Carbon emission of urban vehicles based on carbon emission factor correlation analysis

Abstract

The CO₂ emission factor is the basis for analyzing vehicle CO₂ emissions. This study establishes a correlation model between the fuel CO₂ emission factor and the mileage-based CO₂ emission factor using fuel consumption data, then analyzes the fuel consumption and CO₂ emission situation of vehicles in Beijing with the established models. The main research conclusions are as follows: The proposed correlation models are effective for analyzing urban vehicle CO₂ emissions. Cars can only meet the national standard limit when traveling at an average speed of 42.17 km/h. During rush hours, fuel consumption in most cities in China exceeds national standards, making it urgent to improve urban traffic efficiency. Due to the decline in urban traffic conditions in Beijing in 2023 (average speed decreased from 29.14 km/h to 24.26 km/h), each passenger car consumes an average of 0.282 L more gasoline per day, emitting an additional 619.87 g of CO₂. This study is of great significance for energy conservation and emission reduction in road transportation.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-86119-3.

Introduction

The report from the International Energy Agency (IEA) indicates that the transportation industry accounts for up to 25% of carbon emissions and is also one of the fastest-growing industries in terms of carbon emissions. Over the past decade, China’s carbon emissions in the transportation sector have maintained an average annual growth rate of over 5% (except for 2020, due to the COVID-19 pandemic)¹. According to data from the website of the Ministry of Public Security of the People’s Republic of China, the number of motor vehicles nationwide had reached 435 million by 2023, including 336 million cars, ranking first in the world². There are 94 cities with over one million cars. The increase in the number of motor vehicles has led to urban traffic congestion, and numerous studies have confirmed that traffic congestion has led to increased emissions^3–6. Despite the booming development of the new energy vehicle industry, traditional fuel vehicles will still dominate for a considerable time in the future. Energy conservation and emission reduction of motor vehicles are key to the low-carbon development of China’s urban road transportation. The accounting of motor vehicle CO₂ emissions, which is typically calculated by multiplying CO₂ emission factors by vehicle activity level data, serves as the foundation for motor vehicle emission studies⁷. The CO₂ emission factor is a crucial component in the accounting of motor vehicle CO₂emissions. This study aims to delve into the correlation between the two most commonly used vehicle CO₂ emission factors and then analyze urban transportation energy conservation and emission reduction based on the correlation.

Vehicle CO₂ emission factors are divided into two categories: fuel CO₂ emission factor and mileage CO₂ emission factor. Although both are widely used, they differ in their sources, units, and applicable scopes. Fuel CO₂ emission factors are influenced by factors such as fuel type (diesel, gasoline, natural gas, LPG, etc.), lower heating value of the fuel, carbon content, etc. Specific country or region’s fuel CO₂ emission factor data are used. When specific country data is not available, default emission factors from the IPCC emission factor database can be referenced, with a unit of kg/TJ⁸. Fuel CO₂ emission factors are primarily applicable to the accounting of motor vehicle carbon emissions at the macro and meso scales and are matched with the total fuel consumption data of a country or region within a certain period (typically one year), which is commonly known as the “top-down” approach in carbon emission inventories. Large-scale international inventory programs, such as the European Union’s Emission Database for Global Atmospheric Research (EDGAR), the National Emissions Inventory developed by the US Environmental Protection Agency, and the Multi-resolution Emission Inventory for China (MEIC) led by Tsinghua University, primarily utilize fuel CO₂ emission factors. Mileage CO₂ emission factors are influenced by numerous factors, including vehicle-specific attributes such as vehicle type, deterioration rate, emission standard, and vehicle weight; natural environmental factors like slope, temperature, and humidity; as well as driving habits and traffic situations, among others.

