Introduction

In recent years, driven by a series of national policies and market demand, China’s transportation industry has sustained rapid development. With the continuous growth of transportation activities, the transportation sector has become an important source of global greenhouse gases. CO2 emissions from road transportation currently account for three-quarters of the emissions of the entire transport sector. The resulting environmental pollution and traffic congestion are becoming increasingly serious. Air pollutants not only have side effects on the economy, but also seriously affect people’s physical and mental health (Hergueta et al. 2017; Qin et al. 2019). In order to control global climate change, at the 75th Session of the United Nations General Assembly, President Xi Jinping promised that China’s CO2 emissions will reach the peak before 2030 and carbon neutrality will be achieved before 2060.

With the implementation of a driving-restriction policy in Beijing, opinions on its energy conservation and emission reduction effects have been controversial and can be mainly divided into two different views. The first view is that a driving-restriction policy could effectively reduce air pollution and traffic congestion and help improve residents’ health (Liu et al. 2016). Chen et al. (2021a) used daily data for before and after the COVID-19 shutdown and found that a driving-restriction policy for local fuel vehicles and a driving-restriction policy based on the last digit of the license plate had the best effects in terms of reducing air pollution. Zhang et al. (2014) concluded that strengthening traffic control measures could play an important role in reducing the actual fuel consumption and CO2 emissions of passenger vehicles. Wen and Bai (2017) took Beijing as an example and established a system dynamics model to simulate the impacts of different traffic strategies on urban transport energy consumption and carbon emissions. Their results showed that a driving-restriction policy could effectively reduce energy consumption and carbon emissions. The above-cited scholars studied the policy from the aspects of the environment, health, traffic, etc., and adopted a friendly attitude.

Another view is that a driving-restriction policy would have serious side effects in some areas, such as air pollution and second-car purchase rates. Tu et al. (2021) used the inverse distance weighting model to determine that a driving-restriction policy could significantly reduce the increase in the NO2 content in the Nanjing traffic analysis area, but the increase in the NO2 content would be higher in some areas due to the choices of different decision makers. Ma et al. (2020) evaluated the benefits and risks of a driving-restriction policy using on-site and Internet survey methods, and their results showed that the majority of visitors had a positive view regarding driving-restriction policies. Others reacted negatively to driving-restriction policies by buying another car, driving in unrestricted hours, and traveling in violation of regulations. Hulkkonen et al. (2020) quantitatively assessed the effects of a driving-restriction policy on CO2 emissions, PM10 emissions, and PM10 concentrations and assessed driving-restriction policy sanctions through peer-reviewed analysis. It was not intended as an incentive. Yi et al. (2022) also conducted a horizontal study in 16 cities in China, and the results showed that the driving-restriction policy had different effects in different cities. For example, the control of winter haze, while effective in some cities, was not effective in Luoyang, Tangshan, Xinyang, and Zhoukou. Liu and Lei (2019) mentioned that private car owners tend to buy a second private car to take advantage of the unrestricted travel policy, which leads to a continuous surge in the number of motor vehicles and exacerbates the level of urban traffic congestion. Yang et al. (2014) believed that many families choose to buy a second private car with a different tail number stimulated by the car-restriction policy. The above-cited scholars believe that the driving-restriction policy brings negative results. The purchase of the second car would weaken the implementation effect of the car-restriction policy.

Urban air pollution has become one of the most important factors hindering urban development in most emerging economies. Studies on energy conservation and emission reduction in China and abroad have mainly suggested reducing air pollution by reducing fuel consumption of fuel-fired vehicles (Robertson 2019), controlling the growth rate of motor vehicles (Luo et al. 2021; Jiang et al. 2019), and increasing the scrapping rate of motor vehicles (Chen and Lei 2017). Yang et al. (2018a) calculated the pollutant emission factors of motor vehicles in the Beijing-Tianjin-Hebei region, generated a high-spatiotemporal-resolution motor-vehicle-air-pollutant emission list based on traffic flow data, and evaluated the environmental effects of motor-vehicle pollution control measures. Chen and Lei (2017) used a path analysis model to estimate the direct, indirect, and total impacts of motor vehicles on CO2 emissions from transportation in Beijing and proposed that in order to further curb the growth of carbon emissions from transportation, the quality of the population should be appropriately improved, the scale of motor vehicles should be controlled, clean energy technology should be developed and promoted, and the transportation energy intensity and traffic intensity should be reduced. Dey et al. (2019) and Zhang et al. (2016) used a computer program to calculate the emissions from road traffic, and used the (COPERT) model to establish emission lists of major air pollutants from motor vehicles in Berlin and Beijing, respectively, providing references and support for motor-vehicle pollution prevention and coordinated control. Rivera-González et al. (2020) used the low emissions analysis platform (LEAP) model to forecast and analyze energy-related carbon and pollutant emissions in the transportation sector in China, South Korea, Malaysia, and Ecuador and proposed corresponding improvement measures. Gupta et al. (2019) simulated the carbon-emission-reduction effect of a carbon tax on Indian road passenger transport by establishing a system dynamics model. Based on the double dividend theory of environmental tax, He et al. (2019) empirically studied the environmental and economic consequences of an energy tax and vehicle traffic tax by establishing a panel autoregressive distributed lag model (ARDL) model. Liu and Dong (2021) used a threshold regression model and geographical time-weighted regression model to show that urban traffic development has a significant threshold effect on PM2.5 pollution. These scholars used different methods to study how to effectively reduce CO2 emissions from road transportation, which will subsequently help save energy and reduce emissions.

