Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Transport Emissions Public Disclosure Authorized Analysis Methodology (TEAM) Supporting Measurement, Reporting, and Verification (MRV) and Results-Based Climate Finance John A. Rogers, Jing Xiong December 2025 * * * ## © 2025 The World Bank 1818 H Street NW, Washington DC 20433 ## Telephone: 202-473-1000; Internet: [www.worldbank.org](http://www.worldbank.org/) ## Some rights reserved. This work is a product of The World Bank. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy, completeness, or currency of the data included in this work and does not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, processes, or conclusions set forth. The boundaries, colors, denominations, links/footnotes, and other information shown in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. The citation of works authored by others does not mean The World Bank endorses the views expressed by those authors or the content of their works. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. ## Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Attribution—Please cite the work as follows: John A. Rogers and Jing Xiong. 2025. “Transport Emissions Analysis Methodology: Supporting Measurement, Reporting, and Verification (MRV) and Results-Based Climate Finance.” Washington DC: World Bank. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e mail: [pubrights@worldbank.org](mailto:pubrights@worldbank.org). Cover photo: © Boy Anthony/Shutterstock, © Firadoaw/Shutterstock, © Thongden Studio/FREEPIK. Further permission required for reuse. Cover design and text layout: Duina Reyes * * * # Table of Contents ### Acknowledgments ... 1 ### Executive Summary ... 2 ### Abbreviations and Acronyms ... 3 ## 1\. Background and Rationale ... 4 1.1 Methodology ... 4 ## 2\. Implementing the Model ... 6 2.1 Overview: A continually evolving model and methodology ... 6 2.2 Key steps in applying the model ... 7 ## 3\. Conducting the analysis ... 17 3.1 Calibration: Bottom-up to top-down ... 17 3.2 Preliminary analytical steps ... 21 3.3 Analytical framework ... 23 ## 4\. Applicability of Methodology ... 36 4.1 Determining changes in transport demand ... 36 4.2 Evaluating changes in policy variables ... 42 4.3 Carbon dioxide (CO2) emissions ... 49 i * * * Figures Figure 1. Modeling diagram .....8 Figure 2. Welcome page of the model .....9 Figure 3. Setup page for the WithPolicy scenario (Scenario 1) .....10 Figure 4. Setup page for the WithoutPolicy scenario (Scenario 2) .....10 Figure 5. Typical Gompertz curve .....13 Figure 6. Mortality curve .....13 Figure 7. TCAF approach to crediting baseline .....15 Figure 8. Process for calculating TCAF crediting baseline .....16 Figure 9. Basis of the bottom-up methodology .....17 Figure 10. Passenger cars and two-wheelers .....23 Figure 11. Mass passenger transport (buses and coaches) .....24 Figure 12. Freight transport (light commercial vehicles, medium- and heavy-duty trucks, and tractor-trailers) .....24 Figure 13. Efficiency and emissions (for all vehicles) .....25 Figure 14. Three-parameter Gompertz curve for car + motorcycle ownership Table s * * * # Acknowledgments The transport emissions analysis methodology and model were developed by John Allen Rogers (Consultant) and Jing Xiong (Senior Transport Specialist). This work has benefited from the important input, review, and feedback of World Bank colleagues: including Cecilia M. Briceno-Garmendia (Lead Economist), Xiaoyu Chang (Climate Finance Specialist), Simona Sulikova (Consultant), Maria Carolina Monsalve (Lead Economist), Fei Deng (Practice Manager), Bianca Bianchi Alves (Practice Manager), Joanna Moody (Transport Specialist), Jonathan Coony (Senior Climate Finance Specialist), Hongye Fan (Senior Urban Transport Specialist), Julian Gonzalo Jimenez (Sr. Climate Change Specialist), He He (Economist), Jason James Smith (Senior Climate Finance Specialist) and Nino Pkhikidze (Economist). It has also benefited from the important input, review, and feedback of expert consultants including Christian Brand (Professor in Transport Energy and Environment at the University of Oxford’s Transport Studies Unit and Environmental Change Institute). This work was carried out under the leadership of Nicolas Peltier (Global Director), Valerie Hickey (Global Director), Olivier Mahul (Practice Manager), Binyam Reja (Practice Manager), Said Dahdah (Acting Practice Manager), and Erwin De Nys (Practice Manager). The work was funded by the Transformative Carbon Asset Facility (TCAF), a World Bank trust fund that supports countries’ efforts to implement market-based carbon pricing and to create conditions for private sector investments in low-carbon technologies. The team would like to thank Nuyi Tao (Senior Climate Finance Specialist), TCAF Grant Manager, for her vision and continued support. Finally, the report was edited by Fayre Makeig with graphic design by Duina Reyes, under the leadership of Jonathan Davidar (Senior Knowledge Management and Learning Officer), Faustina Chande (Program Assistant) provided tremendous administrative support. * * * Transport Emissions Analysis Methodology (TEAM) # Executive Summary Reducing greenhouse gas emissions from transport is essential for achieving national climate targets, yet the sector has lacked practical and credible tools to quantify the emissions impacts of policies in a way that supports monitoring, reporting, verification (MRV), and results-based climate finance. This report responds to that gap by presenting a robust, MRV-compatible methodology for analyzing transport emissions. The report introduces a scenario-based transport emissions analysis methodology (TEAM) and model that quantifies emission reductions by comparing a measured WithPolicy scenario against a conservative and credible WithoutPolicy counterfactual. The methodology supports analysis both ex post, using observed data to measure achieved emission reductions, and ex ante, to assess the mitigation potential of proposed policies. The current application focuses on road transport—the largest source of transport emissions—while allowing for future expansion to other transport modes. The analytical framework integrates vehicle fleet dynamics, transport demand, energy consumption, and emissions factors through a bottom-up approach calibrated to national fuel consumption data. This calibration ensures consistency with energy balances and strengthens the credibility of results for MRV and carbon-crediting applications. The model captures key drivers of emissions, including vehicle ownership, usage patterns, technology mix, and fuel types, while accounting for broader economic and demographic trends. The methodology is designed to assess a wide range of transport policies, including vehicle efficiency and emissions standards, restrictions on used-vehicle imports, fuel pricing and subsidy reforms, vehicle scrappage programs, incentives for low- and zero-emission vehicles, and policies that promote modal shift and freight efficiency. It explicitly addresses additionality, baseline setting, leakage, and uncertainty, applying conservative assumptions where needed to safeguard environmental integrity. By providing a transparent and scalable analytical tool, the methodology enables governments and development partners to strengthen transport MRV systems, prioritize high-impact mitigation policies, and improve access to results-based climate finance and carbon market mechanisms. * * * Abbreviations and Acronyms | BAU | Business as Asual | | --- | --- | | BEV | Battery Electric Vehicle | | CCDR | Country Climate and Development Report | | CO2 | Carbon Dioxide | | CO2e | Carbon Dioxide Equivalent | | EU | European Union | | Euro 1 to 6, I to VI | European New Vehicle Emission Standards | | EV | Electric Vehicle | | g/kmCO2 | Grams Of Carbon Dioxide Emissions Per Kilometer | | GDP | Gross Domestic Product | | GHG | Greenhouse Gas | | Gt | Gigaton | | GtCO2e | Gigatons Of Carbon Dioxide Equivalent | | HH | Household | | IPCC | Intergovernmental Panel on Climate Change | | km | Kilometer | | MJ | Megajoule | | MMHE | Mean Monthly Household Expenditure | | MRV | Monitoring, Reporting, And Verification | | NDC | Nationally Determined Contribution | | PHEV | Plug-In Hybrid Electric Vehicle | | PM | Particulate Matter | | SUV | Sports Utility Vehicle | | TCAF | Transformative Carbon Asset Facility | | TRACE | Transport Research and Costing in Europe | \\mathrm{C O\_{2}} \\mathrm{C O\_{2}e} \\mathrm{G t C O\_\_{\\rho}e} * * * Transport Emissions Analysis Methodology (TEAM) ## Background # 1\. and Rationale The analysis presented here is against a backdrop of rapidly growing transport-related emissions on a regional as well as global scale, requiring urgent attention from policy makers. Worldwide, transport emissions in 2024 reached 8.4 gigatons of carbon dioxide equivalent (GtCO₂e). Of this, road sector emissions were just over 6 GtCO₂e, 8 percent higher than in 2015. Emission growth averaged only 0.2 percent annually from 2019 to 2024, down from 1.7 percent per year between 2015 and 2019. Regionally, transport-related emissions are increasing more rapidly in developing economies than in Europe or North America—a trend that is likely to continue in the coming decades. Delivering on the updated Nationally Determined Contributions (NDCs) by 2030 would bring about a decline in emissions. However, the implied global emissions by 2030 still exceed, by a large margin, the emissions on pathways consistent with the Paris Agreement. Available evidence suggests that achieving NDC targets would require limiting transport-related CO2 emissions to about 2–3 Gt by 2050. On a global scale, there is a need for transport to match the ambition of other sectors, both to mitigate greenhouse gas (GHG) emissions and to maintain clear monitoring and documentation of the (potential) impacts of emissions. Meanwhile, a lack of methodological tools has hindered progress toward ambitions by making carbon financing a challenge for transport projects. In particular, there is an urgent need for monitoring, reporting, and verification (MRV)-compatible methods of evaluating, analyzing, and measuring the impacts of specific policies and interventions (both before and after implementation) that demonstrate benefits in terms of GHG emission reductions. ### 1.1 Methodology The present report outlines an MRV-compatible method of evaluating, analyzing, and measuring the impacts of specific transport policies and interventions that deliver GHG emission reductions. The method may be applied either before implementation (ex ante), using projected estimates, or after implementation (ex post). The current version of the methodology evaluates GHG emission reduction due to a shift in vehicle usage and fuel types in the largest contributing subsector: road transport. The methodology may be used to analyze: - GHG emission reductions due to a specific intervention, following implementation and using measurable variables. This, ex post, evaluation method is the focus of the report. * * * 1. Background and Rationale - The potential impacts of different future policy scenarios on GHG emissions. This ex ante analysis can help select policies for evaluation in various countries by identifying their mitigation potential. The analytical process is scenario based. A measured WithPolicy scenario (what happened) is compared against a modeled, conservative, and credible WithoutPolicy counterfactual scenario, which evaluates what would have happened had the policy not been enacted. # Need for ex post analysis Ex post, the magnitude of change in transport sector emissions can be calculated using historical data and compared to a WithoutPolicy counterfactual scenario. The Bank’s lending operations—including investment project financing and development policy financing engagements—provide a substantial market for the evaluation of emission reductions due to vehicle and fuel policies. This analysis will, in turn, expand transport projects’ access to results-based carbon financing and similar mechanisms. # Need for ex ante analysis The methodology may also be used to estimate the GHG emission reductions that could be achieved by policy changes yet to be enacted. One ex ante application is for the quantitative analysis of road vehicle and fuel interventions as required by Country Climate and Development Reports (CCDRs). The World Bank is investing in these country-level analytics to help identify the right policies, plans, and timelines for climate action specific to each country’s local context and development challenges—and transport is recognized as one of the five critical sectors for a low-carbon transition. Across the 93 CCDRs prepared as of November 2025, 1 improving transport infrastructure is highlighted as necessary for reducing GHG emissions, connecting people with jobs and services, improving air quality, and increasing productivity and competitiveness through better logistics. The CCDRs for countries in East and West Africa, including Ghana, the G5 Sahel countries, 2 Angola, and Malawi, mention policies pertaining to imports of used vehicles (including fuel efficiency standards and age restrictions) and motorization management practices (including scrappage) for passenger cars. In East Asia and Pacific and also Latin America and the Caribbean, mitigation scenarios include fleet renewal for road freight trucks, improvements in logistics operations and cargo loading, incentive programs for alternative fuel vehicles (including e-mobility), and adoption of personal cars. Across the Middle East and North Africa, and Europe and Central Asia, there is an additional focus on national fuel standards and fossil fuel pricing reform. The methodology presented here could be a valuable resource for current and upcoming CCDRs in identifying the most impactful vehicle and fuel policies and for articulating how they interact with other economic and sustainable development goals. See World Bank, “Country Climate and Development Reports (CCDRs),” [https://www.worldbank.org/en/publication/country-](https://www.worldbank.org/en/publication/country-) climate-development-reports. Burkina Faso, Chad, Mali, Mauritania, and Niger. * * * Transport Emissions Analysis Methodology (TEAM) ## Implementing # 2\. the Model The modeling approach used to quantify emission reductions examines the effects of policies and regulations that modify the use of road transport, transport modes, technology employed, and the adoption of these technologies by end users, as well as energy demand. The model can estimate potential emission reductions from policies expected to be adopted, as well as those whose implementation is complete. The transport sector is dynamic. There are many simultaneous interactions within the sector as well as other changes that affect transport activity and emissions (e.g., growth of gross domestic product \[GDP\], rural to urban migration, infrastructure development, and changes in vehicle fleets). A before- and-after comparison of business as usual (BAU) and specific projects cannot isolate and identify the impacts of specific policies. The analytical process described herein uses an ex post, measured WithPolicy scenario (what happened) and a modeled, conservative, and credible counterfactual WithoutPolicy scenario. The WithPolicy scenario is compared to the WithoutPolicy scenario. The scenarios will differ only in the implementation of specific policies or interventions. Thus, all activities that alter emission intensity and are outside the scope of an evaluated intervention, are included in the _WithProject measured scenario and directly transferred to the WithoutProject counterfactual scenario._ The comparison of the WithPolicy and WithoutPolicy scenarios quantifies the emission reductions that can be attributed to a policy’s enaction. The methodology is primarily meant for measuring, reporting, and verifying emission reductions achieved by recently adopted policies. ### 2.1 Overview: A continually evolving model and methodology Like most analytical models, the model used for the present analysis and its methodology is continually evolving. So far, the model includes all road transport (e.g., passenger cars, motorcycles, three- wheelers, light-duty commercial vehicles, heavy-duty trucks, and buses). In its next stage, the model will incorporate rail (metro, light rail, suburban, and long distance) for passengers and freight. The model includes an analytical framework to evaluate policy actions such as the choice of standards for vehicle technology, emissions, and fuel consumption; the incorporation of electric vehicles (EVs); * * * 2. Implementing the Model the evaluation of the enforcement (or tightening) of restrictions on the import of used vehicles; the removal of fuel subsidies; and the implementation of vehicle scrappage programs. Other policy choices will be added to the framework as the demand arises. The modal choice for passengers and freight between road transport modes and rail will be further detailed. Subsequent to this, adding waterborne (inland waterways and coastal) and air (regional and national) transport modes to the model could enable accounting for consumers’ choices of other passenger and freight travel modes. This will inform dialogue on improvements at the level of transport systems and how these improvements impact GHG emissions at the country or subnational level. This allows an MRV-based evaluation of achievements that feed into distinct carbon crediting programs. # 2.2 Key steps in applying the model ## Define the modeling boundary Geographical A first step in applying the model is to set the modeling boundary. For countrywide policies, national borders can serve as the boundary. However, if a policy impacts only part of a territory—for example, vehicle activity in the metropolitan area of the capital city—a more limited area may be chosen. However, this will, in many cases, bring up issues of transboundary leakage. For