They generally originate from measured data or motor vehicle emission models built on a large amount of measured data, such as COPERT, MOBILE, EMFAC, MOVES, CMEM, IVE, etc., with the unit being g/km or g/s^9,10. Early motor vehicle emission models focused primarily on exhaust pollutants, and CO₂ emission factors were incorporated later. Mileage CO₂ emission factors are suitable for accounting motor vehicle CO₂ emission inventories at the micro and meso levels and are matched with data such as vehicle types and travel mileage in the study area. They are often used in another “bottom-up” approach for CO₂ emission inventories. For example, motor vehicle emission inventories in cities or regions such as Ireland, the Federal District of Brazil, and Beijing have adopted mileage CO₂ emission factors^11,12. While these two types of motor vehicle CO₂ emission factors are widely used, no study has yet explored the correlation between them.

Although the source, unit, and scope of application of the fuel CO₂ emission factor and mileage CO₂ emission factor are different, the total CO₂ emissions from motor vehicles calculated by the top-down method based on the fuel CO₂ emission factor and the bottom-up method based on the mileage CO₂ emission factor should be the same in the same region during the same time period¹³. These two methods can be mutually verified, and both are widely used at the urban level. Current studies focus on the comparative analysis of vehicle emission results calculated by the top-down method and the bottom-up method, but relevant research reports on the correlation between the two CO₂ emission factors have not been seen yet.

Because of the different sources and units of the two types CO₂ emission factors, an intermediate variable that can link the two must be found to analyze the correlation between them. They can be related to each other through vehicle fuel consumption, and fuel consumption is closely related to vehicle operating conditions such as speed. Speed has a significant impact on fuel consumption and CO₂ emission factors of motor vehicles. For example, motor vehicle emission models such as COPERT, MOBILE, and EMFAC, which are applicable to macro and medium scales, use average speed as a characterization parameter, and speed is also an important parameter of micro emission models such as MOVES, CMEM, and IVE. The relationship between speed and fuel consumption has also been studied by many scholars. Although fuel consumption can correlate two different carbon emission factors, current studies focus on the influencing factors and conditions of fuel consumption^14–16. Further study is needed to analyze the relationships among fuel consumption, CO₂ emission factors, speed, and vehicle CO₂ emissions.

This paper aims to delve into the correlation between the fuel CO₂ emission factor and the mileage CO₂ emission factor, and to construct a correlation model among fuel consumption, CO₂ emission factors, and speed, in order to analyze the impact of traffic conditions on energy conservation and emission reduction for motor vehicles.

Methods

Research area and data

Research area

Beijing is the capital of China, characterized by a warm temperate semi-humid and semi-arid monsoon climate, as shown in Fig. 1. According to the Beijing Statistical Yearbook, by the end of 2023, the number of motor vehicles in Beijing had reached 7,589,000, including 5,431,000 cars. The rapid increase in motor vehicles has led to increasingly serious traffic congestion in Beijing, and numerous studies have confirmed that congestion leads to more emissions. According to data from navigation platforms such as AMap (https://www.amap.com/) and Baidu (https://map.baidu.com/), Beijing has almost always been the most congested city in China in recent years, with the average speed during rush hours in the core area of Beijing in 2023 being only 24.26 km/h.

Fig. 1.

Administrative planning map of Beijing City.

Data

The data used in this study mainly includes the number of motor vehicles, meteorological data, and traffic operation data. The motor vehicle data includes types, ownership, vehicle age, emission standards, driving mileage, and other relevant information, sourced from the Beijing Motor Vehicle Pollution Monitoring Center. Meteorological data mainly includes temperature, humidity, etc., sourced from the China Meteorological Data Network (http://data.cma.cn/). Traffic operation data mainly includes the congestion delay index (which is the ratio of the actual time spent on travel to the time spent in a free flow state), from AMap (https://www.amap.com/) and Baidu (https://map.baidu.com/) navigation platforms.

The Sylphy was the champion in retail sales of sedans in the Chinese passenger vehicle market from 2020 to 2022. Therefore, measurements were conducted using the PEMS-OBEAS600 (portable emission measurement system) to assess the CO₂ emissions of the Sylphy at various speeds. The Sylphy boasts a curb weight of 1287 kg, a fuel consumption of 5.57 L per 100 km, and complies with China’s National VIb emission standards.