In terms of CO2 emission reduction, many scholars have carried out extensive research. Jiang et al. (2022) suggested that the government must properly address the issues of rapid urbanization, improve institutions, and reduce the excessive use of fossil fuel energy. CO2 emissions could be mitigated by prioritizing demand reduction according to oil product value, starting with the most valuable products (Krantz et al. 2022). To reduce CO2 emissions, a set of reduction measures were proposed, including technological methods of CO2 utilization and removal (Hepburn et al. 2019), campaigns to promote shared mobility, economic incentives in public transport, and integrated parking management policies, among others (Sobrino and Arce 2021). These studies provide theoretical support for the optimization of CO2 emission reduction strategies.

Based on the existing literature, there are two main research regrets. First, most of the current research on private-car driving-restriction policies involved static analysis in China, and car driving-restriction policies have seldom been analyzed from a dynamic perspective. Second, most existing studies focused on the evaluation of a car driving-restriction policy, such as evaluation of the pollution reduction, carbon reduction, and traffic congestion mitigation effects. How to improve the car driving-restriction policy should be the key direction of traffic and environmental research in China at present. In view of this, it is of practical significance to explore the effects of driving-restriction policies and propose improvement measures for the medium- and long-term to reduce pollution and carbon emissions from urban traffic.

The main contributions of this study are the following. First, most studies pay less attention to the dynamic changes of key factors, so the system dynamics model of energy conservation and emission reduction was established in this study to reveal the negative effects of the driving-restriction policy (from the perspective of medium- and long-term dynamic simulation) and to compensate for the lack of previous research. Second, combining the gray Verhulst model with system dynamics, a nonlinear function construction method is proposed to dynamically depict the growth rate of GDP and numbers of private cars. Third, by introducing a technology policy, and through continuous debugging technology to reduce the impact of fuel-consumption factor and improved emission-reduction factor, two regulating factors, as well as the adjustment factor interval of the carbon peak target, could be achieved by dynamic simulation. The workflow of this research is shown in Fig. 1.

Fig. 1
figure 1

Flowchart of this research

Full size image

Methodology

The system-dynamics method originated from Professor Forrest of the Massachusetts Institute of Technology. It is a quantitative method based on feedback control theory and computer simulation technology that is usually used to study complex social economic systems. Because it is suitable for dealing with long-term and periodic problems, the system dynamics method was adopted in the present work.

Model boundary hypothesis

In this article, the energy-conservation and emission-reduction system was divided into five subsystems: economic, government, environment, technology, and energy subsystems. The economic and technology subsystems were the input systems, and the environment and energy subsystems were the output systems. The government subsystem was the main body for macro policy control. Figure 2 shows a theoretical model of the energy-conservation and emission-reduction system. The main assumptions of this model are as follows.

  1. (1)

    This article mainly considers CO2 and PM2.5 emissions from private cars.

  2. (2)

    This article assumes steady growth of the gross domestic product (GDP) and does not consider the large impacts of other unexpected events on the economy.

  3. (3)

    In this article, the amount of CO2 pollution and the PM2.5 concentration are used as indicators to measure the amount of pollutants, and the quantity of fuel used in private cars is used as an indicator to measure energy.

Fig. 2
figure 2

Theoretical model of energy-conservation and emission-reduction system

Full size image

Figure 2 includes five subsystems.

  • Economic subsystem: The economic subsystem mainly studies the influences of the GDP on the main variables.

  • Government subsystem: The government subsystem has the policy regulation, supervision, and public service functions. By changing the degree of environmental pollution, the government adjusts the implementation intensity of various policies to achieve carbon reduction and energy saving.

