Editor’s summary
Electric vehicles (EVs) produce much lower emissions of greenhouse gases than internal combustion engine vehicles (ICEVs), but when in the life cycle of an ICEV does it make sense to replace it with an EV? Campbell and Geyer conducted life-cycle analyses of different vehicle types and concluded that the energy impact benefits from swapping an EV for an ICEV begin to accrue as early as the first day of the life of the ICEV, and that replacement early in the life cycle consistently provides reductions in greenhouse gas emissions. Financial considerations will also affect when people choose to make the replacement, of course, but from the perspective of emissions, it is almost never too soon to switch. —Jesse Smith
Abstract
Internal combustion engine (ICE) vehicles represent the single largest source of carbon dioxide emissions for most US households. Although battery electric vehicles (BEVs) considerably reduce use-phase emissions, the transition poses a fundamental life cycle optimization challenge: Is there a net climate benefit to retiring a functional ICE vehicle before its end of life? Retiring a functioning ICE vehicle incurs a fundamental trade-off between BEV manufacturing emissions and decreased vehicle operation emissions. In this study, we evaluated this trade-off across a comprehensive range of retirement scenarios and found that scrappage-and-replacement yields consistent emissions reductions across most contexts, including the retirement of new ICE vehicles. Although scrapping a functional asset remains economically prohibitive under current market conditions, these results demonstrate a major mitigation potential for policy interventions, such as enhanced scrappage subsidies, to address the emissions of an increasingly durable ICE fleet.
Personal vehicles are responsible for more CO2 emissions than all other forms of US transportation combined, including air (international and domestic), rail, water, and commercial trucking and busing (1). Annual emissions from these light-duty vehicles [cars, sport utility vehicles (SUVs), pickup trucks, and vans] are relatively evenly split between daily-driving and long-distance trips (2). Personal vehicle emissions are a primary mitigation target because transportation recently became the single largest sector of US emissions. Establishing a sustainable pathway for this fleet is not only essential for domestic climate goals but also serves as a model for developing economies experiencing rapid motorization. Yet the pace of this transition is fundamentally constrained by the physical and economic life span of the existing internal combustion engine (ICE) fleet.
One strategy to mitigate emissions is the electrification of vehicles. Life cycle studies show that battery electric vehicles (BEVs) can achieve large CO2 reductions relative to ICE vehicles (3–10). These studies account for the use phase (gasoline production and combustion for ICE vehicle; power plant emissions for BEVs) as well as the vehicle phase (emissions from manufacturing, material mining, and end of life). For today’s low-carbon electric grids, BEVs emit only one-third of the greenhouse gases as ICE vehicles. Even in regions that still rely heavily on coal power plants, BEVs still offer performance comparable to the most efficient gas-powered vehicles. Notably, the benefits of BEVs are expected to continue to improve owing to the growing capacity of renewable energy and the transition from coal to natural gas (6).
Although the emission benefits of choosing a BEV over an ICE vehicle during a new vehicle purchase are well established (3–6), a persistent question remains regarding the optimal time to retire (scrap) a fully operational ICE vehicle. Given the large existing stock of ICE vehicles and their increasing durability (11), it is essential to determine whether lower life cycle emissions would result from continued ICE vehicle use or from retiring (permanently scrapping) an ICE vehicle before it reaches its full operational life span and replacement with a BEV. Although the continued use of the ICE vehicle throughout its functional life would avoid additional BEV manufacturing emissions, retiring a functioning ICE vehicle would take advantage of the low use-phase emissions of the BEV. Previous work has established relatively low ratios of ICE vehicle operation emissions relative to manufacturing emissions, high production emissions in BEVs compared with ICE vehicles, and low operating emissions of BEVs (3–10). However, the net of these offsetting emissions has not yet been determined to reveal the optimal retirement age. In general, such questions concerning product retirement involve analysis of lifetime extensions and recycling that are particularly relevant to the material efficiency of circular economy policies such as vehicle scrappage incentives (12–16).