Data preprocessing

The MOVES model, a comprehensive motor vehicle emission model developed by the US Environmental Protection Agency, is used in this paper. It can be applied to multiple scales such as macro, meso, and micro. Numerous studies have confirmed the model’s portability and the accuracy of its simulation results. Based on the geographical information, meteorological factors, and vehicle information of Beijing City, the parameters in the model are locally adjusted, as shown in Table 1.

Table 1.

Natural geographic data and transportation data of Beijing.

Natural geographic data	Geographical position	115° 20′–117° 30′ E, 39° 28′–41° 05′ N
	Temperature	11–13 ℃
	humidity	38%
Transportation data	Model composition	7.589 million units
	Fuel type	Mainly gasoline and diesel
	average speed	24.26 km/h

Methodology

Two types of CO₂ emission factors

Fuel CO₂ emission factor

The CO₂ emission factors for fuels should use data from specific countries or regions. When such data are not available, the default CO₂ emission factors for different fuel types provided in the Emission Factor Database by the IPCC Guidelines can be used, with a unit of kg/TJ. The “Data Table of Fossil Fuel Combustion Activity Levels in Beijing” provides the CO₂ emission factors for different fuel types in the study area, as shown in Table 2. Natural gas vehicles account for less than 3% of the vehicle population in China, so natural gas is not considered in this study.

Table 2.

Reference of CO₂ fuel emission factors (kg/TJ).

Fuel type	Default value	Lower limit	Upper limit	Beijing
Gasoline	74,100	72,600	74,800	67,910
Diesel	74,100	72,600	74,800	72,590

Mileage CO₂ emission factors

The mileage CO₂ emission factor is typically differentiated by vehicle type, fuel type, and emission standards, with a unit of g/km or g/s. Its measurement is complex due to many influencing factors. Motor vehicle emission models such as COPERT, MOBILE, MOVES, and IVE, which are built based on a large amount of measured data or physical principles, can provide emission factor data for different vehicle types, fuel types, and emission standards. These models provide researchers with data support and reduce the difficulty of obtaining original emission data. They are constructed based on local temperature and humidity environments, vehicle types, fuel types, and traffic conditions. Therefore, local corrections are needed when using them. In this paper, the MOVES model, which is suitable for macro, medium, and micro scales, is used to obtain emission factors for different vehicle types, with localization processing conducted. Details are provided in Sect. 1.1.3.

Mileage CO₂ emission factors based on speed

To analyze the influence of traffic conditions on CO₂ emissions, this study uses the MOVES model to simulate CO₂ emissions from various vehicle types at diverse speeds, and establishes the mileage CO₂ emission factors model based on speed. See Eq. (1).

1

here: is the mileage CO₂ emission factor; represents the fuel type; represents the vehicle type; and is the CO₂ emission factor function based on the average speed.

According to data from the Ministry of Transport of China in 2022, CO₂ emissions from road transportation accounted for over 80% of the total. Among these emissions, heavy-duty trucks contributed 54%, while passenger cars contributed 33%. The combined emissions from these vehicle types amounted to 87% of road emissions. In Beijing, diesel vehicles represented over 98% of heavy-duty trucks, while gasoline-powered vehicles constituted 95% of passenger cars. Thus, focusing on these predominant vehicle types-gasoline passenger cars and heavy-duty diesel trucks—a speed-based emission factor model was developed. The CO₂ emission factors of the two vehicle types were simulated across a speed range of 0–120 km/h, with an 8 km/h speed gradient based on the localized MOVES model, as depicted in Fig. 2a. The regression equation fitted for gasoline-powered passenger cars is illustrated in Fig. 2b, and the equation along with the goodness of fit (R²) is presented in Table 3.

Fig. 2.

Mileage CO₂ emission factors based on speed. (a) CO₂ emission factors. (b) regression equation fitted.

Table 3.

Speed-based CO₂ emission factor model.

	Vehicle type	Emission factor estimation function (g/km)	R 2
V < 100	Passenger gasoline cars		0.995
	Heavy-duty diesel truck		0.980
V > 100	Passenger gasoline cars		0.839
	Heavy-duty diesel truck		0.980

As shown in Fig. 2a, CO₂ emission factors differ among various vehicle types, but exhibit similar trends with changes in speed. As illustrated in Fig. 2b, the inverse function fits exceptionally well, demonstrating a high goodness of fit (R²) of 0.995. Nevertheless, beyond speeds of V > 100 km/h, displays a notable upward trajectory, contrary to the declining trend of the inverse function. Hence, for speeds exceeding 100 km/h, a polynomial function may be more suitable.