example, commuters residing outside the area may contribute significantly to transport activity within the metropolitan area; the same goes for freight trucks making a drop-off. Leakage must be dealt with by redefining boundaries, or estimated through an applied discount, where necessary. ### Time period The model is set up for conducting scenario-based analysis over a period of up to 31 years, with the first year being between 2010 and 2030. The analysis period must be divided into three parts: - A historical period, from the base year, when the policy package was enacted, up to the most recent year for which data are readily available. - A period before this base year, which will be used for evaluating tendencies before the policy package was enacted. - A future period after the year for which most recent data are available to forecast future impacts of the enacted policy package over a crediting period. For example, suppose a policy package was enacted in 2017, the year for which most recent data are available is 2020, and the proposed crediting period is through 2030. The first year of the modeling framework could be 2010. This would give 2010–20 as the period for which historical data are available, and a forecasting period through 2040. The analysis from 2017 to 2020 would be results based (ex post), whereas from 2021 to 2030, the future value of policy actions would be forecast. * * * Transport Emissions Analysis Methodology (TEAM) For results-based crediting, new data are introduced each year into the model (with a time lag dependent on data availability). Thus, for example, in 2023, data for 2021 may become available. These data are added to or updated in the model to calculate the emission reductions resulting ex post due to the policy package for 2021. In this way, the model records the outcomes of the MRV program annually. Based on this new information, the model also recalculates the forecast period (to 2040). The model will output data and results over any period that the user selects within these 31 years. # Build the modeling diagram Once the modeling boundary has been selected, historical data are added in the tool for the _WithPolicy operation and the tool is calibrated. As shown in figure 1, which illustrates the modeling_ diagram, there are two types of inputs: (1) scenario-independent inputs, which define the transport system within this framework; and (2) scenario-specific inputs in two groups, which distinguish the _WithoutPolicy operation from the calculated WithPolicy operation. Specific data requirements are_ outlined in a later section. The first column of boxes in figure 1 (“Control Modules”) indicates where each type of information is entered. Figure 1. Modeling diagram _Note: CNG = compressed natural gas; LPG = liquefied petroleum gas; PHEV = plug-in hybrid electric vehicle; W = wheeler._ ## Define scenario-independent inputs The scenario-independent inputs should ideally be publicly available data, so the tool can be used to share data and findings and build consensus across different groups of stakeholders. These data are stored on the model’s “Library” and “Collection” pages. For data that have to be updated at a later date—via the MRV process mentioned earlier—the tool includes separate sheets for each data * * * 2. Implementing the Model type. The updated data (and the reference sources cited for them) can be added to these sheets without affecting the initial load. This ensures data integrity and preserves a data chain of custody, 3 giving a complete and fully documented step-by-step history of the data and of the model users who have modified them. For future years, the assumptions applicable to the Define scenario-specific inputs _WithPolicy operation and the WithoutPolicy_ counterfactual are defined and used as the basis for analysis. The scenario-specific inputs are a mix of historical, measured data that define the WithPolicy drivers, and data and assumptions that will be used to creatively describe the WithoutPolicy drivers of the analysis. ## Understand the model layout For future years, the assumptions applicable to the WithPolicy operation and the WithoutPolicy The model counterfactual are defined and used as the basis for analysis. can be easily customized to specific analyses. It does not require enabling Microsoft Excel’s default functionality and can be navigated using the built-in buttons and links. The model contains one Microsoft Excel workbook with (currently) 53 worksheets. All the sheets that are specific to a scenario # Understand the model layout end in “1” or “2” (the WithPolicy scenario or the WithoutPolicy scenario, respectively). The model can be easily customized to specific analyses. It does not require enabling Microsoft Excel’s Model navigation is based on three sheets.default functionality and can be navigated using the built-in buttons and links. The model contains oneThe model’s welcome page (Error! Reference source not **found.** Microsoft Excel workbook with (curren ) gives easy access to setting up the tly) 53 worksheets. All the sheets that are specific to a scenario _WithPolicy and WithoutPolicy scenarios, running the_ end in “1” or “2” (the WithPolicy scenario or the WithoutPolicy scenario, respectively). calculations, and accessing the results. Model navigation is based on three sheets. The model’s welcome page (figure 2) gives easy access to The setup setting up the pages for the _WithPolicy_ _WithPolicy_ and WithoutPolicy scenario scenarios, running the calculations, and accessing the results. (Scenario 1) (Error! Reference source not found.) and the _WithoutPolicy scenario (Scenario 2) (Error! Reference source not found.) provide links to all necessary_ sheets. The setup pages for the WithPolicy scenario (Scenario 1) (figure 3) and the WithoutPolicy scenario (Scenario 2) (figure 4) provide links to all necessary sheets. **Figure 2. Welcome page of the model** Figure 2. Welcome page of the model _Note: CNG = compressed natural gas; ER = emission reduction; LPG = liquefied petroleum gas; MRV = monitoring, reporting, and_ verification _Note:_ verification; tCO CNG = compressed natural gas; ER = emission reduction; LPG = liquefied petroleum gas; MRV = monitoring, reporting, and ; tCO2 = tons of carbon dioxide = tons of carbon dioxide. . A process that tracks the movement of evidence through the collection, safeguarding, and analysis stages of its life cycle by documenting each person who handled the evidence, the date/time it was collected or transferred, and the purpose for the transfer. ### pg. 9 * * * Transport Emissions Analysis Methodology (TEAM) Figure 3. **Figure 3** Setup page for the **. Setup page for the** WithPolicy _**WithPolicy**_ scenario (Scenario 1) **scenario (Scenario 1)** Figure 4. **Figure 3. Setup page for the WithPolicy scenario (Scenario 1)** Setup page for the WithoutPolicy scenario (Scenario 2) **Figure 4. Setup page for the _WithoutPolicy_ scenario (Scenario 2)** All sheets can be easily accessed via the “INDEX” sheet, which indicates what sector or scenario each **Figure 4. Setup page for the WithoutPolicy scenario (Scenario 2)** sheet applies to and whether a sheet contains user input data. All sheets can be easily accessed via the “INDEX” sheet, which indicates what sector or scenario each sheet applies to and whether a sheet contains user input data. Initial data are stored in the “Library” and “Collection” sheets. Each variable has a specific sheet so that the user can easily update numbers from one year to the next (in accordance with their MRV Initial data are stored in the “Library” and “Collection” sheets. Each variable has a specific sheet so that program) and cite the source of each update, to maintain adequate control. These variable-specific the user can easily update numbers from one year to the next (in accordance with their MRV program) and cite the source of each update, to maintain adequate control. These variable-specific sheets also sheets also make it easy to use the model to evaluate the likely impact of different policy packages. make it easy to use the model to evaluate the likely impact of different policy packages. The model has results sheets and charts that the user can easily adjust to any output period (while the model calculates from 2010 to 2040, the user can select different subperiods within this range pg. 10 \[e.g., 2017–27\] for reports in the “Publish” sheet). All sheets can be easily accessed via the “INDEX” sheet, which indicates what sector or scenario each sheet applies to and whether a sheet contains user input data. Initial data are stored in the “Library” and “Collection” sheets. Each variable has a specific sheet so that the user can easily update numbers from one year to the next (in accordance with their MRV program) and cite the source of each update, to maintain adequate control. These variable-specific sheets also make it easy to use the model to evaluate the likely impact of different policy packages. pg. 10 * * * Analyze impact channels The methodology and model are structured to evaluate different policy and regulatory impact channels, such as: • Regulation of vehicles, for example, fuel efficiency standards, age restrictions, and technology standards applied to imports of new and used vehicles. • Fossil fuel pricing reform (including but not limited to subsidy removal) and carbon taxes. • Monetary or nonmonetary incentives for the purchase and use of newer, cleaner vehicles, including vehicle scrappage programs. • Pricing of vehicle ownership. • Pricing of vehicle use, for example, road pricing. • Accelerated adoption of low-emission vehicles and clean transport modes. • Improvements in infrastructure and services as alternatives to road-based travel to reduce the need to travel and encourage a modal shift, including investments in public transit and active mobility. Typical policy actions that may be included in the model are listed in table 1. Table 1. Examples of policy actions that are considered in the analysis | Policy | Type of policy | Type of intervention | | | | --- | --- | --- | --- | --- | | Avoid | Shift | Improve | | | | Efficiency and emissions standards for new and imported secondhand vehicles (including import restrictions) | regulation | | | X | | Strict efficiency and emissions standards for new vehicles (including phase-out of new ICE vehicles) | regulation | | | X | | Vehicle feebate system based on emissions | pricing | | | X | | Vehicle special sales tax | pricing | | | X | | Vehicle quota systems | regulation, pricing | X | X | X | | Reduction in maximum age of in-use vehicles (including scrappage programs) | regulation, pricing | | | X | | Vehicle registration fees based on emissions(includingCO2) | pricing | | | X | | Cross-funded subsidies that promote the purchase of zero-direct-emission noncommercial vehicles | pricing | | | X | | Incentives that promote the use of zero-direct-emission vehicles in commercial vehicle fleets | regulation, pricing | | | X | | Pricing policies that favor usage of zero-direct-emission noncommercial vehicles | pricing | | | | | Vehicle usage permits based on emissions(includingCO2) | regulation, pricing | | X | X | | Modal shift to mass transport | regulation | | X | | | Usage of larger capacity trucks | regulation, pricing | | X | X | | Distance-based vehicle insurance and registration fees | pricing | | | X | | Congestion or low-emission zone charging | pricing | X | X | | | Parking restrictions and pricing | pricing | X | X | | \\mathrm{O\_{2})} \\mathrm{C{}\_{2})} Note: CO2 = carbon dioxide; ICE = internal combustion engine. \\mathrm{C O\_{2}} * * * Example of an impact channel An impact channel is a pathway between an intervention and its outcomes (and, eventually, its impacts). The example below illustrates how an impact channel is analyzed. A European Union (EU) regulation entered into effect on January 1, 2020, which sets fuel efficiency and emissions standards for new and imported secondhand vehicles (the standard includes a regulation on import age restrictions). This policy package would have two impact channels, outlined as follows. WithPolicy scenario 1. New vehicles. All new vehicles sold after January 1, 2020, must meet the EU 2015 Euro 6 emission standard and the EU 2020 CO2 emission standard (95 percent compliance in 2020 with an emission standard of 95 grams of carbon dioxide emissions per kilometer \[g/km CO2\] and 100 percent compliance from 2022 onward). Motorcycles would be expected to comply with the Euro 5 standard. In future years, these standards will evolve to keep pace with EU standards. 2020,\\mathrm{C O}\_{2} \\mathrm{C{}\_\_{2}\]} 2. Imported secondhand vehicles. All imported secondhand vehicles sold after January 1, 2020, must meet the EU 2015 Euro 6 emission standard and the EU 2015 CO2 emission standard (130 g/km CO2). In future years, these standards will change to maintain a gap of five years with EU standards for new vehicles. 9/\ \\mathrm{k m},\\mathrm{C O}\_{2}) WithoutPolicy scenario Continuation of the previous policy (which, in this case, we are assuming to be): 1. A requirement for all new vehicles to meet Euro 3 emission standards, though there are no CO2 standards. \\mathrm{C O\_{2}} 2. A requirement for all imported used vehicles to be under 15 years old, though there are no emissions or CO2 requirements. \\mathrm{C O\_{2}} Note that in both cases, the regulation does not affect vehicles that are already in use. It only affects the new and secondhand vehicles entering the fleet each year, which slowly modifies the fleet mix every year based on (1) the vehicles in active service, (2) those leaving the active fleet each year, and (3) new and used vehicles entering the fleet each year. The resale of vehicles within the analytical boundary of the model does not affect the total active population: only new and used vehicles entering the fleet from outside the boundary impact the total active fleet. The analysis must consider the used vehicles entering the market that may comply with standards different from those for new original equipment manufacturer sales. * * * Logic flow For the WithPolicy scenario Apply the new policy starting January 1, 2020. The outcome can be measured. Future years are estimated to be able to assess the overall value of the program, but these estimates are not used for crediting purposes. For the WithoutPolicy scenario Suppose it can be shown that the new policy package increased the sales price of new or used vehicles. Then the demand for passenger cars in the WithoutPolicy scenario will be higher than that measured in the WithPolicy scenario. This will increase annual sales volume, and the distribution of used car imports by model year will change. For private motorized transport (passenger cars and motorcycles), the calculations for the in-use fleet and sales are carried out in two parts (for each vehicle category and segment): Vehicle population = Vehicles owned by households + vehicles owned by ministries, corporations, and others: • Vehicles owned by households. This is derived from data from household (HH) expenditure surveys (cars and motorcycles per thousand population vs HH income). Any price change shifts the curve down, increasing ownership in the WithoutPolicy scenario. • Vehicles owned by ministries, corporations, and others. The conservative estimate is inelastic but may vary from country to country (based on change in GDP, for example) The profiles of vehicles entering the fleet will follow the prior logic outlined in the WithoutPolicy scenario. New vehicles with the current year as the model year continue to join the fleet, but local emission standards are more flexible and CO standards are less stringent. 2 Older used vehicles enter the market meeting age law, only with higher local and CO emissions. 2 Figure 5. Typical Gompertz curve * * * Transport Emissions Analysis Methodology (TEAM) 2. Divide up by previous years inventory (segment and category) 3. Subtract from this year’s inventory (segment and category) Note that the average life of a vehicle is defined as the time point at which 50 percent of the original 4. Gives sales (this year inventory – previous inventory + mortality) (segment and category) number of vehicles are no longer in service. For example, an average life of 15 years means that of Where the 100 vehicles in service in year 0, only half (50) are likely to be in service at year 15 and very few at year registration database is not sufficiently detailed or up to date, vehicle sales data can be found for the previous 30. Buying a vehicle that has been serviced earlier does not extend this total life. So, importing older 10 to 20 years and used through a mortality function to reconstruct an estimated fleet profile using the logic of vehicles (in the Equation 1. In other cases, a survey needs to be conducted _WithoutPolicy scenario) will increase future scrappage and replacement volumes to_ . A quick and easy way of doing this is via the “Gas station Survey stabilize the overall population. ” documented separately. _Equation 1_ **Equation 1** 𝑃𝑃𝑃𝑄𝑃𝑅𝑃𝑄𝑃𝑄𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑄𝑃𝑃!"#$ &= 𝑃𝑃𝑃𝑄𝑃𝑅𝑃𝑄𝑃𝑄𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑄𝑃𝑃!"#$ '− 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑇!"#$ '\+ 𝑣𝑣𝑣𝑣 ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑖𝑖 \_𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 \_𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓!"