  • Environmental subsystem: In this paper, we mainly study the influence of the improvement of people’s living standard on environmental tolerance under continuous growth of the population and GDP, and the reaction of environmental pollution loss to the national GDP, such as environmental pollution caused by CO2 pollution.

  • Technology subsystem: The main goal is to achieve carbon reduction and energy saving through government investment in science and technology and the transformation of science and technology to improve the technical level.

  • Energy subsystem: We mainly study the road traffic energy-conservation and emission-reduction models. The implementation effect of the policy is measured using the quantity of fuel used in private cars.

Causal loop diagram

The above five subsystems interact and restrict each other. According to their relationships, a causal loop diagram of the energy conservation and emission reduction model for private cars was constructed (Fig. 3).

  • Loop 1: Amount of CO2 pollution \(\stackrel{+}{\to }\) Degree of air pollution \(\overline{\to }\) GDP \(\stackrel{+}{\to }\) Environmental protection investment \(\stackrel{+}{\to }\) Amount of CO2 process \(\overline{\to }\) Amount of CO2 pollution

  • Loop 2: Amount of CO2 pollution \(\stackrel{+}{\to }\) Degree of air pollution \(\overline{\to }\) GDP \(\stackrel{+}{\to }\) Number of private cars \(\stackrel{+}{\to }\) Private car trips \(\stackrel{+}{\to }\) Amount of CO2 emission \(\stackrel{+}{\to }\) Amount of CO2 pollution

  • Loop 3: Amount of CO2 pollution \(\stackrel{+}{\to }\) Degree of air pollution \(\overline{\to }\) Air quality \(\overline{\to }\) Government decision-making \(\stackrel{+}{\to }\) Driving-restriction policy \(\stackrel{+}{\to }\) Number of private cars \(\stackrel{+}{\to }\) Private car trips \(\stackrel{+}{\to }\) Amount of CO2 emission \(\stackrel{+}{\to }\) Amount of CO2 pollution

  • Loop 4: Technology policy \(\overline{\to }\) Fuel consumption per 100 km \(\stackrel{+}{\to }\) Private car fuel quantity \(\stackrel{+}{\to }\) Amount of CO2 emission \(\stackrel{+}{\to }\) Amount of CO2 pollution \(\stackrel{+}{\to }\) Degree of air pollution \(\overline{\to }\) Air quality \(\overline{\to }\) Government decision-making \(\stackrel{+}{\to }\) Technology policy

Fig. 3
figure 3

Causal loop diagram

Full size image

Loop 1 is a positive feedback loop. The degree of air pollution increases as the amount of CO2 pollution increases. The air pollution causes a decrease in the GDP. When the GDP decreases, environmental-protection investment decreases accordingly, so the amount of CO2 process decreases, and the amount of CO2 pollution in the air increases.

Loop 2 is a negative feedback loop. The economic losses caused by air pollution also lead to a decrease in the per capita GDP, and thus, the number of private cars decreases and the number of private car trips also decreases. Subsequently, the amount of CO2 emissions decreases, so the amount of CO2 pollution decreases. In addition, with continuous economic growth, residents’ incomes increase, which plays a positive role in the growth rate of motor vehicles. Households with a high family income have a higher probability of buying private cars than households with a low family income; thus, income promotes the growth rate of private cars.

Loop 3 is a positive feedback loop. The amount of CO2 pollution leads to serious air pollution, and the reduction of the air quality leads to an increase in the implementation of driving-restriction policies by the government, which stimulates some families to buy a second car. As the number of private cars increases, the amount of CO2 emissions from private cars also increases, which increases the amount of CO2 pollution in the atmosphere.

Loop 4 is a negative feedback loop. With the implementation of a technology policy, the fuel consumption per 100 km decreases significantly, and the quantity of fuel used in private cars decreases accordingly. The amount of CO2 emissions decreases, the amount of CO2 pollution in the atmosphere decreases, and the degree of air pollution decreases. The government sees that improving air quality will reduce the implementation of the technology policy, i.e., negative feedback.

It can be seen that driving-restriction policy has a dual role and can reduce the number of motor vehicle trips in the short term. This will relieve the degree of traffic congestion and reduce CO2 emissions. But in the long run, this policy will induce new demand for cars and increase the amount of CO2 pollution. Technology policy solves the problem of pollutant emissions from the source of private cars, but also can save energy, so that the amount of CO2 pollution is alleviated, and the environmental quality is improved.