Correlation analysis of two carbon emission factors

At the urban scale, the CO₂ emissions accounting based on fuel CO₂ emission factors should be the same as those based on mileage CO₂ emissions factors , meaning that and can be correlated through the vehicle’s fuel consumption; see Eq. (2). The unit of is kg/TJ, while the unit of is g/km. Due to the difference in units, when correlating them through fuel consumption, the fuel mass and the lower heating value of the fuel are also required. The fuel mass is equal to the product of fuel consumption (F) and mileage (S). The values are shown in Table 1, assuming 100% oxidation of carbon, including CO, CH₄, etc.

2

Where is the CO₂ emission factor of fuel , with a unit of kg/TJ; is the fuel type; is the vehicle type; is fuel consumption per unit mileage in L/100km; is the driving distance in km; is the vehicle ownership; is the average lower heating value of fuel in kJ/kg; is the density in kg/L; is the CO₂ emission factor function based on speed, and the unit is g/km.

Combining with Eq. (1), we can obtain that:

3

and can all be regarded as constants, so Eq. (3) can be transformed into Eq. (4):

4

Where: represents the conversion coefficient of fuel , which is the product of and .

Results analysis

Using Beijing as a case study, this paper validates the mileage CO₂ emission factors derived from speed, as constructed based on the localized MOVES model, against measured carbon emission data from gasoline cars. Subsequently, the vehicle fuel consumption and CO₂ emissions are analyzed based on the correlation model between fuel CO₂ emission factors and mileage CO₂ emission factors.

Verification of CO₂ emission factor based on speed

The speed-based CO₂ instantaneous emission data for the Nissan Sylphy 1.6 L gasoline car measured by PEMS is shown in Fig. 3a, and the CO₂ emission factors for passenger gasoline cars simulated based on the localized MOVES model are shown in Fig. 3b.

Fig. 3.

Comparison of speed-based CO₂ emission factors based on the measured and simulated. (a) The measured CO₂ emission, (b) the simulated CO₂ emission.

Figure 3 illustrates that the emission trends of both factors are largely similar. At speeds below 10 km/h, the CO₂ emission factor is notably high. Between 10 and 40 km/h, the factor fluctuates but demonstrates a sharp decline with increasing speed. From 50 to 100 km/h, the emission factor gradually stabilizes at a lower level. However, between 100 and 120 km/h, a slight upward trend in the emission factor is observed. Given that urban driving typically occurs at speeds below 100 km/h, the inverse function proves to be applicable.

Quantitative correlation analysis of two CO₂ emission factors

For the two vehicle types in the study area—passenger gasoline cars and heavy-duty diesel trucks—the correlation between fuel CO₂ emission factors and mileage CO₂ emission factors can be described by Eq. (5) for passenger gasoline vehicles, and by Eq. (6) for heavy-duty diesel trucks.

5

6

Where: and are fuel CO₂ emission factors, using the local value of Beijing, as detailed in Table 3; and are the conversions coefficient of gasoline and diesel, and which are the product of and , as detailed in Table 4. and are the fuel consumptions per unit mileage, according to China’s ‘Limits of Fuel Consumption for Passenger Cars’ (GB 19578 − 2021) for vehicles with a curb weight of 1500 kg, and ‘Limits of Fuel Consumption for Heavy Commercial Vehicles’ (GB 30510 − 2018) for vehicles with a maximum design gross vehicle weight (GVW) greater than 31,000 kg, as specified in Table 4.

Table 4.

Speed-based CO₂ emission factor model.