#$ ' ## Where in each category and segment: Population yearx = The total in-use fleet of vehicles in year X. Where: Population Scrappageyearx = the total inyearx= The number of vehicles leaving the in-use fleet in year X, due to scrappage-use fleet of vehicles in year X **Scrappage** or disuse due to issues related to maintenance, parts, and reliability; or due to the vehicles **yearx** = the number of vehicles that exit the in-use fleet in year X due to scrappage, disuse due to maintenance, parts, and reliability issues, or vehicles being sold outside of the analytical boundary (to being sold outside the analytical boundary (to another state, for example). another state or country for example). _**vehicles**_ vehicles \_entering\_fleet _**\_entering\_fleet yearx**_ = The number ofyearx = The number of vehicles entering the in-use fleet within the analytical vehicles entering the in-use fleet within the analytical boundary in year X. These will include new boundary in year X. These will include new vehicles sold through the vehicle manufacturer’s vehicles sold through the vehicle manufacturer’s dealers, new and used vehicles imported from outside of the analytical boundary (bought in another state for dealers, and new and used vehicles imported from outside of the analytical boundary (bought example). in another state, for example). Often the registration database Suppose the types of vehicles demanded by users change over time (e.g., as users demand more does not have all the required information or some of the fields are free capture resulting in the same data being written in many different form mini and small cars in India, medium and luxury cars in China, or more sports utility vehicles \[SUVs\] s which complicates the analysis. Here the use of a worldwide database (such as: [https://car2db.com/excel](https://car2db.com/excel)) can facilitate filtering the registration database and in the Philippines), the historical data used to define the WithPolicy scenario will reflect this change including needed data that is absent (such as vehicle curb weight). and they will be transferred to the WithoutPolicy scenario. Mortality Function However, suppose the WithPolicy scenario shows a shift to EVs, for example; the WithoutPolicy scenario7 The preferred function to estimate vehicle mortality is the Winfrey S3 Curve 77 will likely not have this shift or will have a slower change because there is no fuel efficiency regulation, B07 and the expected vehicle technology is, for example, only Euro 3. y7 7 7 7 The average kilometers traveled per year by a vehicle varies by the vehicle’s age, though this is 7 7 more a function of the vehicle owner than of the vehicle itself (presuming that persons with higher _Equation 2_ disposable income tend to have higher mobility and own newer vehicles). The reduction in the average 𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = max (0, 𝑌𝑌 ∗ (1 − ((𝑌𝑌 − 𝑊𝑊)^2/𝐴𝐴𝐴𝐴 ^2))^𝑚𝑚)) ' age of imported vehicles due to the regulation is therefore unlikely to have a primary impact on the kilometers traveled per year. However, EV adoption may have an effect if users are reluctant to use them on longer trips but prefer them for shorter trips. 𝑖𝑖𝑃𝑄𝑃𝑃𝑆𝑆𝑃𝑃 𝑃𝑃𝑆𝑆 𝑃𝑄𝑌𝑌𝑆𝑆𝑆𝑆𝑆𝑆 𝑃𝑅𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑃𝑃 𝑃𝑅 7Thus, the modeled WithoutPolicy scenario could have a more extensive fleet than the WithPolicy scenario (defined in \[1\] above), though with annual mileages that may be subject to adjustment and e shorter in-use active life (defined in \[2\] above). The class mix of new and used vehicles entering the fleet in the WithoutPolicy scenario would be similar to that under the WithPolicy scenario. Still, the technology (and corresponding fuel economy) of the new and used vehicles entering the fleet in the _WithoutPolicy scenario could be older (and lower)._ With these data and assumptions, vehicle activity, fuel consumption, and emission for each vehicle type, technology, and age will be calculated for each scenario. The differences between the two scenarios would be due to the impact channels analyzed. * * * 2. Implementing the Model # Account for additionality in calculations 100% 100% GHG emission reductions can be counted only if they are additional. That is, if they would not have occurred without the policy or measure being enacted. Had reductions occurred anyway, without80% 80% causality, then they are not additional. 60% 60% Also, country signatories of the Paris Agreement formally agreed to reduce their GHG emissions and strengthen their commitment over time. Most developed, high-income countries committed themselves 40% 40% to an emission reduction target in absolute terms over an earlier year. Developing countries typically commit to an unconditional target or emission intensity they can achieve on their own, and offer to meet a stricter goal conditional 20% on receiving international technical or financial assistance. Emission 20% reductions required to meet NDC targets but committed to as part of the unconditional pledge would not be considered additional and should be part of the baseline.0% 0% 0 5 10 15 20 25 30 35 40 Since emission reduction units that have been sold cannot be applied to future more ambitious commitments, the crediting baseline should be more conservative than the unconditional NDC target. This is to ensure a high level of environmental integrity and compensate for uncertainties in the process of determining and calculating emission reductions (see figure 7). Figure 7. TCAF approach to crediting baseline BAU emissions Unconditional NDC target emissions TCAF crediting baseline Program emissions Emissions Mt CO2 Total emission reductions 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 _Note: BAU = business as usual; NDC = Nationally Determined Contribution; TCAF = Transformative Carbon Asset Facility._ If any other project-based emission reductions are sold (e.g., from a project-based mechanism such as Article 6.4 under the Paris Agreement), then these would have to be subtracted from the emission reductions credited to the program. In practical terms, this means the Transformative Carbon Asset Facility (TCAF) will make payments against a crediting threshold (or “TCAF baseline”) that could be below the BAU emission trajectory and typically below the target emission trajectory. Single-year targets will conservatively be broken down to a trajectory of annual targets, with the default being linear interpolation to the 2030 goal. * * * Transport Emissions Analysis Methodology (TEAM) Setting a conservative TCAF baseline may require applying a case- and country-specific discount to the calculated emission reduction (figure 8). Figure 8. Process for calculating TCAF crediting baseline **NDC emissions** **Min of NDC &** **BAU emissions** **TCAF crediting** **BAU emissions** **baseline** **Discount** _Note: BAU = business as usual; Min = minimum; NDC = Nationally Determined Contribution; TCAF = Transformative Carbon Asset Facility._ It may be necessary to define and agree on three scenarios—BAU emissions, NDC unconditional emissions, and the crediting threshold—and calculate the emission reduction between the WithPolicy scenario and each of these three scenarios. Suppose another co-financing mechanism is involved or policy implementation receives financial support from other sources (e.g., it is included in a policy lending operation), the calculated emission reduction must be attributed to the funding sources or partners. This attribution would be processed outside the model based on the model’s results. The difficulty in determining additionality is that activities reducing GHG emissions occur all the time. Vehicle fleets continually grow, and vehicles become more efficient as technology evolves. Changes in passenger and freight travel choices are driven by convenience, cost saving, and time saving, with little concern for GHG emissions. Other shifts occur due to structural social and economic changes or other technological and good-practice changes driven by regulatory, policy, and market forces. This precludes trying to evaluate GHG emission reductions by comparing the emission intensity at one point in time against a baseline emission intensity at a previous point in time. Over the intervening period, there are likely to be many different activities that change emissions intensity and are beyond the control of the policy or other measure being evaluated. An analysis of the barriers impeding the implementation of specific policies and measures can help prove the additionality of GHG emission reductions. These barriers may include: - Local traditions. Lack of knowledge of practices, laws, customs, and market conditions may have prevented a more GHG-efficient scenario from being implemented. - Social barriers. Societal actors may exert pressure to continue the status quo. - Financial barriers. Implementation of specific policies and measures may require funding that would not be available without demonstrating a reduction in GHG emissions. This may include an investment feasibility analysis to determine whether a given policy or intervention is most likely to attract funding. - Institutional and political barriers include lack of enforcement or political support for implementing policies and measures to reduce GHG emissions (e.g., the NDCs to the Paris Agreement). - Technological barriers. A sector may lack access to the materials, equipment, or infrastructure necessary to implement a specific policy or measure. So, if this policy or measure were not to be enacted, the BAU scenario would result in higher emissions. * * * Conducting 3\. the analysis Vehicular CO emissions cannot be measured directly. They are calculated based on vehicle activity. 2 Different vehicles (e.g., small, medium, and large passenger cars) have different fuel economy and emission characteristics, that must be accounted for. The basis of this bottom-up methodology is shown in figure 9. Figure 9. Basis of the bottom-up methodology Note: g CO2 = grams of carbon dioxide; km = kilometer; MJ = megajoule; yr = year. \\mathrm{C O}\_{2} Where: • Number of vehicles is subdivided by category, segment, fuel and technology, and emission standard. • Usage. Average travel speed and annual mileage are in three usage bins (urban, rural, and highway). • Energy consumption factors for each combination of vehicle type are calculated using COPERT 5.7.3. 3.1 Calibration: Bottom-up to top-down • Emission factors in gigatons of carbon dioxide (GtCO2) per megajoule (MJ) for each fuel are calculated using local values or default values from the Intergovernmental Panel on Climate Change (IPCC). * * * Transport Emissions Analysis Methodology (TEAM) Applicable international emission factors from the COPERT database are adjusted. While the need for this adjustment could be due to operating and maintenance practices in the countries in question, it is also likely due to needed fine-tuning of the annual vehicle use (kilometers/year) estimates or slight differences in the vehicle urban/rural/highway speed profiles. # Number of vehicles Since the MRV analysis is ex post, based on actual results, the number of vehicles sold annually within the modeling boundary for each category and vehicle segment in the WithPolicy scenario needs to be recorded (see table 2 for a list of common categories). The number of vehicles sold annually is calculated for the WithoutPolicy (baseline) scenario also. This calculation will depend on the specific policies being enacted. If policy entails the use of higher-capacity trucks, then the WithoutPolicy scenario (with smaller trucks) would need a more significant number of vehicles to have the same carrying capacity as the WithPolicy scenario. If policy introduces the sales of EVs instead of internal combustion engine alternatives, then the essential substitution would be one-for-one. Suppose the policy change entails a slight difference in end-user vehicle cost between the two scenarios, in that case, an agreement must be reached on the demand price elasticity of sales to be applied. If the WithoutPolicy scenario results in higher-cost vehicles, then this may result in lower uptake of vehicles. Conversely, if vehicles are less costly to users under the WithoutPolicy scenario, then this may result in higher uptake. The notion of elasticity adopted in the model is based on an econometrics study of the price elasticity of vehicle demand in the absence of country-specific data. The adopted elasticities, derived from meta-analyses, have been adjusted considering sectoral and country contexts. The more conservative short-term elasticities are typically considered for the possible program lifetime. Of course, these should be used if credible country-specific elasticities are available. Under most circumstances, car and motorcycle purchases for household transportation are price sensitive, although purchasers do not look to only optimize transport cost. A combination of needs, budget, desired features, the competition’s offerings, safety, fuel economy, and brand preferences could be more critical. While the model applies elasticities to the purchase of cars for household transportation, it generally considers that the price-demand elasticity for commercial vehicles is basically inelastic. However, this can be changed according to the policy package being evaluated. * * * Table 2. Vehicle categories and segments used in the model | Vehicle category | Buses by segment | | --- | --- | | Passenger cars | Urban buses, midi<=15t | | Light commercial vehicles | Urban buses, standard 15-18t | | Buses | Urban buses, articulated>18t | | Heavy trucks | Coaches, standard<=18t | | Motorcycles and 3-wheelers | Coaches, articulated>18t | | Passenger cars by segment | Heavy trucks by segment | | Mini | >3.5t | | Small | Rigid<=7.5t | | Medium | Rigid7.5-12t | | Large,SUV,luxury | Rigid12-14t | | | Rigid14-20t | | Light commercial vehicles by segment | Rigid20-26t | | N1-I | Rigid26-28t | | N1-II | Rigid28-32t | | N1-III | Rigid>32t | | | Articulated14-20t | | Motorcycles and 3-wheelers by segment | Articulated20-28t | | Mopeds2-stroke<50cm³ | Articulated28-34t | | Mopeds4-stroke<50cm³ | Articulated34-40t | | Motorcycles2-stroke>50cm³ | Articulated40-50t | | Motorcycles4-stroke<250cm³ | Articulated50-60t | | Motorcycles4-stroke250-750cm³ | | | Motorcycles4-stroke>750cm³ | | | Quad3W&ATV | | | Microcar | | > 750,\\mathrm{c m}^{3} Emissions and fuel consumption per vehicle depend on the distance a vehicle travels and can vary significantly depending on the vehicle’s speed. Thus, the annual distance traveled, which will vary by vehicle type and age and the percentage of this travel that occurs on an urban street or rural highway, 4 along with the average travel speed in each environment, needs to be entered into the model. Usage | Fuel and technology | | --- | | Petrol | | Petrol hybrid | | Petrol PHEV | | Diesel | | Diesel PHEV | | LPG bifuel | | CNG bifuel | | BEV | | | | | | | | Emissions standard | | Conventional | | Euro 1/I | | Euro 2/II | | Euro 3/III | | Euro 4/IV | | Euro 5/V, V+ | | Euro 6/VI A/B/C/D/E | | Electric | * * * However, sensitivity is low because of the bottom-up to top-down calibration. If annual mileage were to be increased, the resultant correction factor would reduce. This information can be obtained from: • A traffic simulation model (e.g., VISSUM, EMME2), • A survey (e.g., the Gas Station Survey), or • Expert opinion. In future years, vehicle travel time may change when increased mobility and vehicle usage can impact 5 the typical congestion level and average speed. Another way of obtaining speed estimates from road users is via a level of service estimate, as shown in table 3. Table 3. Relationship between service level and speed, by type of road | Road type | Level of service | Speed/Limit | | --- | --- | --- | | Motorway-national | | | | | Freeflow | >0.91894 | | | Heavy | 0.78231-0.91894 | | | Saturated | 0.38413-0.78230 | | | Stop+go | <0.38413 | | Motorway-city | | | | | Freeflow | >0.93746 | | | Heavy | 0.82821-0.93746 | | | Saturated | 0.49121-0.82820 | | | Stop+go | <0.49121 | | Trunkroad/primary-city | | | | | Freeflow | >0.87848 | | | Heavy | 0.75303-0.87848 | | | Saturated | 0.45306-0.75302 | | | Stop+go | <0.45306 | | Distributor/secondary | | | | | Freeflow | >0.80333 | | | Heavy | 0.66006-0.80333 | | | Saturated | 0.41766-0.66005 | | | Stop+go | <0.41766 | | Local/collector | | | | | Freeflow | >0.82409 | | | Heavy | 0.65358-0.82409 | | | Saturated | 0.42271-0.65357 | | | Stop+go | <0.42271 | | Access-residential | | | | | Freeflow | >0.96347 | | | Heavy | 0.81352-0.96347 | | | Saturated | 0.57726-0.81351 | | | Stop+go | <0.57726 | Source: The “traffic situation” model from Mario Keller, Handbook Emission Factors for Road Transport 3.1: Quick Reference (Bern: INFRAS, 2010), [https://cdn.prod.website-files.com/6207922a2acc01004530a67e/64dbb2e9c9e37c80513e0a4b\_HBEFA31\_Help\_en.pdf](https://cdn.prod.website-files.com/6207922a2acc01004530a67e/64dbb2e9c9e37c80513e0a4b_HBEFA31_Help_en.pdf). 