System dynamics model

As the capital of China, Beijing is China’s political, economic, cultural, and technological innovation center, and its governance measures are more conducive to improving the traffic driving-restriction policy. At the same time, the governance experience of Beijing can also provide a more useful reference for the improvement of relevant policies in other cities. In this article, Beijing is taken as the research object and variables related to energy conservation and emission reduction are selected to construct a system dynamics model (Fig. 4). This article refers to the China Statistical Yearbook, the China Vehicle Emission Control Annual Report, and other materials. According to historical data for the related variables from 2010 to 2020, the following main equations are obtained, and their parameters are listed in Table 1.

Fig. 4
figure 4

Stock and flow diagram for energy conservation and emission reduction

Full size image
Table 1 Road traffic parameters and initial values for Beijing in 2010
Full size table

The main parameters of the energy conservation and emission reduction model and the initial values for 2010 are presented in Table 1. Some of the parameter settings in the model are from previous studies or official data (Table 2). The main equations in the model are as follows.

  1. (1)

    Private car fuel quantity = Private car trips × Average annual mileage of private cars × Fuel consumption per 100 km × Gasoline oxidation coefficient ÷ 100

  2. (2)

    Number of private cars = INTEG (Growth of private cars—Scrapped of private cars, 3.9 × 10.6)

  3. (3)

    Growth rate of private cars = Private car growth rate table function – Purchase restriction policy

  4. (4)

    Private car trips = Number of private cars × Proportion of private car trips × Driving-restriction policy

  5. (5)

    Amount of CO2 emission = Private car fuel quantity × (CO2 emission factor—Improved emission reduction impact factor)

  6. (6)

    Amount of CO2 process = GDP × Proportion of environmental protection investment × Proportion of CO2 environmental protection investment × 1.04 × CO2 treatment efficiency

  7. (7)

    Amount of CO2 pollution = INTEG (Amount of CO2 emission—Amount of CO2 process, 1.34711 × 10.10)

  8. (8)

    CO2 environmental protection investment = Proportion of CO2 environmental protection investment × Environmental protection investment

  9. (9)

    PM2.5 concentration = Amount of PM2.5 emission—Amount of PM2.5 process

  10. (10)

     Fuel consumption per 100 km = 7.5 – Technology to reduce fuel consumption impact factor

  11. (11)

     Degree of parking space tension = IF THEN ELSE (Poor demand for parking spaces >  = 8 × 106, 0.8, IF THEN ELSE (Poor demand for parking spaces >  = 7.5 × 106, 0.7, IF THEN ELSE (Poor demand for parking spaces >  = 7 × 106, 0.6, IF THEN ELSE (Poor demand for parking spaces >  = 6.5 × 106, 0.55, IF THEN ELSE (Poor demand for parking spaces >  = 6 × 106, 0.5, IF THEN ELSE (Poor demand for parking spaces >  = 5.5 × 106, 0.45, IF THEN ELSE (Poor demand for parking spaces >  = 5 × 106, 0.4, IF THEN ELSE (Poor demand for parking spaces >  = 4.5 × 106, 0.35, IF THEN ELSE (Poor demand for parking spaces >  = 4 × 106, 0.3, IF THEN ELSE (Poor demand for parking spaces >  = 3.5 × 106, 0.25, IF THEN ELSE (Poor demand for parking spaces >  = 3 × 10.6, 0.2, 0.1)))))))))))

  12. (12)

     Poor demand for parking spaces = Demand for parking spaces – Actual number of parking spaces

  13. (13)

     Per vehicle area of road = Number of road area ÷ Motor vehicle trips

  14. (14)

     Environmental protection investment = GDP × Proportion of environmental protection investment

  15. (15)

     Air quality = 1 – Degree of air pollution

Table 2 Description of the parameter setting of the simulation model
Full size table

Extreme condition test

The extreme condition test is mainly conducted to investigate whether relevant variables had realistic significance and conform to realistic changes when they were affected by extreme strategies. Figure 5 shows that when the driving-restriction policy was at the extreme values of 1 and 0.1, the number of private car trips corresponded to the maximum and minimum values, respectively, and the amount of CO2 emission also corresponded to the maximum and minimum values, respectively, consistent with the actual situation and with a certain rationality.

Fig. 5
figure 5

Extreme condition test

Full size image

Model validation

The model was tested using historical data to ensure its reasonableness and applicability. The GDP, the number of private cars, and the year-end road area were taken as examples for testing, and the test results are presented in Table 3.

Table 3 Model validation results
Full size table

As can be seen from Table 3, the average errors in the GDP, number of private cars, and road area were 1.8%, 1.7%, and 2.3%, respectively, and the maximum errors were 3.88%, 3.97%, and 4.81%, respectively. The errors were all within 5%, indicating a high degree of model fitting. Thus, it was reasonable to use the model for the subsequent simulations.