Although the optimal time to replace an ICE vehicle with a BEV has not been studied, previous research has explored related questions of replacing an older ICE vehicle with a newer, more efficient ICE vehicle. Such studies were motivated by vehicle scrappage policies (e.g., Cash for Clunkers), which were designed to remove high-polluting vehicles from the roads. Studies of life cycle emissions for vehicle retirement in the US and Canada found that the original ICE vehicle should be operated for 18 years before replacement with the newer ICE vehicle if the goal is to minimize greenhouse gas emissions (17, 18). This 18-year replacement interval is similar to the vehicle lifetime, suggesting that retiring a functioning ICE vehicle for a more efficient ICE vehicle provides no greenhouse gas (GHG) benefits.
These studies did not show GHG advantages for ICE vehicle retirement, but the results cannot be directly extrapolated to the present question for retiring an ICE vehicle with BEV replacement. In particular, use-phase improvements (vehicle efficiency) for the replacement ICE vehicle were small in previous studies. By contrast, the use-phase improvements for BEVs relative to ICE vehicles are generally large (8). Furthermore, the manufacturing emissions of BEVs are considerably higher than those for ICE vehicles (19, 20). Determining the retirement interval needed to minimize CO2 emissions would require a new analysis of the wide range of life cycle emissions associated with both ICE vehicles and BEV production and use. Such an analysis could provide timely information for the design of scrap-and-replace policies that promote BEV adoption (e.g., Clean Cars 4 All, California; ULEZ Scrappage Scheme, UK) (21, 22).
To address the question of ICE vehicle retirement, here we estimate the life cycle CO2 emissions and assess the variability in these results with respect to a wide range of vehicle efficiencies, electricity grids, manufacturing emissions, battery sizes, and vehicle miles traveled. We characterize the variability through a continuous range of these parameters as well as through a discrete analysis based on production-weighted averages and class-leading models by efficiency and sales. The emission benefit of retirement is assessed as the percent difference in emissions between scenarios of continuous operation of an ICE vehicle and scenarios in which the ICE vehicle is retired (permanently scrapped) and replaced with a BEV. Furthermore, we explore specific inflection points at which the early retirement benefit is minimized or reversed to map the boundaries of the mitigation benefit.
Emissions reduction from retirement of ICE vehicle
To evaluate the emissions benefits of ICE vehicle retirement, we first analyze three representative pathways for the production-weighted average SUV on the average US grid over a 16-year vehicle lifetime. In the baseline scenario, the ICE vehicle is operated for its full life span (Fig. 1A). We compare this to two retirement-via-scrappage pathways (year 11 retirement, Fig. 1B; year 2 retirement, Fig. 1C). Although both retirement scenarios incur a transient increase in emissions arising from BEV manufacturing, this carbon debt is offset through lower use-phase emissions. Notably, the earlier retirement scenario yields the greatest benefit, achieving a 44% reduction in cumulative emissions with a carbon payback period of 3 years (Fig. 1D). Because the ICE vehicle manufacturing emissions are common to all pathways (mathematically sunk costs), the relative advantage of the retirement scenarios is derived solely from the forward-looking delta between BEV adoption (operation and production) and the continued ICE vehicle operation.
Fig. 1. Life cycle emission comparisons between baseline (continued ICE vehicle operation) and retirement scenarios.
Emissions include use phase (tailpipe and fuel cycle for ICE vehicle; electricity generation for BEV) and vehicle phase (manufacturing and mining). (A) Scenario 1: The ICE vehicle is operated for its full design life. (B) Scenario 2: The ICE vehicle is scrapped at year 11 and replaced with a BEV. (C) Scenario 3: The ICE vehicle is scrapped at year 2 and replaced with a BEV. (D) Cumulative use and vehicle-phase emissions. The ICE vehicle production emissions are sunk costs because they are common to both the baseline and retirement pathways. Values reflect a sales-weighted average SUV, US average grid mix, and time-varying annual mileage. Comprehensive sensitivity analyses for all parameters are provided in the subsequent figures.