5Average speed in this context equals the distance traveled divided by the travel time and is expressed in kilometers per hour. * * * 3. Conducting the Analysis Annual vehicle use may differ between the WithPolicy and WithoutPolicy scenarios. For example, when passengers switch to buses for commuting, this is unlikely, in many cases, to affect their vehicle ownership but can directly impact the annual use of their vehicles. Similarly, end users who decide to purchase an EV instead of an internal combustion engine vehicle may limit its use for longer journeys. ## Ridership/load factor The relationship between vehicle-kilometer and passenger-kilometer or freight-ton-kilometer traveled is required when the model evaluates modal shift between travel options. Information to analyze this relationship can be collected from: - Origin-destination (O-D) surveys, - Other surveys (e.g., the Gas station Survey), or - Expert opinion. ## Energy consumption The MRV methodology uses COPERT’s empirically derived emission factors to model energy consumption when local figures are not available. Each emission factor in COPERT is a function of the average vehicle speed. This emission factor has been derived using mathematical regression based on experimental data when driving under real-world conditions. The MRV methodology currently uses COPERT version 5.7.3, which was published by EMISIA for the European Environmental Agency in January 2024 and includes updates to previous versions. COPERT is used to model vehicular energy consumption and all criteria pollutants. ## CO2 emission factors CO 2 emissions are modeled from the energy consumption for each fuel. When local figures are not available, default values are used from IPCC’s “Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories” ( [https://www.ipcc-nggip.iges.or.jp/public/gp/bgp/2\_3\_](https://www.ipcc-nggip.iges.or.jp/public/gp/bgp/2_3_) Road\_Transport.pdf). # 3.2 Preliminary analytical steps ## Set the baseline Defining and building agreement around the counterfactual WithoutPolicy baseline for emission reduction crediting can be complex. Since the emission reduction that can be reported is the difference between the WithPolicy and WithoutPolicy scenarios, setting this baseline has a direct impact on the emission reduction. It is a balancing act to demonstrate that the resultant emission reduction is sufficiently conservative while not disadvantaging the host country. The basis for defining * * * Transport Emissions Analysis Methodology (TEAM) the transport activity likely to have been implemented without a given policy package must follow a rationale acceptable to all stakeholders. There will always be changes in GHG emissions that are not captured by the definition of the WithPolicy and WithoutPolicy scenarios. These emissions will be treated as leakage and are held constant between the two scenarios. # Perform scenario calculations The model performs the scenario calculations based on the following approach. First, the model analyzes the change in vehicle ownership and use together with changes in transport demand from the policy package being evaluated in this analysis (e.g., the implementation of emission and fuel economy standards for new vehicles). As historic data under the WithPolicy scenario include the impact of these changes, the counterfactual WithoutPolicy scenario is modified to represent what would most likely have happened had the policy package not been enacted. While part of this change will be deterministic (a new obligatory standard is brought into force), it could, for example, change the price of vehicles or other characteristics that affect the end-user’s choice. When vehicle prices or other characteristics influencing end-users’ choice are altered, the demand adjustment is determined by analyzing the price (or time) effect by employing the most rigorous possible of the following approaches: a) Time-series econometric analysis b) Panel econometric analysis c) General equilibrium analysis d) Adopting robust estimates from the literature The model is set up to use approach (d) in the understanding that other more rigorous methods will be applied whenever possible. There are cases, however, when the end-user choice does not respond to variable cost, or it is difficult to define the response. In these cases, choice elasticity may not be applied, and this can result in a more conservative emission reduction, although this has to be substantiated. This is treated in more detail later in this report’s discussion of transport demand. The between-scenario difference in vehicle ownership, use, and energy demand can be analyzed through the various impact channels. Before the model can produce results, it should be calibrated to local conditions using the energy balance method as described later. After analyzing the change in vehicle ownership and use and changes in transport demand from the policy package under consideration, the model determines the end-user energy consumption by fuel and emissions for the WithPolicy and WithoutPolicy scenarios in the base year, subsequent (historical) years, and—ex ante—in future years through the end of the modeling period. - Any constraint on energy supply by fuel type should be applied. - Any additional energy supply needed to meet the demand for operation in future years should now be defined. Since energy demand in the WithoutPolicy scenario is likely higher, additional capacity may be required. * * * 3. Conducting the Analysis The emission differences between the two runs has to be evaluated now. The results-based emission crediting is informed by the (ex post) analysis based on historical data within the confines of the different baselines. At the same time, the recalculation of the (ex ante) forward-looking estimate reevaluates the expected total value of the program. Additionally, enacting the policy package under consideration may generate a net cash flow difference between the WithPolicy and WithoutPolicy runs. This additional free cash flow could be utilized for other improvements to the transport system that yield further GHG emission reductions. For example, this would be the case when a fossil fuel subsidy reform is used to improve public transport. This can be identified in a separate impact channel and increase the calculated emission reduction from the enactment of the policy package. # 3.3 Analytical framework The analytical framework for road vehicles is shown across figures 10–13 in four groups. The figures contain descriptions of each step within the respective groups: a) Passenger cars and two-wheelers (figure 10) b) Mass passenger transport (buses and coaches) (figure 11) c) Freight transport (light commercial vehicles, medium- and heavy-duty trucks, and tractor- trailers) (figure 12) d) Efficiency and emissions (for all vehicles) (figure 13) At a later stage, rail (metro, suburban, intercity, and high-speed rail), waterborne (inland waterways and coastal shipping), and air (national and regional aviation) may be added to the model. Figure 10.Figure 1 Passenger cars and two-wheelers **0\. Passenger cars and two-wheelers** Macro Scenario GDP Urbanization Population 1. Projected Urban/Rural household vehicle ownership 4. Projected 5. Projected population changes 6. Projected use patterns from household economic expenditure survey data active population Vehicle Market sales mix / Mix of use patterns Income per Average No. of development 100 quantiles household size households info year Change to Total population vehicle active mix / year Average km vehicle Annual Vehicle travelled per vehicle Per capita Household population replacement sales related to age / income expenditure Household ownership / year volumes / year year / ave. speed projections of vehicles Mortality calculations 2. Baseline population of 3. Total baseline vehicle population 7. Baseline ridership 8. Projected travel distance additional vehicles Baseline database Disaggregated by e.g. Baseline passenger Future -km Modal Total active of in-use vehicles vehicle-km travelled / shift vehicle Type (subtype) travelled year assump- population in Fuel tions baseline year Technology Vehicle mortality calculations / year Gross vehicle weight Vehicle Non-household Age passengers per Average no. of passenger- Baseline fixed and variable Vehicle-km travelled / ownership of Historic vehicle Usage (km/yr) vehicle km travelled costs / year vehicles sales data / year by Age year _Note: GDP = gross domestic product; km = kilometer; yr = year._ _Note: GDP = gross domestic product; km = kilometer; yr = year._ pg. 24 * * * Figure 11.Figure 1 Mass passenger transport (buses and coaches) 1. Mass passenger transport (buses and coaches) Note: GDP = gross domestic product; km = kilometer; yr = year. Note: GDP = gross domestic product; km = kilometer; yr = year. Figure 12.Figure Freight transport (light commercial vehicles, medium- and heavy-duty trucks, and 1 Freight transport (light commercial vehicles, medium- and heavy-duty trucks, and tractortractor-trailers) trailers) Note: GDP = gross domestic product; km = kilometer; yr = year. Note: GDP = gross domestic product; km = kilometer; yr = year. * * * Figure 13. Efficiency and emissions (for all vehicles) Figure 2 Efficiency and emissions (for all vehicles) Note: CO2 = carbon dioxide; km = kilometer. Note: CO2 = carbon dioxide; km = kilometer. \\mathrm{C O}\_{2} Analyze passenger car and motorcycle ownershipThe steps in this section refer to Figure 10.. Forecast vehicle ownership The steps in this section refer to figure 10. Forecast vehicle ownership using HH survey data (refers to Figure 10., step 1). Forecast vehicle ownership The World Bank has supported HH surveys in client countries for many years. External users can What is a household? Each Living Standards Measurement Study survey has data on vehicle ownership, including of cars What is a household? and motorbikes, as well as data on HH size and expenditure, which can be used to estimate future A “household” in Living Standards Measurement Study surveys refers to “normal” households and growth in the private motorized vehicle fleet using the methodological steps given further below. includes the people who generally eat together; institutional and homeless households are excluded from the surveys. Thus, the total population of normal households is less than the census figure. Note that each “typical” household in the survey has an associated weight. Households have been assigned weights for the purpose of extrapolation from the limited number of study households to national results.pg. 27 * * * Transport Emissions Analysis Methodology (TEAM) ## Measure income HH expenditure is a proxy measure for income as it is more consistent and reliable. Preparation: Identify location and determine income level Determine urban and rural households adopting the definition of household used by the HH survey: - For each (urban/rural) location, calculate income centiles. - Determine (1) the mean monthly HH expenditure (MMHE) and (2) the number of HH members or HH size. - Calculate the mean per capita expenditure (MPCE) as follows: MMHE/HH size. - Rank households according to MPCE (not by MMHE). - Create 100 quantiles (or centiles), with the centiles having relatively similar numbers of households. Note that HH size decreases with centile. Therefore, the number of persons decreases with centile and the mean HH expenditure and per capita expenditure increase with centile. - Forecast the population of “normal” households over the modeling period using UN census projections (or using national sources) and use the projections of “Adult Population aged 20–64” to define the average number of households by location. Use projections of other age groups as “dependents” to define the average HH size by location. - Adjust the HH size for each centile in future years so that the average (by location and year) matches the value calculated using census data. Identify and project vehicle ownership patterns - Among all households, for each location, and across centiles, determine vehicle ownership by type of vehicle (using the question “Does the household own one or more cars?”) - Among households owning a vehicle, determine the average number of vehicles owned by a household for a given centile for both locations (rural and urban) for each vehicle type (using the question “How many cars does the household own?”) - The analysis is conducted on car and motorcycle ownership only. Analyze ownership - Using the HH survey data, examine vehicle ownership patterns using a simplified model: ‒ Vehicle ownership = f(Average HH expenditurei), where i is the expenditure centile. ‒ Calculate parameters. ‒ Use the parameters calculated from the survey for the duration of the modeling period. ‒ Select the function that provides the best overall fit or use a forced saturation model that improves the fit of the curve. * * * 2. Access to Electricity–– Use the parameters calculated from the survey for the duration of the modeling period Use the parameters calculated from the survey for the duration of the modeling period.. –– Select the function that provides the best overall fit or use a forced saturation model Select the function that provides the best overall fit or use a forced saturation model The following are the three-parameter Gompertz and logistic functions used in modeling whether a The following are the three The following are the three--parameter Gompertz and logistic functions used in modeling whether a parameter Gompertz and logistic functions used in modeling whether a household owns a vehicle: household owns a vehicle: household owns a vehicle: G o m p e r t z:\\quad o w n e r s h i p\_{i,k}=\\beta\_{1,k},^{ _}e^{-e\\left-\\beta\_{2,k},^{_}\\left(M M H E\_{i,k}-\\beta\_{3,k}\\right)\\right\ } i = The Equation 3ith expenditure centile. In both equations: L o g i s t i c:\\enspace\\mathit w{n e n s h s p} _{i,k}=\\beta_{ _{1,k}}\\left//\\left(1+e^{-{\\beta_{\_{2,k}}}^{\*}\\left(\\lambda M{M H E}{}}\_ _i i,k-\\beta_{\_{3,k}}\\right)\\right) In both equations: pg. pg. 2828 i = The ith expenditure centile. k = Location (urban, rural). The parameters are as follows: no. owned no. owned = b + b \* e -e-{b2,k\*(MMHEi,k-b3,k)} β1 = Upper bound of the curve or the saturation level past which ownership can no longer rise. \\beta\_{} β2 = Growth rate or how fast the curve turns toward saturation. The function will continually be increasing if > 0 and decreasing if < 0. \\beta\_{2} β3 = The inflection point, or, in this case, the average HH expenditure (MMHE) level above vehicles owned by a household: which electrification or ownership rises at an increasing rate. The following are the four-parameter S-curves with lower bounds used in modeling the number of \\beta\_{3}= \ \\mathrm{o r}, The following are the four MMHE = Mean gross monthly expenditure of the household.-parameter S-curves with lower bounds used in modeling the number of Figure 14. Three-parameter Gompertz curve for car + motorcycle ownership: Sample * * * Gompertz Gompertz Gompertz Equation 4 n o.o o n e d\_{i,k}=\\beta\_{\\mathrm{0},k}+\\beta\_{\ { _{1}},k},^{\*}e,{^-e-\\left{\\beta_{{ _{2}},k}\ ^\\ast\\left(M M\ E_{i,k}-\\beta\_{{\_{}},k}\\right)\\right}} LogisticLogistic Equation 5 n o.,o w n e d\_{i,k}=\\beta\_{o,k}+\\left(\\beta\_{\ ,{} _{1,{k}}}middle\\middle/\\left(1+e^{-\\beta_{a,k}\\ast\\left(M M H E\_{i,k}-\\beta\_{3,k}\\right)}\\right)\\right) In both equations: : In both equations In both equations: i i= = The The ithithexpenditure centile expenditure centile.. The ithithexpenditure centile expenditure centile.. kk= Location (urban, rural) i = The ith expenditure centile... = Location (urban, rural) k = Location (urban, rural). The parameters = Lower bound of the curve (which in our case is set to 1) The parametersare as follows are as follows: : The parameters = Lower bound of the curve (which in our case is set to 1) The parametersare as follows are as follows: : The parameters are as follows: = The upper bound or saturation level minus the lower bound (i.e., Saturation - b0 = The upper bound or saturation level minus the lower bound (i.e., Saturation - b0 β0 = Lower bound of the curve (which in our case is set to 1). bobtain the saturation level, calculate b0+ b1. \\beta\_{0} bobtain the saturation level, calculate b0+ b1. b β = The upper bound or saturation level minus the lower bound (i.e., Saturation -β). Hence, 1 0 = The growth rate or how fast the curve turns toward saturation b b. to obtain the saturation level, calculate β + β. 0 1 = The inflection point or MMHE level above which electrification or ownership rises at an \\beta\_{} \\mathsf{S a t u r a t i o n}-\\beta\_{o0} \\beta\_{o}+\\beta\_{1}, = The inflection point or MMHE level above which electrification β2 = The growth rate or how fast the curve turns toward saturation. increasing rate. \\beta\_{2} increasing rate. β3 = The inflection point or MMHE level above which electrification or ownership rises at an increasing rate. Figure 4. Four-parameter Gompertz curve for the number of vehicles owned by a household: Sample \\beta\_{3} Figure 4. Four-parameter Gompertz curve for the number of vehicles owned by a household: Sample Figure 15. Four-parameter Gompertz curve for the number of vehicles owned by a household: Sample * * * Gompertz functions: Forced saturation Gompertz functions: Forced saturation Equation 6 \\mathit o w n e r s h i p\_{i,k,t}=\\beta\_{1,\\mathrm{l o o},k}\*e,^{{-e-\\left{\\beta\_{2,k}\*\*\\left(M\\mathit{M H E} _{i,k,t}-\\beta_{3,k}\\right)\\right}}} In the equation In the equation: : i = The i = The ith expenditure centile ith expenditure centile.. k = Location (urban, rural). k = Location (urban, rural). k = Location (urban, rural). t = Period. t = Period. t = Period. The parameters are as follows: The parameters are as follows: b1,100= Saturation level for the highest centile (centile = 100). 1,100= Saturation level for the highest centile (centile = 100). β = Saturation level for the highest centile (centile = 100). 1,100 = How fast the curve turns toward saturation. \\beta\_{1,100} β2 = How fast the curve turns toward saturation. = The MMHE level above which electrification or ownership increases at an increasing rate. \\beta\_{2} β3= The MMHE level above which electrification or ownership increases at an increasing rate. \\beta\_{3} Figure 5 Sample curve for free vs forced saturation Figure 16. Sample curve for free vs forced saturation Note: VND Note: VND = \[\[?\]\] \[\[make all labels consistently sentence case. Insert comma for figures in thousands. Y axis label not clear\]\ \ * * *\ \ Transport Emissions Analysis Methodology (TEAM)\ \ Summary of the methodology\ \ - Prepare data by creating expenditure centiles for both urban and rural locations.\ \ - Project MMHE over the modeling period.\ \ - Adjust future years so that the change in HH size by location is consistent with the UN projections over the modeling period.\ \ - Determine the relationship of vehicle ownership and HH expenditure using saturation/sigmoid/S- curves, such as Gompertz or logistic curves.\ \ - Use the calculated parameters as inputs into the MRV model.\ \ - Maintain the observed vehicle ownership pattern based on the saturation curve analysis during the modeling period. The analysis described above can be performed in:\ 6\ \ - Stata,\ \ - SPSS,\ \ - R, or\ \ - Excel. To convert Stata’s data files into Microsoft Excel (CSV format) using R:\ \ - library(haven)\ \ - yourData = read\_dta(“path/to/file”)\ \ - write.csv(yourData, file = “yourStataFile.csv”) To perform the analysis in R:\ \ - Package “grofit,” “easynls,” or FlexParamCurve or\ \ - Use Microsoft Excel’s Solver to fit parameters.