Fuel type i	Default value (kg/TJ)	Average lower heating value A (TJ/Gg)	Fuel density ρ (kg/l)	Conversion coefficient C	Fuel consumption limit F (L/100km)
Gasoline	67,910	44.3	0.73	0.03169	8.9
Diesel	72,590	43.0	0.84	0.03625	46.3

Using Eqs. (5) and (6) along with data from Tables 3 and 4, the average speed and CO₂ emission factors for passenger gasoline cars and heavy-duty diesel trucks have been calculated for three different fuel consumption scenarios: at the limit, 85% of the limit, and 75% of the limit. The results of these calculations are detailed in Table 5.

Table 5.

CO₂ emission factors, speeds under different fuel consumption.

Vehicle type	EFf (kg/TJ)	Fuel consumption limit			85% limit			75% limit
		F (L/100km)	EFt (g/km)	V (km/h)	F (L/100km)	EFt (g/km)	V (km/h)	F (L/100km)	EFt (g/km)	V (km/h)
Passenger gasoline car	67,910	8.9	195.46	42.17	7.57	166.14	68.42	6.68	146.59	112.78
Heavy diesel truck	72,590	46.3	1381.52	28.73	39.36	1174.291	51.56	34.7	1105.214	70.12

According to the data presented in Table 5, variations in fuel consumption levels across different vehicle types correspond to significant fluctuations in both CO₂ emission factors and associated vehicle speeds. For instance, in the case of gasoline passenger cars, at fuel consumption levels of the limit (8.90 L/100km), 85% of the limit (7.57 L/100km), and 75% of the limit (6.68 L/100km), the average speeds are 42.17, 68.42, and 112.78 km/h, respectively. Correspondingly, the CO₂ emission factors are 195.46, 166.14, and 146.59 g/km, respectively.

Fuel consumption and CO₂ emission of motor vehicles based on the correlation model

The average speed of motor vehicles is not only closely linked to their fuel consumption and CO₂ emission factors, but also serves as a reflection of a city’s traffic conditions. Chinese navigation platforms like Amap and Baidu have developed the congestion delay index (CDI), calculated as the ratio of actual travel time to free-flow travel time (i.e., the ratio of free-flow speed to actual operating speed), based on average speeds. The index has become a key indicator for assessing urban traffic conditions. This study analyzes the fuel consumption and CO₂ emissions of motor vehicles based on the correlation model of two CO₂ emission factors and the average speeds monitored by navigation platforms in the study area.

According to the ‘China Urban Traffic Report’ released by Baidu Map, in 2023, the CDI during rush hours (morning rush 07:00–09:00, evening rush 17:00–19:00) in the core area reached 2.125 (an area with active human and vehicle traffic), showing a 20.13% increase compared to 2022 (CDI was 1.769). The average vehicle speed in 2023 in Beijing was only 24.26 km/h, whereas in 2022, it stood at 29.14 km/h (the free-flow speed is 51.55 km/h). Table 6 displays the vehicle speed, the corresponding fuel consumption, and CO₂ emission factors for speeds of 24.26 km/h (2023), 29.14 km/h (2022), 42.17 km/h (fuel consumption limit 8.9 L/100km), and 51.55 km/h (free-flow) in Beijing, as shown in Table 4.

Table 6.

Fuel consumption and CO₂ emission factors corresponding to different speeds.

Speed (km/h)	Passenger gasoline car		Heavy diesel truck
	F (L/100km)	CO2EFt (g/km)	F (L/100km)	CO2EFt (g/km)
24.26	11.68	256.61	55.78	1467.86
29.14	10.62	233.13	52.25	1375.00
42.17	8.90	197.08	46.00	1210.49
51.55	8.31	182.41	44.63	1174.32