Results and discussion

Simulation results

Impacts of driving-restriction policy

The driving-restriction policy was an administrative measure that prevented a motor vehicle from traveling for a specific period of time based on the last digit of its license plate. Since Beijing officially implemented this driving-restriction policy in 2008, more and more cities have followed suit. Taking Beijing as an example, the existing driving restrictions limit two tail digits on working days, so the driving-restriction policy quantification is 0.8, and the unrestricted case is 1. To investigate whether the driving-restriction policy saves energy and reduces emissions, a comparison was performed (Fig. 6).

Fig. 6
figure 6

The effects of the driving-restriction policy on the major variables

Full size image

It can be seen from Fig. 6a that the development trends of the two curves are the same, but the amount of CO2 pollution under the no-limit scenario is lower than that under the driving-restriction policy scenario, indicating that the amount of CO2 pollution does not decrease after the implementation of the driving-restriction policy, and the carbon emission peak target is not achieved before 2035 under either scenario. As can be seen from Fig. 6b, the curves for the driving-restriction policy and the no-limit scenarios exhibit decreasing trends in the early stage of the simulation, but the degree of air pollution begins to increase over time, indicating that the policy fails to improve the air quality in the long-term. As can be seen from Fig. 6c, in the early stage of the driving-restriction policy scenario, the number of private car trips decreases significantly, but this only lasts for a few years. In about 2013, the number of private car trips under the driving-restriction policy scenario begins to pick up, and even surpasses the number of private car trips under the no-limit travel scenario after 2018. The original intention of the driving-restriction policy was to limit the number of motor vehicles on the road. However, due to the inconvenience of public facilities and the government’s subsidy policy for motor vehicles, more and more families buy a second car with different tail digits to travel, which leads to an increase in the number of private car trips instead of a decrease, which is contrary to the original intention. As private car ownership continues to rise, the degree of parking space tension increases (Fig. 6d).

Therefore, the driving-restriction policy can reduce the number of private car trips and air pollution in the short-term. However, in the long-term, it fails to effectively reduce the amount of CO2 and air pollution emitted by motor vehicles and aggravates the degree of parking space tension, resulting in certain side effects.

Impact of technology policy

In view of the shortcomings of the driving-restriction policy, we introduced a technology policy to explore its impact on carbon emission reduction and energy consumption through dynamic simulations. The technologies to reduce the fuel consumption impact factor and improve the emission reduction impact factor were selected as the policy adjustment factors to analyze the energy saving and emission reduction effects from the perspective of the fuel consumption per 100 km and the CO2 emission factors for gasoline combustion. These two factors were selected for the following reasons. First, according to the available documentation, e.g., the Medium and Long-term Development Plan of Automobile Industry, the average fuel consumption of new passenger cars should be reduced to 4.0 L/100 km by 2025, but the average fuel consumption of existing vehicles in Beijing is 7.5 L/100 km. Second, CO2 emissions are directly related to the CO2 emission factor. The lower the CO2 emission factor is, the larger the direct reduction in the amount of CO2 emissions is, which has a direct effect on carbon reduction. The variations in the output variables under the condition of a science and technology policy were investigated by changing the parameter values of the regulatory factors of the technology to reduce fuel consumption and improve emission reduction.

In order to study the effects of the two regulatory factors on energy conservation and emission reduction, first, we studied the effect of the technology to reduce the fuel consumption impact factor. The improved emission reduction impact factor was held constant at 0.1, and the technology was debugged to reduce the fuel consumption impact factor. The simulation results are shown in Fig. 7a, c, and e. In Fig. 7, CO2 denotes the amount of CO2 pollution, PM2.5 denotes the PM2.5 concentration, and fuel denotes the quantity of fuel used in private cars. Then, the improved emission reduction impact factor was studied. The technology fuel consumption reduction impact factor was held constant at 0.2, and the improved emission reduction impact factor was adjusted. The simulation results are shown in Fig. 7b, d, and f.

Fig. 7
figure 7

Sensitivity analysis of a single regulatory factor

Full size image

On the horizontal coordinate, adjusting only one regulatory factor is not conducive to the attainment of the carbon peak target. From the ordinate view in Fig. 7a, c, and e, when the improved emission reduction impact factor is held constant at 0.1, the impact factor of the technological fuel consumption reduction changes significantly between 0.2 and 1 and changes slowly from 1 to 3 but remains basically the same, and curves 3 and 4 overlap. This shows that with the improvement of the regulatory factors of the technology to reduce fuel consumption, the amount of CO2 pollution gradually decreases, the air quality gradually improves, and the fuel consumption of private cars gradually decreases. However, the technology to reduce the fuel consumption impact factor should not exceed 3, and values between 1 and 3 are the most appropriate. In Fig. 7b, d, and f, when the technology to reduce the fuel consumption impact factor is held constant at 0.2, the improved emission reduction impact factor changes slowly but significantly between 0.1 and 0.2 and changes slowly between 0.2 and 0.6. The three curved lines overlap between 0.6 and 0.7. This shows that with the improved emission reduction impact factor, the amount of CO2 pollution, and the PM2.5 concentration gradually decrease, and the air quality gradually improves, but the value should not exceed 0.6.