\ \ \ # Baseline population of other vehicles\ \ Add non-household-owned vehicles in figure 10, step 2; figure 11, step 1; and figure 12, step 1.\ \ The active vehicle fleet also includes vehicles not owned by households (in use by different institutions, for commerce, and by industry, for example). This vehicular fleet has to be estimated and included in the analysis. The basis for this estimation is generally vehicle registration data. For passenger cars and motorcycles, this estimation is based on the difference between the number of registered in-use vehicles and the number of vehicles owned by households as defined above.\ \ Stata: see Stata, “The Complete Statistical Software for Data Science,” [https://www.stata.com/](https://www.stata.com/); SPSS: see IBM, “IBM SPSS Statistics,” [https://www](https://www/). ibm.com/products/spss-statistics?utm\_content=SRCWW&p1=Search&p4=43700050715561164&p5=e&gclid=Cj0KCQjw8O-VBhCpARIsACMv- VLOdLb6OknPqowr26Vto39TssJK1uAJCWdbpNgFhvInytalXQfpfI2gaAkfVEALw\_wcB&gclsrc=aw.ds; R: see R Foundation, “The R Project for Statistical Computing,” [https://www.r-project.org/](https://www.r-project.org/).\ \ * * *\ \ 2. Access to Electricity Note that the number of vehicles not owned by households is likely to change from year to year, although following a trajectory different from that for vehicles owned by households. The basis for estimating this change will differ between countries, depending on local knowledge. In some countries, it can be expected that the number of vehicles will not increase significantly over the coming years, while in others, estimating an increase in this fleet similar to GDP growth could be more logical.\ \ ## Total base year passenger car population\ \ Disaggregate the active fleet by vehicle category and segment (see figure 10, step 3; figure 11, step 1; and figure 12, step 1).\ \ Vehicles differ in size, weight, use, fuel, and technology, and the active fleet has to be disaggregated to apply the corresponding emission factors to each vehicle type. The level of detail needed is shown in figure 3.\ \ ## Use of vehicle registration data\ \ If a country has reliable vehicle registration data, then this is the most reliable source for disaggregating the active fleet in the base year. There are two cases that affect how registration data can be used.\ \ # Case 1 (normal forecast)\ \ We have the inventory of the previous year and sales for this year.\ \ Calculate mortality for the total vehicular fleet using Winfrey applied to the previous year.\ \ 1. New inventory = Last year’s inventory – Mortality + Sales.\ \ # Case 2 (normal backcast)\ \ We have the inventory and sales in the current and previous years:\ \ 1. Calculate mortality for the total vehicular fleet using Winfrey applied to the previous year.\ 2. Divide by the previous year’s inventory (subcategory).\ 3. Subtract from the current year’s inventory (subcategory).\ 4. We arrive at sales (Current year’s inventory – Previous year’s inventory + Mortality) (subcategory). Where the registration database is not sufficiently detailed or up to date, vehicle sales data can often be found for the previous 20 years. These data can be used through a mortality function to reconstrue an estimated fleet profile utilizing the logic of Equation 1. In all other cases, a survey needs to be conducted. A quick and easy way is via the “Gas station survey,” which is documented separately. Often, the registration database does not have all the required information, or some of the fields are free capture, resulting in the same data being written in multiple forms. Here, a worldwide database (e.g., [https://car2db.com](https://car2db.com/)) can help filter the registration database and include any needed data that are absent (e.g., vehicle curb weight).\ \ * * *\ \ 2. Divide up by previous years inventory (segment and category)\ 3. Subtract from this year’s inventory (segment and category)\ 4. Gives sales (this year inventory – previous inventory + mortality) (segment and category)\ Where the registration database is not sufficiently detailed or up to date, vehicle sales data can be found for the previous 10 to 20 years and used through a mortality function to reconstruct an estimated fleet profile using the logic of Equation 1. In other cases, a survey needs to be conducted. A quick and easy way of doing this is via the “Gas station Survey” documented separately.\ \ ## Equation 1\ \ 𝑃𝑃𝑃𝑄𝑃𝑅𝑃𝑄𝑃𝑄𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑄𝑃𝑃!"#$ &= 𝑃𝑃𝑃𝑄𝑃𝑅𝑃𝑄𝑃𝑄𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑄𝑃𝑃!"#$ '− 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑇!"#$ '\+ 𝑣𝑣𝑣𝑣 ℎ𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑖𝑖 \_𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 \_𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓!"#$ '\ \ Where: Populationyearx= the total in-use fleet of vehicles in year X **Scrappageyearx** = the number of vehicles that exit the in-use fleet in year X due to scrappage, disuse due to maintenance, parts, and reliability issues, or vehicles being sold outside of the analytical boundary (to another state or country for example). _**vehicles\_entering\_fleetyearx**_ = The number of vehicles entering the in-use fleet within the analytical boundary in year X. These will include new vehicles sold through the vehicle manufacturer’s dealers, new and used vehicles imported from outside of the analytical boundary (bought in another state for example). Transport Emissions Analysis Methodology (TEAM) Often the registration database does not have all the required information or some of the fields are free capture resulting in the same data being written in many different forms which complicates the analysis. Here the use of Mortality function a worldwide database (such as: [https://car2db.com/excel](https://car2db.com/excel)) can facilitate filtering the registration database and including needed data that is absent (such as vehicle curb weight).7 The preferred function to estimate vehicle mortality is the Winfrey S3 Curve, modified so that during the vehicle warranty period, little or no mortality is expected (while vehicles are under warranty, most Mortality Function damage is repaired). The preferred function to estimate vehicle mortality is the Winfrey S3 Curve7 which can be modified so that during the Where the average service life is defined as the period between the introduction of new assets and vehicle warranty little or no mortality is expected (whilst under warranty, most damage will be repaired). their active use and when 50 percent of those initial assets are still in use, the function takes the Where the average service li following form: fe is defined as the period between the introduction of new assets to active use and when 50 percent of those initial assets are still in use the function takes the form of: **Equation 7** _Equation 2_\ \ 𝑆𝑆𝑆𝑆 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 = max (0, 𝑌𝑌'∗ (1 − ((𝑌𝑌 − 𝑊𝑊)^2/𝐴𝐴𝐴𝐴 ^2))^𝑚𝑚))\ \ Where: Where: Survival = Survival = Percentage of vehicles still in service. percent of vehicles still in service _Y0_ = 1.\ \ 7Y = Years of use. _Y0_ = 1 Y = Years of use W = Warranty period in years. e W = warranty period in years Aa = Factor adjusted to give 50 percent for “Average age at retirement of A years” (years of Aa = factor adjusted to give use when only 50 percent of the original fleet is left in service). 50% for Average Age at Retirement of A years (Years of use when only 50% of original fleet is left in service) m = 3. m = 3 Note that the average age at retirement for different vehicle types can be determined from the “Gas Note that the average age at retirement station survey” or using Equation 1for different vehicle types can be determined from the “gas station when vehicle population and annual sales are known over several survey” or years. using Equation 1 when the population of vehicles and annual sales are known over several years.\ \ An alternative is to use An alternative is to use the survival function shown in the IPCC (2006) guidelines V2\_3\_Ch3\_Mobile\_ the survival function show in IPCC 2006 guidelines V2\_3\_Ch3\_Mobile\_Combustion Box\ \ 3.2.3 Combustion Box 3.2.3. _Equation 3_\ \ ## Equation 8\ \ 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑇𝑆𝑆𝑆𝑇𝑆𝑆𝑆𝑇 = 1 − 𝐸𝐸𝐸𝐸𝐸𝐸(−𝐸𝐸𝐸𝐸𝐸𝐸(𝑎𝑎 + 𝑏𝑏 ∗ 𝑌𝑌))\ \ Where: Where: Survival = percent of vehicles still in service after Y years Survival = Percentage of vehicles still in service after Y years. _a= 1.798_ b = -0.137 _a = 1.798._ Y = Years of use _b = -0.137._\ \ _Y = Years of use._ However, Equation 3 does not accommodate the difference in the in-use lives of different segments, classes, and types of vehicles.\ \ # Projected active\ \ See Edwin Bernard Kurtz,\ \ # population\ \ Life Expectancy of Physical Property Based on Mortality Laws, a foundational 1930 engineering textbook (Iowa State Active population of College, 1932); and Harold Andrews Cowles Jr, “Prediction of Mortality Characteristics of Industrial Property Groups,” Iowa State University ProQuest Dissertations & Theses, 1957, vehicles by year ( [https://www.proquest.com/docview/301919796/](https://www.proquest.com/docview/301919796/)\ \ Figure 6)\ \ .\ \ The total population of vehicles by category is then forecast over the modeling period based on the growth rate determined for privately owned cars (from the household living standards survey analysis) and for additional non-household-owned vehicles (Figure 6)\ \ # Projected active population changes\ \ Active population mix of vehicles determined for each future year (Figure 6)\ \ Using the forecast total population of vehicles by category and the estimated annual mortality8 8888888888 88888888 B0888888 C7\ \ 8888888 88888\ \ ## see step 3\ \ * * *\ \ However, Equation 8 does not accommodate the difference in the in-use lives of different classes\ and types of vehicles.\ \ Projected active population\ \ Active fleet by year (in reference to figure 10, step 5; figure 12, step 6; and figure 13, step 6).\ \ Vehicle fleet by category is then forecast for the entire the modeling period. The basis for this forecast\ is the growth rate calculated for privately owned cars (from the HH living standards survey analysis)\ and for other vehicles not owned by households.\ \ For heavy-duty trucks and for buses (figures 12 and 11, respectively), the future population of vehicles\ is forecast as follows:\ \ • Population by category and segment is determined in the base year.\ \ • The people and goods transported in the base year are calculated from ridership or load factor\ data.\ \ • The need for people and goods to be transported in future years is estimated.\ \ • The number of vehicles needed each year is calculated based on their carrying capacity and\ expected load factor.\ \ Projected changes in active population\ \ 8\ Using the forecast total population of vehicles and the estimated annual mortality, the total annual\ sales (of new and used vehicles entering the active fleet) are determined using Equation 1. Information\ on market development is used to evaluate the configuration of the vehicles included in annual sales\ by class of vehicles, technology type, and fuel (with a level of detail consistent with that shown in figure\ 3). Market demand changes, and the fact that, for example, a SUV was scrapped, does not necessarily\ mean that a new SUV with the same fuel and technology will be sold that year.\ \ This differentiation is vital since it directly impacts vehicle emission factors, and it is important to\ identify which changes in the vehicle market are due to the policy (or intervention) being evaluated\ and which are exogenous.\ \ BAU approach\ \ Three basic approaches can be employed.\ \ 8See step 3.\ \ * * *\ \ Transport Emissions Analysis Methodology (TEAM)\ \ # Marketing study approach\ \ All vehicle manufacturers maintain sales forecasts by dress level for their respective markets. While estimates of future sales mixes of individual manufacturers are often considered proprietary information, market-level forecasts, indicating expected changes in consumer preferences and regulations, can frequently be accessed via manufacturers’ (or importers’) associations. These data have higher validity over the short term.\ \ # Market vision approach\ \ This approach takes the information available on the expected longer-term market tendencies and backcasts to the current sales mix. This is useful for informing a discussion on the probable longer- term changes in consumer demand and the possible implementation of regulations affecting fuel consumption (and electrification), fuel economy, vehicle size, vehicle weight, and the future of private motorization. Constructing different scenarios can allow the model to evaluate the impact of other decisions on emissions.\ \ An example of this would be based on the International Council on Clean Transportation’s (ICCT’s) 9 report “Pathways to Decarbonization: The European Passenger Car Market in the years 2021–2035,” which outlines four future scenarios, illustrated in figure 17:\ \ 1. Adopted Policies scenario, where the share of both battery electric vehicles (BEVs) and plug-in hybrid electric vehicles (PHEVs) increases, from 7 percent in 2021 to 14 percent in 2025 and 27 percent in 2030 (table 7). Without stronger EU regulation, no further uptake of BEVs and no further technical progress is assumed beyond 2030. The Adopted Policies scenario ensures that the current EU regulatory targets of 15 percent and 37.5 percent lower CO emission\ 2 levels under the Worldwide Harmonized Light Vehicles Test Procedure (WLTP), relative to a 2021 baseline, for the new passenger car fleet are reached by 2025 and 2030, respectively.\ \ 2. Lower Ambition scenario, where the assumption is that there is progress on the fleet average WLTP CO emission target for 2025 with fleet average WLTP CO emissions falling from the\ 2 2 current 15 percent to 20 percent (table 8). From 2021 to 2025, the fleet penetration assumptions for BEVs and PHEVs remain the same as under the Adopted Policies scenario. While the share of BEVs increases to 56 percent, the share of PHEVs decreases to 13 percent. In sum, WLTP CO emissions of new car fleets decrease 70 percent in 2035 relative to a 2021 baseline. 2\ \ 3. Moderate Ambition scenario, in which a complete phaseout of combustion engine vehicles by 2035 is assumed (table 9). An interim target of 30 percent lower WLTP CO emissions by\ 2 2025, relative to 2021, is applied, with improvement to 70 percent lower WLTP CO emissions 2 in 2030. For 2030, the share of BEVs grows strongly to 54 percent, whereas the share of PHEVs drops to 14 percent. By 2035, all new vehicles registered are BEVs.\ \ 4. Higher Ambition scenario, in which the phaseout of combustion engines by 2030 is assumed, again relying entirely on BEVs. Between 2025 and 2030, the share of BEVs will rise rapidly to 100 percent, whereas all other technologies are phased out.\ Peter Mock and Sonsoles Díaz, “Pathways to Decarbonization: The European Passenger Car Market, 2021–2035,” ICCT white paper, International Council on Clean Transportation, May 19, 2021, [https://theicct.org/publication/pathways-to-decarbonization-the-european-pas-](https://theicct.org/publication/pathways-to-decarbonization-the-european-pas-) senger-car-market-2021-2035/#:~:text=Pathways%20to%20decarbonization%3A%20The%20European%20passenger%20car%20market%2C%20 2021–2035,-May%2019%2C%202021&text=As%20part%20of%20the%20roadmap,by%20the%20middle%20of%202021.\ \ \ * * *\ \ effect will be different from country to country, since it depends on the age and growth of the passenger2. Access to Electricity\ \ car fleet, the historical technology level in the fleet, and the mix of new/used vehicles entering the in-\ \ use fleet every year.Each of these scenarios—or variants (e.g., moderate ambition with a five-year lag)—can be evaluated by the model for any client country, and their impact on GHG emissions can be determined. Note that the effect will be different from country to country, since it depends on the age and growth of Also, it is important to note that these forecast estimations feed into the exthe passenger car fleet, the historical technology level in the fleet, and the mix of new/used vehiclesante calculation, which entering the in-use fleet every year.\ \ # evaluates the possible future impact of policy decisions. They do not impact the emission reduction\ \ # determination for crediting, which depends on the MRV of passenger car activity.\ \ Also, it is important to note that these forecast estimations feed into the ex ante calculation, which evaluates the possible future impact of policy decisions. They do not impact the emission reduction determination for crediting, which depends on the MRV of passenger car activity.\ \ **Figure 6. ICCT Pathways to Decarbonization for the EU passenger car market**\ \ Figure 17. ICCT pathways to decarbonization for the EU passenger car market\ \ _Note: CO2_ = carbon dioxide _Note: CO = carbon dioxide; ICCT = International Council on Clean Transportation._; ICCT = International Council on Clean Transportation.\ \ ## pg. 36\ \ * * *\ \ Transport Emissions Analysis Methodology (TEAM)\ \ ## Applicability of\ \ # 4\. Methodology\ \ ### 4.1 Determining changes in transport demand\ \ The methodology described here is used in the WithoutPolicy operation only to determine changes in end-user transport demand from the policy (or other interventions) being evaluated.\ \ The methodology seeks to construct the transport demand level in the counterfactual WithoutPolicy operation. In other words, it seeks to develop what would have been the level of demand had the policy package not been enacted.