulations using scripts. The detail settings of each model are summarized in Table 1.


**Table 1.** Baseline EnergyPlus model settings.

Wuhan's hourly weather data were used to simulate annual building energy consumption [19]. Figure 5 shows the simulation results of a baseline residential building. The simulated total building electricity consumption is 27.8 kWh/m2. It can be observed that heating and cooling energy consumptions only account for approximately 30% of the total annual electricity consumption. People tend to use the heat pumps only when the weather is too cold or too hot, to reduce their electricity bills. The occupants' behavioral energy saving patterns can be found.

**Figure 5.** Reference residential building electricity consumption breakdown [kWh/m2].

Figure 6 shows the distribution of electricity consumption in the hot summer and cold winter climate zone from the CRECS survey. The mean value (25.8 kWh/m2) matches well with the EnergyPlus simulation result, which further validates the reference building model. It is of note that the electricity consumption is expected to be 35.3 kWh/m<sup>2</sup> (37% higher than the actual mean value) in the Guideline for Energy Consumption Quota of Civil Buildings in Wuhan [20]. Therefore, it is critical to consider occupants' energy use behavior and reflect the actual energy usage when making regional building energy consumption standards.

**Figure 6.** Distribution of electricity consumption from CRECS data.

Similarly, the HVAC schedules of small office and large office buildings were calibrated using the survey data. The annual energy consumptions were simulated in EnergyPlus. Figure 7 shows the simulation results. The total electricity consumption is 61.0 kWh/m<sup>2</sup> for the small office and 130.9 kWh/m<sup>2</sup> for the large office.

**Figure 7.** Reference office building electricity consumptions [kWh/m2] (top: small office, bottom: large office).