Continuous parameter sensitivity
We next explore emission benefits across the full extent of vehicle efficiencies and grid parameters (Fig. 2). Variations in three critical factors drive the observed retirement emissions benefit: vehicle efficiency, electricity grid carbon intensity, and BEV production emissions. Here we consider the retirement of ICE vehicles and hybrid electric vehicles (HEVs) with BEV replacement; plug-in hybrid electric vehicles (PHEVs) are addressed separately below. Considering the complete range of US vehicle efficiencies (Fig. 2A) introduces the highest degree of variability in the retirement benefit, which ranges from an 82% mitigation of emissions to a 77% increase in emissions. Notably, 92% of the modeled vehicle scenarios achieve a net reduction in emissions, with fleet-average efficiencies (gray lines) yielding a 58% retirement benefit (see tables S5 to S7 for specific model comparisons). Variability in the charging electricity source also reveals boundary cases in which retirement results in a net emission increase; the retirement benefit can be negated when BEV electricity consumption exceeds 30 kWh 100 km−1 and grid emission surpasses 500 kg MWh−1 (Fig. 2B). Additional boundary cases are explored in the discrete comparisons that follow. Although BEV production emissions are sensitive to battery size, manufacturing electricity source, and assembly processes (Fig. 2C), these vehicle-phase parameters contribute less than half as much variability to the total retirement benefit as vehicle operational efficiencies.
Fig. 2. Cumulative emission benefit of retirement relative to a comprehensive range of vehicle efficiency, grid carbon intensity, and manufacturing emissions.
The benefit represents the difference between cumulative emissions in continuous petroleum-fueled scenarios (inclusive of ICE vehicles and HEVs) and retirement scenarios occurring at year 1 (normalized by use-phase emissions). Emissions include the operational phase (tailpipe and refining and upstream for petroleum-fueled vehicles; power plant and upstream for BEVs) and the BEV production phase (manufacturing and material acquisition). Parameter ranges incorporate the full spectrum of efficiencies (A) reported by the US EPA, specifically using high-efficiency hybrid electric vehicle (HEV) benchmarks for the minimum consumption bound, as well as the range of grid emissions (B) and BEV production emissions associated with varying battery sizes and manufacturing intensities (C). The retirement benefit is independent of ICE vehicle and HEV manufacturing, which is mathematically a sunk cost. Gray lines indicate US sales-weighted mean parameter values.
Discrete parameter sensitivity of high-volume vehicles
The continuous analysis identifies critical thresholds across the theoretical efficiency spectrum; we supplemented this with discrete scenarios to evaluate the retirement benefits of production-weighted averages against both class-leading efficient vehicles and the highest selling vehicles (Fig. 3). This discrete approach also enables the assessment of highly comparable vehicle pairings (e.g., a Ford F-150 Lightning BEV replacing an ICE F-150).
Fig. 3. Sensitivity of the retirement benefit (B) to discrete parameter settings.
The cumulative emissions benefit over the vehicle lifetime when retirement occurs at year 1, incorporating all production- and use-phase impacts (truck, purple; SUV, pink; car, yellow). (A to C) Efficiency variants: Sensitivity is evaluated for production-weighted class averages (A), class-leading efficient vehicles (B), and the highest-selling vehicles (C). The latter highlights the climate impact of current consumer behavior, in which a small number of high-volume models dominate the market share. (D to G) Vehicle and technology variants: Sensitivities for low (D) and high (E) battery production intensities, baseline battery capacity (F), and the retirement of plug-in hybrid electric vehicles [PHEVs; (G)]. (H to J) Infrastructure variants: Impact of regional grid intensities across the range of US subgrids (see supplementary materials for more details). All results are invariant to production emissions of the petroleum-fueled vehicles, which function as mathematically sunk costs that cancel out in all comparative calculations. Where applicable [(D) to (J)], open markers indicate the retirement of class-leading HEVs, while closed markers represent retirement cases for the highest-selling ICE vehicles. Error bars represent 95% confidence intervals derived from the propagation of error.