As shown in Table 5, the average speed of passenger gasoline cars in Beijing during rush hours was 24.26 km/h in 2023 and 29.14 km/h in 2022, with corresponding average fuel consumptions of 11.68 L/100km and 10.62 L/100km. Both fuel consumption levels exceeded the limit of 8.9 L/100 km set by the “Passenger Car Fuel Consumption Limit” standard. When the average speed is greater than 42.17 km/h, the fuel consumption of passenger gasoline cars can meet the standard. In the past decade, the CDI of Beijing in 2022 was the lowest, indicating that severe traffic congestion has caused passenger car fuel consumption to surpass the limit. The free-flow speed in Beijing is only 51.55 km/h, slightly higher than the speed associated with the limit standard (8.9 L/100km), which is approximately 42.17 km/h (resulting in 8.31 L/100km fuel consumption). According to the “China Urban Transportation Report,” among the 100 cities monitored in 2023, Huzhou had the highest average speed during rush hours at 43.65 km/h, slightly above the speed corresponding to the limit (42.17 km/h). The average fuel consumption in the other 99 cities during rush hours exceeded the prescribed limit. In 2022, only three cities—Sanya (44.91 km/h), Xining (44.39 km/h), and Huzhou (42.79 km/h)—met the limit standard. The situation for heavy trucks is similar.

Taking small gasoline passenger cars as an example, this paper analyzes the impact of traffic conditions on carbon emissions in Beijing based on carbon emission factors and average speed. In 2022, the average fuel consumption of gasoline passenger cars in the core area of Beijing was 10.62 L/100km, and the carbon emission factor was 233.13 g/km. In 2023, due to the aggravation of traffic congestion, the fuel consumption per kilometer increased by 0.011 L, and the CO₂ emission increased by 23.48 g per kilometer. In 2023, the average commuting distance in Beijing was 13.2 km. Due to the deterioration of traffic conditions (reduction in average speed), each vehicle consumed an average of 0.282 L more fuel per day, emitted 619.87 g more carbon, and consumed 70.54 L more gasoline per vehicle during the 250-day working year, increasing carbon emissions by 154.97 kg per vehicle. In 2023, the number of small gasoline passenger cars in the downtown area of Beijing was approximately 3.5 million, and the average rate of private cars in the downtown area was 47.1%. Due to traffic congestion, gasoline cars consumed 85,900 tons of gasoline more in 2023, and the increased carbon emissions reached 258,500 tons.

Using gasoline passenger cars as an example, this paper analyzes the impact of traffic conditions on CO₂ emissions in Beijing based on CO₂ emission factors and average speed. In 2022, the average fuel consumption of gasoline passenger cars in the active areas of Beijing was 10.62 L/100km, with a CO₂ emission factor of 233.13 g/km. In 2023, due to worsening traffic congestion, fuel consumption per kilometer increased by 0.011 L, resulting in a CO₂ emission increase of 23.48 g per kilometer. The average commuting distance in Beijing in 2023 was 13.2 km. Due to deteriorating traffic conditions (reduced average speed), each vehicle consumed an additional 0.282 L of fuel per day, emitted 619.87 g more CO₂, and consumed 70.54 L more gasoline per car over the 250-day working year, leading to an increase in CO₂ emissions of 154.97 kg per car. In 2023, there were approximately 3.5 million gasoline passenger cars in downtown Beijing, with a private car usage rate of 47.1%. Due to traffic congestion, these cars consumed an additional 85,900 tons of gasoline in 2023, resulting in increased CO₂ emissions of 258,500 tons.

Discussion

Analysis of CO₂emission factor model

Difference analysis caused by data sources

This paper compares the speed-based CO₂ emission factors derived from measured data and those simulated by the MOVES model (Fig. 3). Qiusi Jin et al. obtained the relationship curve between speed and comprehensive CO₂ emission factors (Fig. 4) using the vehicular specific power (VSP) method (the MOVES model measures emission factors based on VSP) based on approximately 8 million floating car data from the Beijing Floating Car Traffic Information Collection System from January 2013 to October 2014.

Fig. 4.

Comprehensive CO₂ emission factor at different speeds.

From Figs. 2, 3 and 4, despite some differences in data sources, the trend of CO₂ emission factors varying with speed is generally consistent for motor vehicles. The measured CO₂ emission data is from a single vehicle, while the CO₂ emission factors simulated by the MOVES model are averages for the same type of vehicle. The data source in Fig. 4 originates from Beijing’s floating car data, encompassing taxis, buses, and heavy goods vehicles. Therefore, the CO₂ emission factor calculated in Fig. 4 is comprehensive and it is higher than that of passenger gasoline vehicles, but lower than that of heavy diesel vehicles depicted in Fig. 3.