\ \ The methodology accounts for user behavior and its response to investment, pricing, and regulatory policies. It emphasizes mode choice as well as technology choice. The methodology recognizes the complementarity between investment in low-emission transport infrastructure and equipment and policies that signal the social costs of transport to users, which significantly differ between environments. Transport GHG emissions depend not on the volume of transport demanded but the structure of that demand. Discrete choices on which transport mode to use, which technology to use, and on-route decisions are responsible for the eventual emissions.\ \ #### Dimensions of choice\ \ Transport emissions rely on which transport modes users will use. The choice is between private motorized transport, mass transit, and nonmotorized transport. Vehicle purchase locks in technology and fuel intensity for years, as much of the purchase price is sunk. Whether new or improved infrastructure generates additional demand or reroutes existing traffic flows, depends on how users change their _trip plans or routes._\ \ Not including these choice dimensions in the GHG analysis runs a high risk of a mismatch between supply and demand for modal transport services.\ \ #### Determinants of behavior\ \ Income growth, cost of mobility, and the quality of services of different modes immediately affect transport demand and emissions. They determine the changes in trends for a transport system in the absence of the investments or policy action whose consequences for GHG emissions are being evaluated. Infrastructure investment, public finance, and regulation shape transport demand indirectly\ \ * * *\ \ 4. Applicability of Methodology through impacts on costs, time, and service availability and quality. Investments and policy action change the trend or impact baseline development in transport through these channels. The choice of transport mode and technology and the volume of modal transport demand are strongly influenced by income development. For most countries, the demand for mobility increases more than proportionately with income—a trend that subsides only at very high income levels. In many cases, the demand for freight transport is proportional to GDP growth.\ \ # Monetary costs of transport services\ \ Changes in the monetary costs of transport services determine changes in transport demand across modes. For private motorized mobility, fuel costs will dominate changes in transport expenditure. Substantial rises in fuel price diminish transport activity, leading to a shift to more fuel-efficient modes. For public transport, monetary costs are administered mainly by public authorities that prescribe ticket prices for operators. Generally, ticket prices vary slightly in the short term, much less than the prices of gasoline and diesel.\ \ # Time costs of transport services\ \ Time costs of transport services are, in general, more important for transport demand than monetary costs. Time costs are expected to increase with income as opportunities to earn additional income during the time lost in traffic increase. With freight deliveries relying heavily on time, depending on the logistics system, the value of time in freight transport can be high, as delays create a need for more extensive inventory or lead to costly disruption of production processes. In-vehicle times are only one dimension of time costs in transport. Wait times are an additional dimension.\ \ # Quality of transport services\ \ Finally, modal demand in transport is strongly influenced by service quality. Modal demand may depend on the quality of transport infrastructure or the comfort provided by vehicles. Often, service quality depends on the safety and security of a mode.\ \ # Estimating the impact on transport demand\ \ There are multiple ways to estimate the price and time impacts on the demand for transport. While reducing the uncertainty of this estimate is associated with increasing effort/cost, this more significant effort may yield a more attractive emission reduction, without affecting the conservativeness of the calculation. A few approaches are discussed below.\ \ ## Unadjusted transport demand\ \ Much evidence supports a general economic theory that transport demand is negatively correlated with transport cost and time. Therefore, one would expect the level of demand (thus emissions of CO2e) to be higher under the counterfactual WithoutPolicy scenario. This suggests that leaving demand unadjusted in the reference case likely establishes a conservative emission level for the measurement of the between-scenario emission difference.\ \ * * *\ \ Transport Emissions Analysis Methodology (TEAM)\ \ Measuring the price and time effect for adjusting demand\ \ Some of these interventions may seem relatively straightforward to model. A standard, for example, may require all new vehicles in year X to achieve a particular minimum fuel economy. The model estimates vehicle sales (by type and category) in that year, X, based on data on transport requirements and a mortality analysis for fleet replenishment. The model uses historic (or survey) data on operational use (kilometers/year) and a load factor to estimate operating costs, fuel consumption, emissions, and goods transported.\ \ However, the analysis could become more complicated when the following considerations are made:\ \ - The fuel economy standard may drive up the purchase cost for new vehicles. This increase in cost may modify the end-user’s decision on what vehicle to purchase and when to purchase.\ - The improved fuel economy may reduce operating costs and lead to an increase in operational use (kilometers/year).\ - The improved fuel economy could even lead to a change in freight charges, making the fleet more competitive and changing its operating dynamics. The example above looks at freight, but it also applies to passenger transport. The passenger travel demand curve appears to have a long tail, meaning that if travel costs to the user (monetary costs, time costs, discomfort, and costs of risks) decline sufficiently, people will tend to travel more. This will result in an increase of marginal-value travel, which offers minimal benefits to users; and if this marginal- value travel imposes external costs (e.g., GHG emissions), then its net benefits are often negative (total benefits are less than total costs). This lower-value travel tends to be highly pricing sensitive.\ 10\ \ When the alternative future low-carbon scenario includes the implementation of the above fuel economy standard (whereas the “normal” development BAU scenario does not), between-scenario differences in fleet size, operation, and fleet dynamics between could be triggered, and the overall impact on fuel economy and CO2 emissions may be starkly different than any calculation that ignores these price-sensitive effects.\ \ The demand adjustment (if applied) should be determined by analyzing the specific price-sensitive effect and by employing the most rigorous possible of the following approaches:\ \ a) Time series econometric analysis\ b) Panel econometric analysis\ c) General equilibrium analysis\ d) Adopting robust estimates from the literature\ Todd Litman, “Understanding Transport Demands and Elasticities; How Prices and Other Factors Affect Travel Behavior,” Victoria Transport Policy Institute, December 8, 2025, [https://www.vtpi.org/elasticities.pdf](https://www.vtpi.org/elasticities.pdf).\ \ * * *\ \ 4. Applicability of Methodology\ \ # Time series econometric analysis\ \ The dependent and explanatory variables have to be defined according to the scope of the policy (i.e., economywide, for specific sectors, or for specific consumer classes within sectors). For a universal change, time series data for total national-level transport by mode and average transport cost may be used. For specific end-user classes, an econometric analysis should be conducted individually for each user group.\ \ A few broad principles of the analysis in this subsection are as follows:\ \ 1. The econometric specification should be chosen based on the literature and country-specific evidence. Control variables typically include per capita income, weather conditions/climate index, the prices for other travel modes, and demographic and geographical control factors.\ 2. Appropriate co-integration techniques and robustness tests are recommended to avoid spurious correlation and to establish credible elasticity estimates.\ 3. The time series should consist of historical observations for over 30 years. The data series should have the entire range of historical variation in line with the magnitude of the change in question.\ 4. The period of the time series should not contain significant market or political changes that could markedly skew the elasticity of demand.\ 5. Elasticity estimates should reflect short-run elasticity, since the methodology captures change in annual time steps.\ 11\ \ ## Panel econometric analysis\ \ The panel co-integration technique may be used when a series of equivalent surveys are available across multiple years. At least four rounds of surveys spanning more than 15 years, with the difference between the ending period and the measurement year under or equal to five years, should be available.\ \ ## General equilibrium analysis\ \ The linkage between transport price and demand could also be analyzed using a computable general equilibrium model, typically constructed from a country’s input–output (I–O) table or social accounting matrix. For this methodology, a static computable general equilibrium model well calibrated for the assessment year can be used to estimate the response of transport demand to price shock.\ \ Adopting a robust estimate from the literature\ \ Existing estimates in the literature may be adopted provided that (1) the design of the analysis and the dependent variable used for the analysis are appropriate for the type and scope of the policy under consideration and (2) the econometric method follows the requirements outlined under the section _“Time-series econometric analysis” above and the procedure outlined therein, which involves unit-_ root and co-integration tests, as well as an estimate of the short-term elasticity of transport demand\ \ The use of short-run price elasticity is consistent with the principle of a conservative baseline. This is also considered pragmatic for the purpose of carbon crediting or for results-based payments against outcomes on emission reduction achieved using a manageable credit- ing/payment period of, say, five years, although the methodology leaves out possible long-run impacts (where there are no fixed capitals/ factors).\ \ * * *\ \ based on an error correction model. Adopting short-term price elasticities of demand for substitution\ transport modes from the literature follows the same guidelines and requirements for statistical rigor.\ \ This model and methodology, by default, adopts elasticities from the recent compilation “Understanding\ Transport Demands and Elasticities How Prices and Other Factors Affect Travel Behavior (June 2021),”\ Victoria Transport Policy Institute (Transportation Elasticities \[vtpi.org\], as well as other elasticities\ from the older seminal TRACE study and handbook (G. de Jong, Elasticity Handbook: Elasticities for\ Prototypical Contexts, TRACE project deliverable D5, EC contract RO-97-SC.2035 \[Brussels: European\ Commission, Directorate-General for Transport, 1999\]). TRACE was a comprehensive research program\ carried out by a consortium of European consultants and universities to understand and demonstrate\ the relationship between travel costs and time and the demand for car travel in the short and long\ terms. The program produced an easy-to-use Elasticity Handbook for including elasticity values for\ a range of “prototypical contexts,” to use for assessing first-order impacts on car travel demand at\ different planning levels.\ \ Elasticities\ \ Travel demand refers to the amount and type of goods and passenger travel that would most likely\ be chosen in a particular situation.\ \ Prices are the direct, perceived costs of using a good or service. Transport prices include monetary\ costs, travel time, discomfort, and risk. Price changes can affect trip frequency, route, mode, destination,\ scheduling, vehicle type, parking location, type of service selected, and location decisions. The impacts\ of pricing are commonly measured using elasticities, which is the percentage change in consumption\ (in this case, in travel activity) from every 1 percent change in price.\ \ The preferred method for calculating transportation elasticities (symbolized as η) is the arc elasticity\ method and its variation, the mid-point arc elasticity method. The arc elasticity method reflects the\ change in consumption (Q) from every 1 percent change in price (P).\ Equation 9\ \ Arc elasticity\ Equation 9\ \ \\mathfrak{n}=\\frac{\\Delta l o g Q}{\\Delta P}=\\frac{l o g Q\_{2}-l o g Q\_{1}}{l o g P\_{2}-l o g P\_{1}}\ \ Equation 10\ Equation 10\ \ * * *\ \ Mid-point arc elasticity\ Equation 11\ \ Equation 11\ Equation 11 Equation 11\ \ Equation 12\ Equation 12\ Equation 12\ \ Q\_{2}=\ Q Q\_{1}\\times\ \\left\[\\frac{P\_{1}(\\eta-1)-,P\_{2}(\\eta+1)}{P\_{2}(\\eta-1)-,P\_{1}(\\eta+1)}\\right\]\ \ Where:\ \ η = Arc elasticity value.\ Figure 6\ Figure 6\ \ \\gamma\ \ Figure 6\ P = The first price point.\ 1\ \ \\mathsf{P}\_{\\uparrow}\ \ P = The second price point.\ 2\ \ \ \ P{}\_{2}\ \ Q = The quantity point associated with the first price point (P1).\ 1\ \ \ \\up\_{1}\ \ \ ({\\sf P}\_{1})\ \ Q = The quantity point associated with the second price point (P2).\ 2\ \ {\\bf O}\_{2}\ \ \ \ (\ mathsf P\ \ For example, if the price of a good with -0.4 price elasticity rises 20 percent, then the resulting\ consumption can be calculated as 1.2^(-0.4) times the old consumption of that good, or 0.93, indicating\ a 7 percent reduction in consumption.\ \ Considerations\ \ • Elasticity analysis should use real (consumer price index–adjusted) prices instead of nominal or\ current prices (unadjusted for inflation). For example, if there is 10 percent inflation and nominal\ prices do not change during a period, then real prices will have declined 10 percent.\ \ • Estimation of mode substitution factors, such as the change in automobile trips from a change in\ transit trips, helps in the analysis of cross-elasticities. These factors vary depending on circumstances.\ \ • The mid-point arc elasticity method should be used if one of the price values is zero (e.g., parking\ charge).\ \ For example, when bus ridership increases due to reduced fares, typically no more than 10–50\ percent of the added trips will substitute for an automobile trip. That is, for two to ten additional\ \ 12Mark Wardman, Jeremy Toner, Nils Fearnley, Stefan Flügel, and Marit Killi, “Review and Meta-Analysis of Inter-Modal Cross-Elasticity Evidence,”\ Transportation Research Part A: Policy and Practice 118 (2018): 662–81, [https://doi.org/10.1016/j.tra.2018.10.002](https://doi.org/10.1016/j.tra.2018.10.002).\ \ * * *\ \ transit trips, automobile trips reduce by one. Conversely, when a disincentive, such as parking fee\ or road toll, causes automobile trips to decline, there is generally a 20–60 percent shift to transit,\ depending on the conditions.\ \ 13\ Some studies indicate that losses from an original endowment are valued at 2.25 times gains. For\ example, a typical motorist could be expected to respond 2.25 times as much to a new parking\ fee (they pay more if they use a parking space) as a parking cash-out incentive (they receive a\ 14\ rebate for reducing their use of parking spaces) of the same amount.\ \ • Multiple changes. With various changes, travel reductions are multiplicative, not additive. For\ example, if three strategies are proposed for implementation, which individually provide a 5\ percent, 6 percent, and 7 percent reduction in vehicle travel, the total predicted reduction is 17\ percent, calculated as (1-0.05) x (1-0.06) x (1-0.07) = 17.0, not 18 percent (5 + 6 + 7 = 18).\ \ • Elasticities vary by trip and traveler type. Commercial (business) travel tends to be less price\ sensitive than personal travel. Commuter trips tend to be less elastic than shopping or recreational\ trips. Higher-income travelers tend to be less sensitive to price than lower-income travelers.\ Weekday trips may have very different elasticities than weekend trips. Urban peak-period trips\ tend to be price inelastic because congestion discourages lower-value trips, leaving only highervalue automobile trips.\ \ • Quality and price of alternatives. Price sensitivity tends to increase with the availability of quality,\ affordable alternatives. For example, highway tolls tend to be more price sensitive if there are\ parallel untolled roadways, and car driving is less price sensitive when transport alternatives are\ poor.\ \ 4.2 Evaluating changes in policy variables\ \ The elasticities used in this model and methodology are structured into two groups:\ \ • Vehicle use (e.g., the impact of fuel price change on the demand for transport).\ \ • The impact of increases in fuel prices can be evaluated directly. If a policy measure is to increase\ the fuel tax, then the effect of that tax on the average fuel price must be calculated first. If the\ policy is to apply a carbon tax, then the impact of this tax on the price of each fuel type needs to\ be calculated.\ • Distance-based road pricing can also be translated to an equivalent change in fuel price and\ \ For example:\ \ Any policy change can be evaluated by translating first in terms of these groups.\ \ 13Richard H. Thaler, “Mental Accounting Matters,” Journal of Behavioral Decision Making 12, no. 3 (1999): 183–206, [https://onlinelibrary.wiley](https://onlinelibrary.wiley/).\ com/doi/10.1002/%28SICI%291099-0771%28199909%2912%3A3%3C183%3A%3AAID-BDM318%3E3.0.CO%3B2-F.\ 14Donald Shoup, “Evaluating the Effects of California’s Parking Cash-Out Law: Eight Case Studies,” Transport Policy 4, no. 4 (1997): 201–16.