Although production-weighted averages and market-dominant models yield similar retirement benefits (55, 57, and 55% averages for cars, SUVs, and trucks, respectively), focusing on the class-leading efficient vehicles (e.g., Toyota Prius and Tesla Model 3) decreases the mitigation potential (Fig. 3, A to C). The reduction in retirement benefits for the most efficient vehicle models are 56, 34, and 23% for cars, SUVs, and trucks, respectively. This variance occurs because even though high-selling BEV models are often among the most efficient in their class, the broader petroleum-fueled fleet distribution includes a larger proportion of low-efficiency models.
Given the impact of battery manufacturing on vehicle-phase emissions, we examined the sensitivity of the benefit to varying battery production intensities and capacities. A range of manufacturing emission factors (52 to 173 kg CO2eq kWh−1; Fig. 3D) resulted in a 13% variation in the net benefit. Notably, the retirement benefit approaches zero in the worst-case permutation: replacing a class-leading HEV while assuming the maximum battery production emission factor. Furthermore, although our core analysis utilizes the conservative assumption of extended-range batteries, accounting for a baseline battery capacity results in only a 3% improvement in the retirement benefit (Fig. 3E).
Retiring the PHEV for a BEV provides a consistent boundary case (Fig. 3G). The benefit potential is near zero for SUVs, while cars result in an 11% increase in emissions. Although these results depend on the PHEV utility factor estimated by the Environmental Protection Agency (EPA), recent work also suggests that charge-depleting mode is underutilized in real-world PHEV driving (23).
Furthermore, we account for the substantial geographic variability in the carbon intensity of regional electrical grids (Fig. 3, H to J). Whereas the retirement benefit is maintained across all US subgrids for high-selling models, the retirement of class-leading efficient HEVs yields marginal benefits or a net increase in emissions in regions where grid emission factors exceed 400 kg MWh−1. Notably, subgrids meeting this high-intensity definition currently account for only 33% of total US net generation (see table S3 for subgrid definitions). This finding suggests that for most of the current US power mix, the emissions reduction from early retirement is robust even when replacing the most fuel-efficient petroleum-fueled vehicles.
The retirement benefit estimates in Fig. 3 are derived using the use- and vehicle-phase emissions from CarbonCounter. Here we examine the robustness of these estimates using the alternative emissions dataset of Santero et al. (8), which consists of 459 vehicle models from the EPA Automotive Trends Report (Fig. 4). Methodological variations between the two datasets include distinct city and highway driving cycles, assumed vehicle miles traveled, and material inventories. Despite these differences, the production-weighted averages of the retirement benefits using the Santero et al. data yield consistent results of 55, 57, and 47% for cars, SUVs, and trucks, respectively. This strong alignment across independent datasets reinforces the conclusion that the climate benefit of early retirement is primarily driven by the fundamental efficiency gap between petroleum-fueled and electric drivetrains, rather than specific life cycle modeling assumptions.
Fig. 4. Distributions of retirement benefits (B) across varying annual utilization rates.
Benefits are calculated using use- and vehicle-phase emissions from the Santero et al. (8) dataset. BEV efficiencies are derived from EPA Automotive Trends data for cars (A), SUVs (E), and trucks (I). MPGe, miles per gallon equivalent. Retirement benefits are estimated for low-utilization [8045 km yr−1; (B), (F), and (J)], average-utilization (18,289 km yr−1; (C), (G), and (K), and high-utilization [46,661 km yr−1; (D), (H), and (L)] scenarios for each vehicle model. Continuous analysis of annual mileage (M) for production-weighted average emissions reveals the break-even thresholds where the carbon debt of BEV manufacturing is offset by operational savings.