Applicability and limitations of the models

The speed-based mileage CO₂ emission factor model and the correlation model between the two CO₂ emission factors constructed in this paper are both valid within the urban scope, but they are not applicable when the speed exceeds 100 km/h. In other words, these models are suitable for analyzing urban vehicle emissions but not for carbon emission analysis on highways or urban expressways with higher vehicle speeds.

This study conducts simulations based on the most common National V vehicle standards in Beijing. Currently, emission standards primarily target pollutants such as CO, NOx, and PM, with relatively little impact on fuel consumption, i.e., CO₂ emission factors. For instance, NOx emissions under the National V standard are reduced by 25% compared to the National IV standard, and the PM (particulate matter) emission limit is newly introduced (the National IV standard did not have this indicator). Although the measured CO₂ emission data can confirm the applicability of the speed-based CO₂ emission factor model constructed from the localized MOVES model, it represents the average data for this vehicle type. In reality, even for similar models, emission factors at the same speed may vary due to differences in vehicle weight, deterioration rate, temperature, and humidity. There are many other factors that affect the vehicle mileage emission coefficient, so improving the accuracy of mileage CO₂ emission factor estimation is crucial for analyzing urban vehicle carbon emissions.

Recommendations related to China’s transport policy

Based on the research presented in this paper, our suggestions for urban transportation in China are as follows: Firstly, develop new energy vehicles, as congestion does not increase the energy consumption of electric vehicles. For China’s increasingly deteriorating traffic conditions, electric vehicles represent the best pathway to achieve transportation emission reductions. Secondly, enhance transportation efficiency through intelligent means (i.e., increasing the average urban vehicle speed), which is also a significant approach to transportation energy conservation and emission reduction in Chinese cities where traditional vehicles still constitute the majority (as of the end of 2023, the number of new energy vehicles in China reached 20.41 million, accounting for 6.07% of the total number of vehicles). Lastly, actively develop public transportation to reduce private vehicle trips, which is another crucial way to decrease transportation emissions in China.

Fuel consumption

Correlation analysis of two CO₂ emission factors based on fuel consumption

For the sources and units of the fuel CO₂ emission factor and the mileage CO₂ emission factor are different, this paper uses fuel consumption to correlate the two types of CO₂ emission factors, as shown in Eqs. (4), (5), and (6). Vehicle fuel consumption varies with vehicle weight, degradation level, average speed, environment, and other factors, which are very complex. When analyzing specifically, fuel consumption is used as a constant in this paper. The CO₂ emission factor and speed are analyzed under different fuel consumption levels. Although the complexity of fuel consumption is simplified, it can still reflect the relationship between vehicle speed, fuel consumption, and carbon emissions.

Fuel consumption data

This paper uses the limit data from two standards: “Limits of Fuel Consumption for Passenger Cars” (GB 19578 − 2021) and “Limits of Fuel Consumption for Heavy-duty Commercial Vehicles” (GB 30510 − 2018). The limit data in this paper are the average data for the same vehicle type. Taking passenger gasoline vehicles as an example, the fuel consumption limit of 8.9 L/100km corresponds to a curb weight of 1500 kg. The 1500 kg is an average, and the limit of 8.9 L/100km is also average data. The curb weight of small cars is relatively lower, while that of SUVs is relatively higher. For example, in 2022, the most popular car in China, the Nissan Sylphy 1.6 L, has a curb weight of approximately 1250 kg, while the Haval H6 2021 model (compact) has a curb weight of 1659 kg. The reason for adopting 1500 kg as the average curb weight is the significant increase in SUV sales. For instance, in 2022, SUV sales reached 9.406 million, accounting for about 45.8% of the passenger car market.

Due to the complex traffic conditions in Chinese cities and the low average speed, actual fuel consumption is generally higher than the labeled WLTC fuel consumption of the vehicle. For example, the labeled fuel consumption (L/100km) of the Haval H6 2022 1.5T automatic two-wheel drive Supreme+ (with a curb weight of 1560 kg) with WLTC certification is 7.13, and the low and high fuel consumption measured by the Ministry of Industry and Information Technology are 7.78 and 10.68 L/100km, respectively. Meanwhile, the average comprehensive fuel consumption measured by Xiaoxiong Fuel Consumption Official (http://androil.sinaapp.com/modelyh/59969.html) is 9.23 L, based on a total mileage of 1,808,914 km. The implementation of the National VI B emission standard began on July 1, 2023. As vehicle models with lower emission standards gradually exit the market, fuel consumption is expected to further decrease.