When only the technology to reduce the fuel consumption impact factor is adjusted from 0.2 to 1, the amount of CO2 pollution decreases by 9.84% and the air quality increases by 24.7%. When only the improved emission reduction impact factor is adjusted from 0.1 to 0.2, the amount of CO2 pollution decreases by 3.82% and the air quality increases by 12.1%.

This indicates that the energy-conservation and emission-reduction models are more sensitive to the technology chosen to reduce the fuel consumption impact factor than to the improved emission-reduction impact factor. However, neither model contributes to the smooth achievement of the carbon peak target, indicating that the models need to be improved.

According to the statistics for Beijing private car fuel consumption per 100 km, the average value was calculated. The modeled technological advances benefit energy efficiency and reduced gasoline consumption. The analysis shows that changing a single regulatory factor cannot achieve the carbon peak goal. Thus, the two regulatory factors of the technology policy were implemented together. The simulation results are shown in Fig. 8.

Fig. 8
figure 8

Sensitivity analysis for two regulatory factors

Full size image

In Fig. 8, CO2 denotes the amount of CO2 pollution, PM2.5 denotes the PM2.5 concentration, and fuel denotes the quantity of fuel used in private cars. It can be seen from Fig. 8 that the effect of the technology to reduce the fuel consumption impact factor is significant between 1 and 3, and the effect of the improved emission reduction impact factor is significant between 0.2 and 0.6. Therefore, in the remainder of this article, we mainly study the combined influence of these two regulatory factors on the main variables within the significant range, and the specific change trends are presented in Table 4.

Table 4 The influences of different regulatory factor values on the major variables (2035)
Full size table

As can be seen from Fig. 8a, when the technology to reduce the fuel consumption impact factor increases to > 2.5 and the improved emission reduction impact factor increases to > 0.5, the amount of CO2 pollution reaches the carbon peak target in 2033. The air quality also changes significantly when these regulatory factors are (2.5, 0.5). It can be seen from Table 4 that the air quality changes the most when the values increase from (2.25, 0.4) to (2.5, 0.45), resulting in an increase of 25.5%. The amount of CO2 pollution and the PM2.5 concentration also decreases by 32.1% and 6.04%, respectively, as the factors increase from (2.25, 0.4) to (2.5, 0.45). In this scenario, the technology to reduce the fuel consumption impact factor is between 1 and 3 and the improved emission reduction impact factor is between 0.2 and 0.6, which also indicates that the emission reduction effect is for this scenario. As can be seen from the quantity of fuel used in private cars (Fig. 8d), with the improvement of the technology policy, the fuel consumption decreases more and more. When the two regulatory factors change to (3, 0.6), the private car fuel consumption decreases by 6.98%. The fuel consumption of private cars also reaches the minimum under this scenario, which also shows that the stronger the implementation of the technology policy, the lower the fuel consumption will be. This has the dual effects of reducing carbon emissions and reducing the consumption of non-renewable resources.

In summary, it is appropriate to control the impact factor of the technological fuel consumption reduction at 2.5 and the impact factor of the improved emission-reduction at 0.5.

Impact of a combination policy

In this study, a driving-restriction policy was combined with a technology policy, and the amount of CO2 pollution, PM2.5 concentration, air quality, and quantity of fuel used in private cars were taken as variables to investigate the energy conservation and emission reduction effects of the combination policy. The driving-restriction policy, no-limit scenario, technology policy, and combination policy are labeled as scenarios 1, 2, 3, and 4, respectively, and the simulation results are shown in Fig. 9. The regulatory factors for the four different scenarios are listed in Table 5. The changes in the four major variables under the four different scenarios were compared. The specific data for 2020 and 2035 are presented in Table 6.