Mileage thresholds
Although the preceding analysis uses average annual mileages, we also evaluate a range of driving rates to account for both high-utilization duty cycles (e.g., taxi service) and secondary, low-usage vehicles (Fig. 4M). The break-even annual mileage (threshold below which the carbon debt of BEV manufacturing is not recovered) is 7054, 6837, and 10,794 km for cars, SUVs, and trucks, respectively. Notably, these thresholds represent only 35 to 54% of an average 20,000-km driving distance. This indicates that for most of the fleet, even many vehicles driven less than the national average can still yield a net climate benefit upon retirement.
ICE vehicle resale and market expansion
If BEV adoption is associated with resale of ICE vehicles instead of retirement, the influx of used vehicles to the market could reduce used vehicle costs and induce a modal shift from low-emission transport (e.g., mass transit). To examine this modal shift effect, we evaluate a displacement factor D to assess the break-even point where electrification no longer provides a benefit (D ≤ 1; D = 1, one-to-one displacement of used ICE vehicles for low-emission transport; D = 0, no displacement). This bounding analysis reveals that the net benefit is eliminated when the displacement factor exceeds 81, 65, and 42%, for the minimum, average, and maximum grid intensities, respectively (Fig. 5).
Fig. 5. Critical displacement thresholds for transportation modal shift.
If the ICE vehicle is resold rather than retired, a decrease in used-vehicle prices may induce a modal shift from low-emission transit to personal vehicle use. The critical displacement factor D represents the threshold at which induced emissions from this modal shift completely offset the net BEV adoption benefit. (A to C) Grid intensity scenarios: Variability in the displacement threshold is explored across three grid intensities—110 (A), 347 (B), and 699 g CO2 eq kWh−1 (C)—as a function of BEV efficiency (x axis) and ICE vehicle/HEV efficiency (y axis). These scenarios assume a shift from zero-emissions modes (e.g., active transport); sensitivity to higher-intensity modes (e.g., mass transit) is provided in the supplementary materials. In regions where the BEV life cycle intensity exceeds that of the ICE vehicle, the critical displacement factor is negative (brown shading), representing an undefined with respect to the modal shift effect. Gray lines denote production-weighted averages for the current US fleet.
Beyond modal shift, economic literature suggests further market-mediated effects. If the resale of ICE vehicles reduces used vehicle costs, induced modal shift emissions could be mitigated by an increase in fleet-wide scrappage rates for the oldest and potentially least efficient ICE vehicles in the existing fleet (24, 25). Conversely, BEV policies may cause an increase in used vehicles costs (26), reversing both the modal shift effect and the fleet-turnover benefit.
Although the market-mediated effects provide prognostic information on fleet turnover, the primary focus of this study is on life cycle assessment for a diagnostic purpose: to quantify the emissions for scrappage-and-replacement. Such scenarios are directly relevant to existing policies, such as the Clean Cars 4 All program—which offers rebates for scrapping functional ICE vehicles—and may become more critical as climate targets necessitate the accelerated decommissioning of the petroleum-fueled fleet.
Discussion
In this study, we found that retirement of a functional ICE vehicle for a BEV, even when the ICE vehicle is brand new, provides large carbon mitigation benefits across a wide range of scenarios. Given these results and the outsized impact of transportation emissions on the US carbon budget, policy-makers could achieve substantial GHG mitigation through investments in vehicle scrap-and-replace programs. Optimization of such policies would require design to account for both vehicle efficiency and regional electric grid emissions. Moreover, the consistent GHG advantage of BEVs demonstrated here suggests that further innovation should not only prioritize BEV sustainability but also overcoming operational challenges, such as charger network reliability and equitable access (27–29).
Our results are not directly comparable with the conclusions of previous studies that found that the lifetime of an ICE vehicle should be extended if the goal is to minimize carbon emissions. A critical element of previous work is the relatively small reduction in use-phase emissions between the original ICE vehicle and the replacement ICE vehicle. BEVs were not considered a viable option at the time of past retirement analyses. However, electrifying vehicle powertrains provides an order-of-magnitude-larger reduction in use-phase emissions than the replacement ICE vehicles considered previously, which leads to the retirement recommendation suggested by the present analysis.