\ \ com/doi/10.1002/%28SICI%291099-0771%28199909%2912%3A3%3C183%3A%3AAID-BDM318%3E3.0.CO%3B2-F.\ 14Donald Shoup, “Evaluating the Effects of California’s Parking Cash-Out Law: Eight Case Studies,” Transport Policy 4, no. 4 (1997): 201–16.\ \ * * *\ \ 4. Applicability of Methodology\ \ - As seen in cordon-based, congestion zone charging, area-based road pricing can be evaluated using parking charge elasticities because the price to be paid is a more or less fixed amount per trip, and not a price per kilometer.\ - If a policy measure is to decrease maximum speed, then the impacts on the average time to travel by car need to be calculated to be able to use the travel timetables. If the change in the speed limit impacts certain types of roads more than others, then these tables cannot give the changes in route choice, which would have to be estimated from a link-based planning model.\ - If the impact on the overall travel time can be estimated, then the travel timetables can give the impact of congestion on mode choice and distribution/generation. Many policy changes do not have an immediate, complete response from users. An example is modal shift. Modal shift happens when one mode gains a competitive advantage in a similar market. Factors that can give this advantage include costs, capacity, time, flexibility, and reliability of transport. Depending on what is being transported, the importance of each factor may be different.\ Figure 18. Mechanism of modal shift\ \ _Source: Jean-Paul Rodrigue, “Principles of Modal Shift,” figure in The Geography of Transport Systems, 5th ed. (New York: Routledge, 2020)._\ The modal share in the inertia phase may differ from that in the maturity phase, which is managed by applying short- and long-term elasticities (figure 18). The time needed for the market potential to reach a new equilibrium in modal shares will vary from case to case. Some studies show ranges of elasticities. These can be usefully applied if the income effect is considered. For example, the elasticity of demand for vehicle travel against fuel price can be expected to be more inelastic for high-income groups. Some studies also show a number enclosed in parentheses. This is the most frequently found elasticity.\ \ * * *\ \ Transport Emissions Analysis Methodology (TEAM)\ \ **Source of the travel elasticities used in this model**\ \ This model currently uses, by default, elasticities from: G. de Jong,  Elasticity Handbook: Elasticities _for Prototypical Contexts, TRACE project deliverable D5, EC contract RO-97-SC.2035 (Brussels:_ European Commission, Directorate-General for Transport, 1999); and TRIMIS (Transport Research and Innovation Monitoring and Information System), “Costs of Private Road Travel and their Effects on Demand, Including Short and Long Term Elasticities,” [https://trimis.ec.europa.eu/project/](https://trimis.ec.europa.eu/project/) costs-private-road-travel-and-their-effects-demand-including-short-and-long-term.\ \ However, the model can be readily adapted to use values from other studies, or elasticities can be determined by the methods described above. Locally derived elasticities are preferred.\ \ # Conduct fuel efficiency and emission analysis\ \ The steps in this section are in reference to figure 13.\ \ This section describes how the methodology described here determines the vehicle emission factors (in grams/kilometer) that are applied to the vehicle activity measured ex post and forecast for the entire modeling period.\ \ The methodology covers all the vehicles and technologies shown in figure 3. However, while the interest is in fuel consumption and emissions of CO2e only, a simplifying assumption can be applied to emission standards, dividing them into two groups:\ \ - Lower: All emission standards for all fuels up to and including a selected level (e.g., Euro 3/III) and\ - Higher: All emission standards for all fuels above this selected threshold.\ \ ## Vehicle operating conditions and expenditure\ \ This section (in reference to figure 13, steps 1 and 2) describes the methodology to calculate the fuel consumption and GHG emissions from the in-use fleet of vehicles. The input data collection and analysis are described in the previous section of this paper.\ \ ## Vehicle emissions and fuel consumption\ \ Calculate vehicle emissions and fuel consumption (in reference to figure 13, step 3). The methodology is set up to use the empirical Tier 3 15 emission factors included in the COPERT 5.7.3 database. 16\ \ 15The Tier 1 approach used in this methodology is based on known fuel sales only, which would have to be disaggregated to estimate emis- sions from passenger cars only. The Tier 2 approach is also based on fuel sales and subdivides fuel sales into each of the different technology groups according to emission control legislation. The Tier 3 approach calculates fuel consumption and emissions based on vehicle activity data. As contained in Chapter 1.A.3.b.i–iv in the EEA/EMEP Road Transport 2021 Guidebook (version 2019, updated in October 2021). See EEA (European Environment Agency), EMEP/EEA Air Pollutant Emission Inventory Guidebook 2019 (Luxembourg: Publications Office of the European Union, 2019), [https://www.eea.europa.eu/publications/emep-eea-guidebook-2019/part-b-sectoral-guidance-chapters/1-energy/1-a-combustion/1-a-3-b-i/](https://www.eea.europa.eu/publications/emep-eea-guidebook-2019/part-b-sectoral-guidance-chapters/1-energy/1-a-combustion/1-a-3-b-i/) view.\ \ * * *\ \ The methodology allows calculating all exhaust emissions of carbon monoxide, oxides of nitrogen,\ non-methane volatile organic compounds, methane, CO2, nitrous oxide, ammonia, oxides of sulfur,\ exhaust particulate matter (PM), polycyclic aromatic hydrocarbons and persistent organic pollutants,\ dioxins, furans, persistent organic pollutants, hexachlorobenzene, and heavy metals present in the\ fuel and lubricant (lead, arsenic, cadmium, copper, chromium, mercury, nickel, selenium, and zinc).\ Nitric oxide emissions are split into nitric oxide and nitrogen dioxide emissions.\ \ \\mathrm{C O}\_{2^{i}}\ \ PM is also divided into emissions of elemental and organic carbon as a function of vehicle technology.\ Detailed speciation for non-methane volatile organic compounds is also provided, covering homologous\ series, such as alkanes, alkenes, alkynes, aldehydes, ketones, and aromatics compounds. PM mass\ emissions in vehicle exhausts mainly fall in the PM2.5 size range. Therefore, all PM mass emission\ factors are assumed to correspond to PM2.5. Emission factors for particle number and surface are\ also provided for different particle size ranges.\ \ The methodology does not cover non-exhaust emissions, such as fuel evaporation from vehicles, tire\ wear, brake wear, or road wear. However, by default, in the current model, only fuel consumption and\ CO2e emissions are reported, and the other emission factor equations are not loaded, to speed up\ calculations.\ \ {\\mathrm{C O}}\_{\\mathrm{e}}\ \ In this Tier 3 method, exhaust emissions are calculated using a combination of firm technical data\ (e.g., emission factors) and vehicle activity data, as discussed in the previous section. Total vehicular\ exhaust emissions are calculated as the sum of hot emissions (when the engine is at its normal\ operating temperature) and emissions during transient thermal engine operation (termed “coldstart” emissions). It should be noted that, in this context, the word “engine” is used as shorthand for\ “engine and any exhaust after-treatment devices.” The distinction between emissions during the “hot”\ stabilized phase and the transient “warming-up” phase is necessary due to the substantial difference in\ vehicle emission performance during these two conditions for certain pollutants. The concentrations\ of certain pollutants are many times higher during the warming-up period than during hot operation,\ and a different methodological approach is required to estimate the additional emissions during the\ warming-up period.\ \ To summarize, total emissions can be calculated by means of the following equation:\ To summarize, total emissions can be calculated using this equation:\ \ EHot= emissions (g) during stabilized (hot) engine operation\ E = Total emissions (in grams) of any pollutant for the spatial and temporal resolution of\ ECold= emissions (g) during transient thermal engine operation (cold start). Total\ the application.\ \ Table 3 indicates Efor which vehicle classes and pollutants cold start emissions need to be calculated. = Emissions (in grams) during stabilized (hot) engine operation.\ Hot\ \ \\mathsf\_{E\_{\\mathrm{T o t a}}\|\ \ E\_{T o t a l}=E\_{h o t}+,E\_{c o l d}\ \ \\mathrm{E}\_{\\mathrm{H o t}}\ \ E = Emissions (in grams) during transient thermal engine operation (cold start).\ Cold\ Table 3 - Summary of calculation methods applied for the different vehicle classes and pollutants\ \ \\mathsf{E}\_{\\mathrm{C o d d}}\ \ * * *\ \ Table 4 indicates for which vehicle classes and pollutants cold-start emissions have to be calculated.\ 𝐸𝐸() _#+= 𝐸𝐸,)_ \+ 𝐸𝐸-)+.\ \ Table 4. Summary of the calculation methods applied for the different vehicle classes and pollutants\ Total; = total emissions (g) of any pollutant for the spatial and temporal resolution of the application,\ \ | Vehicle category | EC | CO2 | NOx | CO | NMVOC | CH4 | PM | N2O | NH3 | SO2 | Pb | HM |\ | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |\ | Petrol passenger cars | A | C | A | A | A | A | B | A | A | C | C | C |\ | Diesel passenger cars | A | C | A | A | A | A | A | C | C | C | C | C |\ | LPG passenger cars | A | C | A | A | A | A | | C | | | | |\ | 2-stroke passenger cars | C | C | C | C | C | C | | C | C | C | C | C |\ | E85 passenger cars | A | C | A | A | A | A | B | A | A | C | C | C |\ | CNG passenger cars | A | C | A | A | A | A | B | A | A | C | C | C |\ \ (square square\_{B}\ \ Key:\ A\ \ Petrol passenger cars A C A A A A\ A Hot emissions + cold-start over emissions + evaporative losses\ Diesel passenger cars A C A A A A\ B Hot emissions + evaporative losses\ \ Diesel passenger cars\ B Hot emissions + evaporative losses\ LPG passenger cars\ C Hot emissions only\ \ LPG passenger cars\ C Hot emissions only2-stroke passengercars\ \ E85 passenger cars A C A A A A B A A C C C\ Note: CH4 = methane; CNG = compressed natural gas; CO = carbon monoxide; COCNG Passenger cars A C A A A A B = carbon dioxide; EC = elemental carbon; HM = A A C C C\ 2\ heavy metals (lead, arsenic, cadmium, copper, chromium, mercury, nickel, selenium, and zinc)Key:; LPG = liquefied petroleum gas; NMVOC =\ non-methane volatile organic compoundsA; N2O = nitrous oxide; NOx = oxides of nitrogen; NHHot emissions + Cold start over emissions + evaporative losses = ammonia; Pb = plumbum (lead); PM =\ 3\ particulate matter; SO2 = sulfur dioxide. BHot emissions + evaporative losses\ C Hot emissions only\ \ {\\mathrm{C H}}\_{4}=\ \ \\mathrm{N}\_{2}\\mathrm{O}\ \ Vehicular emissions rely heavily on the engine operation conditions. Engine operation conditions\ Vehicle emissions are heavily dependent on the engine operation conditions.and emission performance are distinctly different under different driving situations. In this respect, Different driving situations impose\ different engine operation conditions, and therefore a distinct emission performance. a distinction is made between driving in urban, rural, and highway environments. Different activity In this respect, a\ distinction is made between urban, rural and highway driving.data and emission factors are attributed to each driving situation. Cold-start emissions are attributed Different activity data and emission factors are\ mainly to urban driving (and secondarily to rural driving), as it is expected that a limited number of\ attributed to each driving situation. Cold-start emissions are attributed mainly to urban driving (and secondarily\ trips start at highway conditions. Therefore, as far as driving conditions are concerned, total emissions\ to rural driving), as it is expected that a limited number of trips start at highway conditions. Therefore, as far as\ can be calculated using this equation:\ driving conditions are concerned, total emissions can be calculated by means of the equation:\ \ Equation 14\ \ E\_{T o t a l}=E\_{U r b a n}+,E\_{R u r a l}+,E\_{H i g h w a y}\ \ \\mathsf{E} _{\\mathsf{U r b a n}^{\\prime}}\\mathsf{E}_{\\mathsf{R r r a}\|^\ \ illustrates the flow chart of the calculation. Total emissions are calculated by combining activity data for each vehicle category with appropriate\ emission factors. The emission factors vary according to the input data (driving situations, climatic\ conditions). Figure 19 illustrates the flow chart for the calculation of total emissions.\ \ \ mathsf E{}mathsf\_{\\mathsf i i g h w a y}\ \ * * *\ \ Figure 19. Flow chart for the application of the methodology Figure 19. Flow chart for the application of the methodology\ \ Calculation of annual emissions of all pollutants for all road traffic source categories at all defined territorial units and road classes\ \ Source: Chapter 1.A.3.b.iSource: –ivChapter 1.A.3.b.i–iv from the EEA/EMEP Road Transport 2021 Guidebook (version 2019, updated in October 2021). from the EEA/EMEP Road Transport 2021 Guidebook (version 2019, updated in October 2021).\ Note: RVP = reid vapor pressure.\ Note: RVP = reid vapor pressure.\ \ Hot exhaust emissions depend upon a variety of factors, including the distance that each vehicle travels, its Hot exhaust emissions depend upon various factors, including the distance traveled by each vehicle,\ Hot exhaust emissions depend upon various factors, including the distance traveled by each vehicle, its its speed (or road type), age, and weight. Hot exhaust emissions can be calculated using this equation:\ speed (or road type), its age, and its weight.\ speed (or road type), age, and weight. Hot exhaust emissions can be calculated using this equation:\ \ Hot emissions\ Hot emissions\ \ Mk,r = mileage per vehicle \[km/veh\] driven on roads of type r by vehicles of technology k, N = Number of vehicles of technology k in operation in the period concerned.\ k\ ; i, k, r = emission factor in \[g/km\] for pollutant i, relevant for the vehicle technology k, operated on\ \ \ {cal E E}\_{\\mathsf{H o t;i,k,r}}\ \ E\_{H o t:i,k,r}=N\_{k}\\times,M\_{k,r}\\times,e\_{H o t:i,k,r}\ \ \\mathbb{N}\_{\\mathrm{k}}\ \ speed (or road type), age, and weight. Hot exhaust emissions can be calculated using this equation:Equation 15\ \ \\mathrm{M}\_{\\mathrm{k}}\ \ ; i, k, r = Memission factor in \[g/km\] for pollutant i, relevant for the vehicle technology k, operated on = Mileage per vehicle (kilometers per vehicle) driven on roads of type r by vehicles of\ k,r\ roads of type r. technology k.\ \ * * *\ \ e = Emission factor (in grams per kilometer) for pollutant i, relevant for vehicle technology\ Hot:i,k,r\ k, operated on roads of type r.\ \ \ e\_{\\tt o t t;i,k,r}\ \ r = Road class (“urban,” “rural,” and “highway”).\ \ i = The pollutants in table 4.\ \ k = All the vehicle technologies in figure 3.\ \ The hot exhaust emission parameters to calculate the 51,670 emission factors for all the covered\ vehicle types and technologies (emission factors \[in grams/kilometer\] or energy consumption factors \[in\ megajoules/kilometer\] or #/kilometer or #/kilowatt-hour or grams/kilowatt-hour) are found in appendix\ 4 to section “1.A.3.b.i-iv Road Transport,” of the EMEP/EEA, Air Pollutant Emission Inventory Guidebook\ 2023 (Luxembourg: Publications Office of the European Union, 2023), [https://www.eea.europa.eu/en/](https://www.eea.europa.eu/en/)\ analysis/publications/emep-eea-guidebook-2023?activeTab=22266594-97f5-4524-946f-095a50759ae7.\ \ Cold-start emissions\ Cold-start emissions\ \ Cold-start emissions\ Cold-start emissions are calculated as additional emissions over the emissions that would be\ Cold-start emissions are calculated as an extra emission over the emissions that would be expected if all vehicles\ expected if all vehicles were operated only with hot engines and catalysts warmed up. A relevant\ were only operated with hot engines and warmed-up catalysts. A relevant factor, corresponding to the ratio of\ factor, corresponding to the ratio of cold-start emissions and hot exhaust emissions, is applied to the\ cold over hot emissions, is applied to the fraction of fraction of kilometers driven with a cold engine. kilometers driven with a cold engine.\ \ Cold-start emissions are introduced into the calculation as additional emissions per kilometer using\ Cold-start emissions are introduced into the calculation as additional emissions per km using the following\ this equation:\ \ Equation 16\ \ E\_{C o l d;i,j}=,\\beta\_{i,k}\\times N\_{k}\\times M\_{k}\\times e\_{H o t;i,k}\\times\\left((^{{C o l d}}\\div e^{{t o t}}\|\_{i,k}-1\\right)\ \ where:\ COLD;i,j = cold-start emissions of pollutant i (for the reference year), produced by vehicle technology k,\ \ = cold-start emissions of pollutant i (for the reference year), produced by vehicle technology k,\ E = Cold-start emissions of pollutant i (for the reference year), produced by vehicle\ Cold;i,j\ = fraction of mileage driven with a cold engine or the catalyst operated below the light-off\ technology k.\ temperature for pollutant i and vehicle technology k,\ \ temperature for pollutant i and vehicle technology k,\ β = Fraction of miles driven with a cold engine or with the catalyst operated below the light-\ = number of vehicles \[veh\] of technology k in circulation,i,k\ off temperature for pollutant i and vehicle technology k.\ = total mileage per vehicle \[km/veh\] in vehicle technology k,\ \ = total mileage per vehicle \[km/veh\] in vehicle technology k,\ N = Number of vehicles \[veh\] of technology k in circulation.\ = hot emission factor for pollutant i and vehicles of k technology, k\ \ i,k M= = Total mileage per vehicle \[km/veh\] of technology k.cold/hot emission quotient for pollutant i and vehicles of k technology.