2.3.1. The Fuel Cycle

The fuel cycle consists of the well-to-tank (WTT) stage and the tank-to-wheel (TTW) stage. As for the WTT stage, the primary energy including coal, liquefied gasoline gas, and natural gas are inputted to produce the terminal energy of gasoline and electricity. In 2017, the electricity mix in China is shown in Figure 1. The conversion efficiency of primary energy, the proportion of fuel consumption in various processes and the transportation distance of primary energy can be obtained or calculated based on the data from official yearbooks and other related publications [8,15,16].

As for the TTW stage, the fuel efficiencies of BEVs and PHEVs, as shown in Table 1, are provided by the car marker and have been verified through a fuel consumption record website, where the real-world energy efficiency data are reported by vehicle users [17]. The energy consumption and GHG emissions in the TTW stage are calculated by Equation (1).

$$E\_{\rm TTW} = E\_{\rm electrolyte} \times \text{UF} + \left(E\_{\rm upstream} + E\_{\rm combustion}\right) \times (1 - \text{UF}) \tag{1}$$

where *E*TTW denotes the energy consumed per kilometer in the TTW stage, *E*electricity represents the upstream energy consumption of electricity, while *E*upstream and *E*combustion represent the upstream and the combustion emissions of gasoline, respectively. The utilization factor (UF) is defined as the distance fraction that is powered by electricity whereas (1-UF) represents the fraction of travel powered by gasoline [18]. For BEVs, the UF equals to 1 while that for PHEVs is assumed as 40% in this paper based on the assumption by Hou, Wang and Ouyang (2013) [18]. Similar methodology is applied to calculate GHG emissions. The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation Model (GREET) [19] is used to calculate the energy consumption and GHG emissions in the fuel cycle.

**Figure 1.** The electricity profile of China in 2017.



#### 2.3.2. The Vehicle Cycle

As shown in Table 2, the vehicle and battery related parameters are provided by the car manufacturer, or assumed after personal communication with the car marker and car users.


**Table 2.** The key parameters of representative vehicles.

<sup>1</sup> The lifetime mileage is assumed by the author after personal communication with the car marker and some vehicle owners.

The vehicle cycle includes five phases: material production and the vehicle production, the vehicle assembly, the transportation of the vehicle, the maintenance service and the end-of-life stage.

As for the material production and the vehicle production stage (the vehicle production, for short), the inventory of pre-manufacturing, such as the raw material extracting and processing, is based on published studies and reports [9,19–22], and GaBi software [23], which is an LCA computational platform and accommodates thousands of background processes and elementary flows. This paper splits the vehicle into three parts: the vehicle body (excluding the battery and fluids), the battery and the fluids (including engine oil, brake fluid, transmission fluid, powertrain coolant, and wiper fluid); production-related inputs and outputs of each part are specified. Table 3 contains the list of materials

for each vehicle technologies, and the material breakdown of vehicle body and fluids is based on the reports given by Sullivan and Gaines (2010) [24] Mayyas, et al. (2012) [25] while that of battery packs is based on estimations given by Peters, Baumann, Zimmermann, Braun and Weil (2017) [20], Peters and Weil (2018) [21], Majeau-Bettez, Hawkins and Str Mman (2011) [22]. It is noted that the main composition difference between BEVs and PHEVs is the powertrain, where a PHEV consists of both an electric motor and internal combustion engine, while a BEV is exclusively propelled by the electric motor. In the manufacturing phase, main material transformation processes of the vehicle body are considered, including the stamping, casting, forging, extrusion, and machining, and the inventory is estimated on the basis of previous reportedly data [24–27]. In terms of the battery packs, extensive studies have focused on the cell manufacturing and pack assembly stage. Among these studies, the modelling approach of energy demand (one is to allocate the total energy demand of a plant by its output; another is to use data from theoretical considerations for specific processes) is identified as a major cause of deviated results [20]. However, this comparative analysis will not be affected much by the modelling approach when these vehicles come from the same manufacturing plant. Therefore, we estimate the values based on an LCA review study reported by Peters and Weil (2018) [21]. By following these steps, the energy and GHG emissions associated with vehicle production stage are calculated by using GaBi software.

The assembly stage mainly includes stamping, welding, final assembly, injection molding, and painting. The production of heating, ventilation and air conditioning are not included in the comparative study since almost the same products are used for these different vehicles. In the assembly process, the energy consumption and GHG emissions are based on Mayyas, Omar, Hayajneh and Mayyas (2017) [25], J. L. Sullivan (2010) [28], Papasavva et al. (2002) [29].

The transportation of the vehicle includes two parts, from the production plant to the service shop, and from the maintenance shop to the dismantling sites [21]. The distance is set as 1600 km and 500 km, respectively, and diesel is assumed to be used in the road transportation.

Concerning the maintenance and replacement, we make assumptions based on previous studies, our communication with vehicle users and field investigation in the automobile service factory. As shown in Table 4, it is assumed that the tires and the engine oil should be replaced every 62,500 km, 6250 km, respectively and the wiper fluid, brake fluid, and powertrain coolant are completely consumed every 12,500 km, 62,500 km, and 62,500 km, respectively. In this paper, it is assumed that only one transmission oil is replaced during the life cycle of the car and the lifetime of the battery equals the lifetime of the vehicle.

For the end-of-life stage, this paper considers the energy consumption in the disassembly process and the avoided energy by recycling steel, aluminum, copper, and iron. Although batteries contain some valuable metals that need to be recycled, huge uncertainties exist when recycling activities are not conducted at a large scale. Additionally, most studies conclude that the end of life phase makes a small contribution to the whole life cycle [30–32]; therefore, we disregard the battery recycling in the baseline scenario but discuss it in the following sensitivity analysis. Besides, it is assumed that fluids, glasses and other non-metal materials are not recycled for their relatively cheap price. The energy consumption and regeneration rates are shown in Table 5, which are based on the recycling inventory reported by De Kleine et al. (2014) [33], Ruan et al. (2010) [34].


*Energies* **2019**, *12*, 834

**Table 3.** The material component

 and the mass percentage of each vehicle technology (Unit: kg).


**Table 4.** The maintenance and replacement of vehicle materials.

**Table 5.** The energy consumption in the end-of-life stage.


#### **3. Results**

In this section, the performances of different vehicle technologies are presented. The results of the fuel cycle are calculated in per km, while results of the vehicle cycle are firstly presented in the unit of per vehicle and then presented as per km by dividing the lifetime mileage of vehicles.

#### *3.1. Fuel Cycle*

Based on previous studies, the energy consumption and GHG emissions for BEVs and PHEVs in the fuel cycle are found to be primarily affected by the energy conversion efficiency, carbon intensity of fuels and the fuel efficiency of vehicles.