Uncertainty is associated with the future development of the electric grid and liquid transportation fuels. Our analysis focused on data for the present emissions (e.g., existing vehicles, current electric grid). Although we did not attempt to forecast future trends, these may further the advantage for the BEV. In particular, the trends to develop low-carbon fuels for ICE vehicles have been hampered by technological and land-use constraints, while the pace of renewable electricity for BEVs has rapidly progressed (7, 30).
Although we considered a wide range of scenarios, there are some model parameters that could not be evaluated with the available information owing to the current stage of BEV development. The global BEV fleet has grown to 40 million, but the vehicles are still relatively new, and the batteries are lasting longer than anticipated. The result of these two factors is that the battery-recycling industry is still in its infancy because of a low stock of retired BEV batteries. Although we did not account for battery recycling, these processes are likely to further increase the carbon savings for the BEV because recycling offsets the energy required to acquire raw materials for batteries. In particular, the energy inputs to the recycling processes are much lower than the energy inputs to obtain raw materials (4). Future development of more-sustainable battery chemistry (e.g., low-cobalt and cobalt-free) may also alter the BEV vehicle-phase emissions (31).
Current scrap-and-replace rebates are insufficient to make the retirement of newer ICE vehicles financially desirable. Nevertheless, we examined the full range of scrappage scenarios (including new ICE vehicles) to demonstrate the theoretical upper-boundary conditions for emissions benefits when an ICE vehicle is scrapped with a BEV replacement. Newer vehicle scrappage is economically prohibitive with existing scrap-and-replace subsidies, necessitating the use of these limited resources for only the least-efficient ICE vehicles in the fleet. However, as climate impacts accelerate and the ICE vehicle fleet becomes more durable, a major shift in policy to support vehicle retirement may become necessary. The results presented here demonstrate that such a policy shift would result in emissions reductions across most contexts.
Acknowledgments
We thank L. Jackson, S. Perkins, J. Khurana, and J. West for helpful discussions and data collection.
Funding:
This work was supported by the US NSF no. 1931667 (to J.E.C).
Author contributions:
Conceptualization, Data curation, Formal analysis: J.E.C.; Investigation, Methodology, Software, Visualization, Writing – original draft, Formal analysis, Investigation, Methodology, Software, Visualization, Writing – review & editing: R.G.
Competing interests:
The authors declare that they have no competing interests.
Data, code, and materials availability:
All data are available in the main text or the supplementary materials. No new materials were generated for this study.
License information:
Supplementary Materials
This PDF file includes:
References and Notes
2
US Department of Transportation, Federal Highway Administration, Office of Highway Policy Information, “National Household Travel Survey” (2026); https://nhts.ornl.gov/.
3
M. Miotti, G. J. Supran, E. J. Kim, J. E. Trancik, Personal vehicles evaluated against climate change mitigation targets. Environ. Sci. Technol. 50, 10795–10804 (2016).
4
M. Gutsch, J. Leker, Costs, carbon footprint, and environmental impacts of lithium-ion batteries–From cathode active material synthesis to cell manufacturing and recycling. Appl. Energy 353, 122132 (2024).
5
B. Cox, C. Bauer, A Mendoza Beltran, D. P. Van Vuuren, C. L. Mutel, Life cycle environmental and cost comparison of current and future passenger cars under different energy scenarios. Appl. Energy 269, 115021 (2020).
6
Z. A. Needell, J. McNerney, M. T. Chang, J. E. Trancik, Potential for widespread electrification of personal vehicle travel in the United States. Nat. Energy 1, 16112 (2016).
7
J. E. Campbell, D. B. Lobell, C. B. Field, Greater transportation energy and GHG offsets from bioelectricity than ethanol. Science 324, 1055–1057 (2009).