Fig. 9
figure 9

The influences of different scenarios on the main variables

Full size image
Table 5 Regulatory factors for the different scenarios
Full size table
Table 6 The influences of the different scenarios on the main variables
Full size table

As can be seen from Fig. 9a and b, the carbon peak target cannot be reached under the driving-restriction policy and no-limit scenarios. Scenario 4 is more conducive to the realization of the carbon peak goal (Fig. 9c and d). Compared with scenario 1, under scenarios 2–4, the amount of CO2 pollution is 0.49%, 39.11%, and 40.19% lower, respectively, in 2020 (Table 6). Compared with scenario 1, under scenarios 2–4, the amount of CO2 pollution is 1.02%, 52.29%, and 52.45% lower, respectively, in 2035, and the air quality improves by 0.01%, 130.01%, and 144.79%, respectively, in 2035. Compared with scenario 1, under scenarios 2–4, the quantity of fuel used in private cars is 0.7%, 28.92%, and 32.6% lower, respectively, in 2020. At the end of the simulation (in 2035), the quantity of fuel used in private cars is 0.12%, 29.43%, and 33.42% lower in scenarios 2–4, respectively, compared to scenario 1.

Therefore, the combination policy can not only effectively reduce CO2 and PM2.5 emissions and improve air quality, but it can also save energy and improve energy efficiency.

Discussion

Studies have shown that the driving-restriction policy has had no obvious effect on CO2 and PM2.5 emission reduction from private cars, which is consistent with the results of Chen et al. (2021b) and Zhang et al. (2020). Xiao et al. (2019) reported that the policy of scrapping old cars is more cost-effective than a driving-restriction policy. Zhang et al. (2021) suggested increasing investment in technological innovations in energy-intensive industries and deepening cooperation between provinces and cities in environmental governance to improve regional air quality. In order to optimize the driving-restriction policy model, we introduced a technology policy, and we combined the technology policy with the driving-restriction policy to explore the long-term effects.

Time lag of effect of technology policy

The technology policy has multiple effects; not only can it reduce environmental pollution, but it can also save non-renewable energy. Through the implementation of the technology policy, the amount of CO2 pollution can be reduced, the PM2.5 concentration can also be gradually reduced, and the air quality can be greatly improved. Private cars use much less fuel under the driving-restriction policy, which not only saves gasoline but also protects the environment by reducing air pollution caused by gasoline combustion. However, the implementation of the technology policy needs to introduce a large number of high-tech talents, and it takes a certain amount of time to study how to reduce the fuel consumption of private cars. This process may consume a lot of the human, material, and financial resources of the country. Therefore, it is necessary to consider other policies (such as administrative policies and fiscal policies) to compensate for the side effects caused by the time lag of the effect of the technology policy.

The previously presented sensitivity analysis of a single regulatory factor for the technology policy revealed that it is wrong to assume that larger regulatory factors are better. When the regulatory factors are increased to a certain extent, the influences on the major variables such as the amount of CO2 pollution, air quality, and quantity of fuel used in private cars do not increase, and they even decrease. According to the simulation analysis, when the regulatory factor of the technical fuel consumption reduction reaches 3 and the regulatory factor of the improved emission reduction reaches 0.6, increasing their values further has almost no effect on the main variables. Therefore, these two regulatory factors should not exceed 3 and 0.6, respectively. When the regulatory factor of technology for fuel consumption reduction is 2.25–2.5 and the regulatory factor for improved emission reduction is 0.45–0.5, the carbon stock in the air caused by private car emissions reaches its peak in 2035. This has to do with the time lag effect of the technology policy, which requires time to see how effective it is and a lot of time to figure out how to reduce fuel consumption in terms of technology and reduce the pollutants released by gasoline combustion. In the early stage of the simulation, the effect is not obvious, but with the passage of time, the effect gradually appears. This effect is the opposite of the effect of the driving-restriction policy, which can have an immediate effect on controlling the number of private car trips (Yang et al. 2018b) and reducing the pollutants released by private cars in the short-term, and therefore, it can ease traffic congestion (Gao and Hao 2021; Tang and Ruan 2018). However, in the long-term, it induces residents to buy a second car (Wang et al. 2014), causing greater environmental pollution.

Cumulative effect of the combination policy

The existing driving-restriction policy and technology policy were combined to form a combination policy, and the effects of the combination policy and driving-restriction policy were compared (Fig. 10). It can be clearly seen from Fig. 10 that the combination policy has significantly greater effects than the driving-restriction policy in terms of the amount of CO2 pollution, PM2.5 concentration, air quality, and quantity of fuel used in private cars. As can be seen from Table 6, under the combination policy, the amount of CO2 pollution is 40.19% and 52.45% lower in 2020 and 2035, respectively, compared with the driving-restriction policy; meanwhile, under the technology policy, the amount of CO2 pollution is 39.11% and 52.29% lower in 2020 and 2035, respectively, compared with the driving-restriction policy. It can be seen that the effect of the combination policy is better than that of the technology policy alone, and the combination of the technology policy and driving-restriction policy makes up for the shortcomings of the driving-restriction policy (Tu et al. 2021; Ma et al. 2020). Figure 10 also illustrates the time lag of the effect of the technology policy. Overall, the effect of the technology policy becomes stronger. This is contradictory to some scholars’ belief (Jiang et al. 2019) that the effect of the policy gradually weakens over time. It also shows that a technology policy is effective but not perfect, so it must be improved iteratively.