\ k\ The b-parameter depends upon ambient temperature (for practical reasons the average monthly\ \ \ {sqrt V}\_\ \ The b-parameter depends upon ambient temperature (for practical reasons the average monthly e = Hot emission factor for pollutant i and vehicles of technology k.\ Hot;i,k\ temperature can be used), and the pattern of vehicle use — in particular the average trip length.\ \ temperature can be used), and the pattern of vehicle use Cold Hot — in particular the average trip length.\ e /e \| = Cold-start/hot exhaust emission quotient for pollutant i and vehicles of technology k.\ i,k\ However, since information on trip length is not available in many countries for all vehicle classes,\ \ simplifications have been The β parameter depends on the ambient temperature (for practical reasons, the average monthly introduced, and a value of 12.4 km can be used unless a firm national estimate\ temperature can be used) and the pattern of vehicle use—in particular, the average trip length. However,\ is available.\ since information on trip length is not available in many countries for all vehicle classes, simplifications\ have been introduced, and a value of 12.4 km can be used, unless a firm national estimate is available.\ Electricity consumption and Battery Electric Vehicles\ \ * * *\ \ Electricity consumption and BEVs\ \ The COPERT database does not have emission factors for BEVs because they have no direct\ 17\ emissions. However, the methodology and model described herein are set up to calculate and\ report the electricity consumption of vehicles (in kilowatt-hours). For all gasoline and diesel PHEVs, the\ methodology calculates energy consumption from both energy sources (e.g., gasoline and electricity\ used to charge the battery).\ \ The utility factor (as shown in table 5) is applied to the COPERT emission factors, so that, for example,\ 37 percent of the annul mileage of a gasoline PHEV consumes electricity (from the plug-in charger)\ and 63 percent consumes gasoline (or electricity from gasoline). These factors are derived from the\ 18\ real-world use of PHEVs.\ \ For BEVs, the data set from Electric Vehicle Database, “All Electric Vehicles,” [https://ev-database.org/](https://ev-database.org/)\ is loaded into the model with each vehicle classified into subcategories and the average energy use\ (in watt-hours per kilometer) used in the calculations.\ \ Table 5. Utility factors for different technologies and fuels\ \ | | Utility Factor |\ | --- | --- |\ | CNG Bifuel:-Gasoline | 0.22 |\ | CNG Bifuel:-CNG | 0.78 |\ | Diesel PHEV:-Diesel | 0.63 |\ | Diesel PHEV:-Electricity | 0.37 |\ | Gasoline PHEV:-Electricity | 0.37 |\ | Gasoline PHEV:-Gasoline | 0.63 |\ | LPG Bifuel:-Gasoline | 0.22 |\ | LPG Bifuel:-LPG | 0.78 |\ \ Note: CNG = compressed natural gas; LPG = liquefied petroleum gas; PHEV = plug-in hybrid electric vehicle.\ \ 4.3 Carbon dioxide (CO ) emissions\ 2\ \ Emissions of ultimate19 CO stem from three sources:\ 2\ \ • Combustion of fuel,\ \ • Addition of carbon-containing additives in the exhaust.\ \ 17Since the methodology does not include non-exhaust emissions, such as tire wear, brake wear, or road wear.\ 18Patrick Plötz, Cornelius Moll, and Yaoming Li, “Real-World Usage of Plug-in Hybrid Electric Vehicles: Fuel Consumption, Electric Driving,\ \ 18Patrick Plötz, Cornelius Moll, and Yaoming Li, “Real-World Usage of Plug-in Hybrid Electric Vehicles: Fuel Consumption, Electric Driving,\ and CO2 Emissions,” ICCT white paper, International Council on Clean Transportation, September 27, 2020, [https://theicct.org/publication/](https://theicct.org/publication/)\ real-world-usage-of-plug-in-hybrid-electric-vehicles-fuel-consumption-electric-driving-and-co2-emissions/.\ 19Ultimate in this case means that the carbon contained in the three sources is fully oxidized into CO2.\ \ \\mathrm{C O}\_{2}\ \ * * *\ \ Combustion of fuel\ due to fuel combustion Combustion of lubricant oil CO due to fuel combustion\ 2\ due to fuel combustion\ If the fuel composition Addition of carbon of an oxygenated fuel described by the generic chemical formula C-containing additives in the exhaust\ \ \\mathrm{c0}\_{2}\ \ If the fuel composition Addition of carbon of an oxygenated fuel described by the generic chemical formula C-containing additives in the exhaust x HyOz is known from\ If the fuel composition If the composition of an oxygenated fuel described by the generic chemical formula C of an oxygenated fuel described by the generic chemical formula Cx HxyHOyz Ois known from z is known\ ultimate chemical analysis, then the mass fractions of carbon, hydrogen and oxygen atoms in the fuel are c, h,\ ultimate chemical analysis, then the mass fractions of carbon, hydrogen and oxygen atoms in the fuel are c, h, from ultimate chemical analysis, then the mass fractions of the carbon, hydrogen, and oxygen atoms\ and o, where c + h + o = 1. In this case, the ratios of hydrogen to carbon and oxygen to carbon in the fuel are\ due to fuel combustion\ and o, where c + h + o = 1. in the fuel are, respectively, In this case, the ratios of hydrogen to carbon and oxygen to carbon in the fuel are c, h, and o, where c + h + o = 1. In this case, the hydrogen–carbon atom\ respectively calculated as:\ If the fuel composition of an oxygenated fuel described by the generic chemical formula Cx HyOz is known from\ respectively calculated as: and oxygen–carbon atom ratios in the fuel are calculated as, respectively:\ ultimate chemical analysis, then the mass fractions of carbon, hydrogen and oxygen atoms in the fuel are c, h,\ \ C x\_{y}O\_{z}\ \ c,h\ \ c+h+o=1\ \ and o, where c + h + o = 1. Equation 17\ respectively calculated as:\ \ r\_{H,C}=11.916\\frac{h}{c}\ \ r\_{O,C}=0.7507\\frac{o}{c}\ \ 𝑜𝑜Equation 9\ With these ratios, the mass of the CO2 emitted by vehicles of technology 𝑟𝑟 = 0.7507 k and combusting fuel m\ \ > ,;\ > 𝑐𝑐𝐹𝐹𝐹𝐹:-#+-,@\ > can be calculated as: -#+-\ > With these ratios, the mass of CO 𝐸𝐸;>?,:,2 @emitted by vehicles in technology k, combusting fuel m can be calculated as: = 44.011 × 𝐹𝐹𝐹𝐹\ > -#+-:-#+-,@\ > 𝐸𝐸 = 44.011 × 12011 + 1.008𝑟𝑟 + 16.000𝑟𝑟\ \ \\mathrm{C O\_{2}}\ \ :Equation 18\ :\ ? is the fuel consumption of those vehicles for the time period considered.\ \ E\_{C O2,k,m}^{c a l c}=44.011\\times\\frac{F C\_{k,m}^{c a l c}}{12011+1.008r\_{H C,m}+16.000r\_{0C C,m}}\ \ Table 5 gives the ratios of hydrogen to carbon and oxygen to carbon atoms for different reference blend fuels CALC\ FCWhere: is the fuel consumption of those vehicles for the time period considered.\ Table 5 gives the ratios of hydrogen to carbon and oxygen to carbon atoms for different reference blend fuels\ (REF) used in vehicle testing and estimated values for non\ (REF) used in vehicle testing and estimated values for non\ \ (REF) used in vehicle testing and estimated values for non-reference fuels and blends. Oxygen in the fuel may be\ increased due to blending with oxygenated components and/or biofuels.\ calc\ FC = Fuel consumption of these vehicles for the time period considered.\ increased due to blending with oxygenated components and/or biofuels.\ Table 5 gives the ratios of hydrogen to carbon and oxygen to carbon atoms for different reference blend fuels\ \ Table 5 gives the ratios of hydrogen to carbon and oxygen to carbon atoms for different reference blend fuels\ When reporting CO2 emissions, only the fossil fuel statistical consumption should be taken into account in the\ (REF) used in vehicle testing and estimated values for non Table 6 gives the hydrogen–carbon atom and oxygen–carbon atom ratios for different reference -reference fuels and blends. Oxygen in the fuel may be\ When reporting CO2 emissions, only the fossil fuel statistical consumption should be taken into account in the calculation. This is consistent with the IPCC 1996 and IPCC 2006 guidelines, according to which emissions\ increased due to blending with oxygenated components and/or biofuels. blend fuels (REF) used in vehicle testing and the estimated values for non-reference fuels and blends.\ calculation. This is consistent with the IPCC 1996 and IPCC 2006 guidelines, according to which emissions\ associated with use of biofuels are attributed to the Land Use, Land-Use Change and Forestry sector under IPCC.\ Oxygen in the fuel may be increased due to blending with oxygenated components and/or biofuels.\ associated with use of biofuels are attributed to the Land Use, Land-Use Change and Forestry sector under IPCC.\ Hence, for reporting, the CO2 calculated per vehicle category should be corrected and the calculated CO2\ Hence, for reporting, the CO2 calculated per vehicle category should be corrected When reporting CO2 emissions, only the fossil fuel statistical consumption should be taken into account in the and the calculated CO2\ \ Hence, for reporting, the CO2 calculated per vehicle category should be corrected When reporting CO2 emissions, only the fossil fuel statistical consumption should be taken into account in the and the calculated CO2\ emission should be derived from Equation 9, without considering the oxygen content of the biofuel part.\ calculation. This is consistent with the IPCC 1996 and IPCC 2006 guidelines, according to which emissions When reporting CO emissions, only fossil fuel statistical consumption should be considered in the\ emission should be derived from2Equation 9, without considering the oxygen content of the biofuel part.\ associated with use of biofuels are attributed to the Land Use, Land calculation. This is consistent with the IPCC 1996 and IPCC 2006 guidelines, according to which -Use Change and Forestry sector under IPCC.\ Hence, for reporting, the CO2 calculated per vehicle category should be corrected emissions associated with the use of biofuels are attributed to the land use, land-use change and and the calculated CO2\ emission should be derived from forestry sector under the IPCC. Hence, for reporting, the CO Equation 9, without considering the oxygen content of the biofuel part. calculated per vehicle category should\ 2\ be corrected, and the calculated CO2 emissions should be derived using Equation 18, without\ considering the oxygen content of the biofuel part.\ \ \\mathrm{C O\_{2}}\ \ \\mathrm{C O}\_{2}\ \ \\mathrm{C O}\_{2}\ \ * * *\ \ Table 6. Hydrogen–carbon atom and oxygen–carbon atom ratios for different reference blend fuels (REF)\ used in vehicle testing and estimated values for non-reference fuels and blends\ \ | Fuel(m) | Typical Molecule | Ratio of hydrogen to carbon(rH:C) | Ratio of oxygen to carbon(rO:C) | kgCO2per kg of fuel | Max |\ | --- | --- | --- | --- | --- | --- |\ | Petrol | \[CH1.86\]x | 1.86 | 0 | 3.169 | 2.13 |\ | Diesel | \[CH1.86\]x | 1.86 | 0 | 3.169 | 1.7 |\ | Ethanol | C2H5OH | 3 | 0.5 | 1.911 | 2.13 |\ | Methanol | CH3OH | 4 | 1 | 1.373 | 2.13 |\ | Biodiesel | \[CH\]x-COOH | 1.95-2.03 | 0.11-0.13 | 2.797-2.727 | 2.13 |\ | ETBE | C6H14O | 2.33 | 0.167 | 2.584 | 1.7 |\ | MTBE | C5H12O | 2.4 | 0.2 | 2.496 | 2.13 |\ | Diesel | New | 1.28 | 0.43 | | 2.13 |\ | Natural gas/ Biogas(REF) | CH4, market fuels also contain heavier HC | 4 | 0 | 2.473 | 2.13 |\ | LPG(REF) | C3H8(15%)-C4H10(85%), market fuels may contain different proportions | 2.525 | 0 | 3.024 | 1.7 |\ | E5 | | 1.92 | 0.026 | 3.063 | 2.13 |\ | E10(REF) | | 1.98 | 0.053 | 2.694 | 2.13 |\ | E75 | | 2.73 | 0.38 | 2.111 | 2.13 |\ | E85(REF) | | 2.84 | 0.429 | 2.026 | 1.7 |\ | ETBE11 | | 1.91 | 0.018 | 3.094 | 2.13 |\ | ETBE22 | | 1.96 | 0.036 | 3.021 | 2.13 |\ | B7(REF) | | 1.86 | 0.007 | 3.144 | 2.13 |\ | B10 | | 1.86 | 0.01 | 3.133 | 1.7 |\ | B20 | | 1.87 | 0.02 | 3.096 | 2.13 |\ | B30 | | 1.88 | 0.03 | 3.059 | 2.13 |\ \ Source: EEA, EMEP/EEA Air Pollutant Emission Inventory Guidebook, [https://www.eea.europa.eu/emep-eea-guidebook](https://www.eea.europa.eu/emep-eea-guidebook).\ \ Note: CO2 = carbon dioxide; ETBE = ethyl tertiary-butyl ether; kg = kilogram; m = Fuel; REF = reference blend fuel; MTBE = methyl tertiarybutyl ether.\ \ CO2 resulting from lubricant oil\ \ \\mathrm{C O\_{2}}\ \ New and well-maintained vehicles generally need small amounts of lubricant oil, because there is an\ oil film developed on the inner walls of the cylinders. Prolonged engine operation generally increases\ lube oil consumption, and this consumption should therefore be expected to increase, on average,\ with vehicle age. Table 7 presents the typical oil consumption factors for different fuels and vehicle\ ages for passenger cars. An “old” vehicle is defined as one that is beyond its typical useful life (normally\ 150,000 km for a passenger car).\ \ \\mathrm{c0}\_{2}\ \ * * *\ \ Table 7. Oil consumption factors for passenger cars\ \ | | | kg/10,000km | | |\ | --- | --- | --- | --- | --- |\ | Fuel | Age | Mean | Min | Max |\ | Gasoline | Old | 1.45 | 0.85 | 2.13 |\ | Gasoline | New | 1.28 | 0.85 | 1.7 |\ | Diesel | Old | 1.49 | 0.85 | 2.13 |\ | Diesel | New | 1.28 | 0.43 | 2.13 |\ \ Note: kg = kilogram; km = kilometer.\ \ Emission degradation due to vehicle age\ \ Corrections are applied to the results to accommodate the variation in emissions due to vehicle age\ (kilometers). The baseline emission factors correspond to a fleet of average mileage (30,000–60,000\ km), and a degradation factor is therefore inherent. For petrol cars, further emission degradation—due to\ increased mileage—should be modeled using additional degradation factors, shown in tables 8 and 9.\ \ | MC=AM×MMEAN+BM | Capacityclass\[l\] | Averagemileage\[km\] | A M | B M Valueat0km | B M Valueat12000km |\ | --- | --- | --- | --- | --- | --- |\ | Correction forV<19km/h(MCURBAN) | | | | | |\ | CO-MCURBAN | 0.8-1.4 | 29057 | 1.523E-05 | 0.557 | 2.39 |\ | 1.4-2.0 | 39837 | 1.148E-05 | 0.543 | 1.92 | |\ | >2.0 | 47028 | 9.243E-06 | 0.565 | 1.67 | |\ | NOx-MCURBAN | ALL>0.8 | 44931 | 1.598E-05 | 0.282 | 2.2 |\ | HC-MCURBAN | ≤1.4 | 29057 | 1.215E-05 | 0.647 | 2.1 |\ | 1.4-2.0 | 39837 | 1.232E-05 | 0.509 | 1.99 | |\ | >2.0 | 4028 | 1.208E-05 | 0.432 | 1.88 | |\ | Correction forV>63km/h(MCROAD) | | | | | |\ | CO-MCROAD | 0.8-1.4 | 29057 | 1.689E-05 | 0.509 | 2.54 |\ | 1.4-2.0 | 39837 | 9.607E-06 | 0.617 | 1.77 | |\ | >2.0 | 47028 | 2.704E-06 | 0.873 | 1.2 | |\ | Nox-MCROAD | ALL>0.8 | 47186 | 1.220E-05 | 0.424 | 1.89 |\ | HC-MCROAD | 0.8-1.4 | 29057 | 6.570E-06 | 0.809 | 1.6 |\ | 1.4-2.0 | 39837 | 9.815E-06 | 0.609 | 1.79 | |\ | >2.0 | 47028 | 6.224E-06 | 0.707 | 1.45 | |\ \ Source: Section 1.A.3.b.i–iv, “Road Transport,” of the EMEP/EEA Air Pollutant Emission Inventory Guidebook 2023.\ \ * * *\ \ Table 9. Emission degradation due to vehicle age for Euro 3 and Euro 4 petrol passenger cars and\ light commercial vehicles (and Euro 1 and Euro 2 vehicles in case of an enhanced inspection and\ maintenance scheme)\ \ | MC=AM×MMEAN+BM | Capacityclass\[l\] | Averagemileage\[km\] | A M | B M Valueat0km | B M Valueat16000km |\ | --- | --- | --- | --- | --- | --- |\ | Correction forV<19km/h(MCURBAN) | | | | | |\ | CO-MCURBAN | $^{\\prime}\\leq1.4$ | 32407 | 7.13E-06 | 0.769 | 1.91 |\ | $^{\\prime}>1.4$ | 16993 | 2.67E-06 | 0.955 | 1.38 | |\ | NOx-MCURBAN | $^{\\prime}\\leq1.4$ | 31313 | 0 | 1 | 1 |\ | $^{\\prime}>1.4$ | 16993 | 3.99E-06 | 0.932 | 1.57 | |\ | HC-MCURBAN | $^{\\prime}\\leq1.4$ | 31972 | 3.42E-06 | 0.891 | 1.44 |\ | $^{\\prime}>1.4$ | 17913 | 0 | 1 | 1 | |\ | Correction forV>63km/h(MCROAD) | | | | | |\ | CO-MCROAD | $^{\\prime}\\leq1.4$ | 30123 | 1.50E-06 | 0.955 | 1.2 |\ | $^{\\prime}>1.4$ | 26150 | 0 | 1 | 1 | |\ | NOx-MCROAD | ALL | 26150 | 0 | 1 | 1 |\ | HC-MCROAD | ALL | 28042 | 0 | 1 | 1 |\ \ Source: Section 1.A.3.b.i–iv, “Road Transport,” of the EMEP/EEA Air Pollutant Emission Inventory Guidebook 2023.\ \ Note: AM = emission performance degradation per kilometer; BM = relative emission level of brand-new vehicles; CO = carbon monoxide;\ HC Hydrocarbons km = kilometer; I = liter; MC = hot emission factor correction; MMEAN = mean fleet mileage in km; Nox = nitrogen\ oxides; V = vehicle mean travelling speed in km/h.\ \ Calibrate to local conditions\ \ Calculate vehicle emissions and fuel consumption (in reference to figure 13, step 4). The methodology\ described here and the current model are set up to allow an energy balance calculation to match,\ wherever possible, the fuel consumption calculated for vehicles in the base year and reported statistical\ fuel sales. Often, this is difficult for diesel, because this fuel has multiple other uses (including, for\ example, use in off-road construction, and in agricultural machinery, generators, and pumping sets).\ However, this is sometimes possible for gasoline.\ \ * * *\ \ adjustment algorithm. Where possible, in calculating vehicle fleet energy consumption, a mileage correction factor is applied\ to the mean activity to balance the statistical and calculated energy consumption. Figure 20 presents\ the adjustment algorithm.\ Figure 20. Flow chart for the fuel energy balance algorithm\ \ Figure 20. Flow chart for the fuel energy balance algorithm\ Figure 20. Flow chart for the fuel energy balance algorithm\ \ Source: Section 1.A.3.b.i–Source: iv, “ Section 1.A.3.b.i–iv, “Road Transport,” of the EEA/EMEP Road Transport 2021 Guidebook (version 2023, updated in October 2023). Road Transport,” of the EEA/EMEP Road Transport 2021 Guidebook (version 2023, updated in\ Note: g = gram; MCF = Emissions Factor Correction; MJ = megajoule; veh-km = vehicle kilometer.\ \ Note: g = gram; MCF = Emissions Factor Correction; MJ = megajoule; veh-km = vehicle kilometer.\ Determine energy consumption and emission reductions\ Determine energy consumption and emission reductions\ \ Determine energy consumption and emission reductions\ Calculate annual vehicle emissions and fuel consumption (in reference to figure 13, steps 5–7).\ \ The only difference between the two scenarios is the change in the implementation of the specific\ policy or intervention. Care must be taken to identify and report likely sources of leakage and ensure\ that all considerations in the modeling lead to conservative results.\ \ * * *\ \ Mobility and Transport Connectivity Series