Conclusion

This paper focuses on analyzing the relationship between the two most commonly used vehicle CO₂ emission factors: fuel CO₂ emission factor and mileage CO₂ emission factor. To correlate the two types of carbon emission factors with different sources and units, this paper develops a speed-based mileage CO₂ emission factor model based on the localized MOVES model, and further establishes a correlation model between the two CO₂ emission factors. The validity of the speed-based mileage CO₂ emission factor model constructed in this paper is verified through measured data. Furthermore, taking Beijing as an example, the validity and applicability of the models, the fuel consumption and CO₂ emission situation of motor vehicles in Beijing are analyzed based on the correlation model and average speed. The results confirm that the correlation models proposed in this paper for the fuel CO₂ emission factor and mileage CO₂ emission factor are effective for analyzing urban motor vehicle CO₂ emissions. The national standard limit can only be met with an average speed of 42.17 km/h. Fuel consumption in most Chinese cities during rush hours exceeds national standards, making it urgent to improve urban traffic efficiency. Due to the deterioration of urban traffic conditions in Beijing in 2023 (average speed decreased from 29.14 km/h to 24.264 km/h), each passenger car consumes an average of 0.282 L more gasoline per day, emitting an additional 619.87 g of CO₂. Gasoline cars consumed 85,900 tons of gasoline more in 2023, and the increased carbon emissions reached 258,500 tons.

Although the proposed speed-based mileage CO₂ emission factor model and the correlation models between the two CO₂ emission factors are effective, they are not applicable when the speed exceeds 100 km/h. That is, the models are suitable for analyzing urban vehicle emissions but not for carbon emissions on highways or expressways where vehicle speeds are higher. The mileage CO₂ emission factors discussed in this paper are speed-based, but in reality, a vehicle’s mileage CO₂ emission factor is influenced by various factors. While average speed is a key influencing factor, the motor vehicle CO₂ emission data for Beijing under different operating conditions analyzed in this paper are estimated values. With continuous technological advancements, fuel quality is improving, and vehicle fuel consumption is decreasing, meaning the CO₂ emission factor per unit distance will gradually decline. There are numerous factors that affect a vehicle’s mileage emission factor, so improving the accuracy of calculating this factor is crucial for analyzing urban vehicle carbon emissions.

This study is of great significance for energy conservation and emission reduction in road transportation.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

IEA

International Energy Agency

IPCC

Intergovernmental Panel on climate change

LPG

Liquefied petroleum gas

MEIC

Multi-resolution emission inventory for China

COPERT

Computer program to calculate emissions from road transport

EMFAC

Emission factor

MOVES

Motor vehicle emission simulator

CMEM

Comprehensive modal emission model

IVE

International vehicle emission mode

Author contributions

Jingwen Wang, writing and methodology; Hongna Dai, methodology; Haixia Feng, methodology; Meng Guo, data processing and validation; Vladimir Zylianov, modification and polishing; Zhongke Feng and Jipeng Cui, fund.

Funding

This research was funded by the National Natural Science Foundation of China (52102412, 42330507); Natural Science Foundation of Shandong Province (ZR2022MG077); Key Research and Development Program of Shandong Province (Soft Science Project) (2024RZB0703); Foreign Expert Project (DL2023023003L); Foreign Experts Double Hundred Program of Shandong Province (WSP2024007); 5·5 Engineering Research & Innovation Team Project of Beijing Forestry University (BLRC2023A03); Natural Science Foundation of Beijing (8232038, 8234065); the Key Research and Development Projects of Ningxia Hui Autonomous Region (2023BEG02050).

Data availability

The data has been uploaded in the supplementary materials.

Declarations

Competing interests

The authors declare no competing interests.