Fig. 10
figure 10

The cumulative effect of the combination policy

Full size image

Similarly, under the combination policy, the PM2.5 concentration and quantity of fuel used in private cars decrease by 21.98% and 33.4%, respectively, in 2035, while the air quality increases by 144.79%.

Conclusions

In order to study the effects of the driving-restriction policy, in this study, we used a system dynamics approach and took historical data for Beijing from 2010 to 2020 as an example to establish a system dynamics model of energy conservation and emission reduction. By analyzing the relationships between the internal and external factors of the system, the implementation effect of the driving-restriction policy was analyzed in the medium- and long-term, centering on five subsystems: the economic, government, environment, technology, and traffic subsystems. In addition, sensitivity analysis of a technology policy was conducted to quantitatively investigate the effects of a technology policy and the effects of policies with different implementation intensities on the amount of CO2 pollution, PM2.5 concentration, air quality, and quantity of fuel used in private cars. The conclusions of this study are as follows.

(1) The driving-restriction policy can reduce the number of private car trips in the short-term, but in the long-term, it fails to effectively reduce the CO2 emissions and air pollution from private cars. (2) Through sensitivity analysis of the two regulatory factors of the technology policy, it was found that when the regulatory factor of the technology to reduce fuel consumption is between 1 and 3 and the regulatory factor of improved emission reduction is between 0.2 and 0.6, the energy conservation and emission reduction effects are better. The simulation curves almost coincide when their values exceed 3 and 0.6, respectively. In the model of energy conservation and emission reduction, the regulatory factor of the technology to reduce fuel consumption is more sensitive than that for improved emission reduction. (3) When the regulatory factor of the technology to reduce fuel consumption and that for improved emission reduction are 2.5 and 0.5, respectively, the change is great and the carbon emission peak can be achieved. (4) Compared with the driving-restriction policy, the combined policy reduces the amount of CO2 pollution by about 52.45%, the PM2.5 concentration by 21.98%, and the quantity of fuel used in private cars by 33.42% in 2035, and it improves the air quality by about 144.79%. It is beneficial to achieving the carbon emission peak goal.

Based on these results, the following suggestions for energy conservation and emission reduction are proposed.

First, the existing driving-restriction policy should be improved since it is not suitable for long-term implementation. Energy conservation and emission reduction comprise a complex, multi-faceted, and long-term process that requires the joint efforts of all subsystems. Therefore, it is necessary to achieve the coordination between the implementation of policies to form a policy chain with twice the result with half the effort and obtain the synergistic effect of policy implementation. In view of this, the car driving-restriction policy can be improved from the perspective of publicity to improve urban residents’ awareness of the negative impact of the policy. Social publicity and guidance should be strengthened, and differentiated guidance strategies and measures should be adopted for different travel groups. For example, more people should take public transportation, the travel rate of private cars should be reduced, and the awareness of environmental protection among residents should be strengthened.

The second recommendation is to increase the development and progress of technology, to reduce energy consumption by reducing the fuel consumption of vehicles per 100 km, and to reduce the amount of CO2 released by the combustion of gasoline through filtration. The implementation of a technology policy can not only reduce energy consumption and save non-renewable resources, but it can also reduce CO2 emissions, protect the environment, and improve the air quality.

Finally, the government should dynamically adjust the factors and policy mix of each policy according to the process and target of pollution and carbon reduction. On one hand, there may be a phenomenon of diminishing effect margins with the passage of time, so there is a possibility that the prevention effect will gradually diminish or even be absent. On the other hand, due to the long time taken by research and development and preparation of a science and technology policy, the effect of the policy will be delayed. Therefore, when the policy is actually implemented, it is necessary to adjust the policy mix dynamically and moderately make a reasonable arrangement to achieve greater efficiency.

Owing to the limitation of data, whether Beijing’s research results are equally effective for energy conservation and emission reduction in other cities must be further studied horizontally. In addition, we only conducted research on the restriction policy for private cars in this study, and no specific research has been carried out for other models. Exploring these deficiencies is also the direction and focus of future research.