8
N. Santero, L. Nelson, Y. Chen, M. Meredith, P. Busch, A. Kendall, Electrifying light vehicles in the United States shows emission reduction potential for all vehicle types and powertrains. Commun. Sustain. 1, 23 (2026).
9
L. B. Lave, H. L. MacLean, An environmental-economic evaluation of hybrid electric vehicles: Toyota’s Prius vs. its conventional internal combustion engine Corolla. Transp. Res. Part D Transp. Environ. 7, 155–162 (2002).
10
J. L. Sullivan, A. Burnham, M. Wang, “Energy-consumption and carbon-emission analysis of vehicle and component manufacturing” (Argonne National Laboratory, Energy Systems Division, Center for Transportation Research, 2010); https://publications.anl.gov/anlpubs/2010/10/68288.pdf.
12
E. G. Hertwich, S. Ali, L. Ciacci, T. Fishman, N. Heeren, E. Masanet, F. N. Asghari, E. Olivetti, S. Pauliuk, Q. Tu, P. Wolfram, Material efficiency strategies to reducing greenhouse gas emissions associated with buildings, vehicles, and electronics—A review. Environ. Res. Lett. 14, 043004 (2019).
13
A. M. King, S. C. Burgess, W. Ijomah, C. A. McMahon, Reducing waste: Repair, recondition, remanufacture or recycle? Sustain. Dev. 14, 257–267 (2006).
14
R. Geyer, B. Kuczenski, T. Zink, A. Henderson, Common misconceptions about recycling. J. Ind. Ecol. 20, 1010–1017 (2016).
15
D. R. Cooper, T. G. Gutowski, The environmental impacts of reuse: A review. J. Ind. Ecol. 21, 38–56 (2017).
16
E. Muehlegger, D. S. Rapson, Subsidizing low-and middle-income adoption of electric vehicles: Quasi-experimental evidence from California. J. Public Econ. 216, 104752 (2022).
17
H. C. Kim, G. A. Keoleian, D. E. Grande, J. C. Bean, Life cycle optimization of automobile replacement: Model and application. Environ. Sci. Technol. 37, 5407–5413 (2003).
18
D. V. Spitzley, D. E. Grande, G. A. Keoleian, H. C. Kim, Life cycle optimization of ownership costs and emissions reduction in US vehicle retirement decisions. Transp. Res. Part D Transp. Environ. 10, 161–175 (2005).
19
P. Wolfram, T. Wiedmann, Electrifying Australian transport: Hybrid life cycle analysis of a transition to electric light-duty vehicles and renewable electricity. Appl. Energy 206, 531–540 (2017).
20
M. Li, J. Lu, Z. Chen, K. Amine, 30 years of lithium-ion batteries. Adv. Mater. 30, e1800561 (2018).
21
E. Figenbaum, Perspectives on Norway’s supercharged electric vehicle policy. Environ. Innov. Soc. Transit. 25, 14–34 (2017).
22
J. H. Langbroek, J. P. Franklin, Y. O. Susilo, The effect of policy incentives on electric vehicle adoption. Energy Policy 94, 94–103 (2016).
23
A. Mandev, P. Plötz, F. Sprei, Factors impacting real-world fuel economy of plug-in hybrid electric vehicles in Europe–an empirical analysis. Environ. Res. Commun. 6, 051001 (2024).
24
M. R. Jacobsen, A. A. Van Benthem, Vehicle scrappage and gasoline policy. Am. Econ. Rev. 105, 1312–1338 (2015).
25
H. K. Gruenspecht, Differentiated regulation: The case of auto emissions standards. Am. Econ. Rev. 72, 328–331 (1982).
27
D. Keith, J. Womack, Building and sustaining reliable public EV charging in the United States. Environ. Res. Lett. 18, 011004 (2023).
28
D. Rempel, C. Cullen, M. Matteson Bryan, G. Vianna Cezar, Reliability of open public electric vehicle direct current fast chargers. Hum. Factors 66, 2528–2538 (2024).
29
Q. Yu, T. Que, L. J. Cushing, G. Pierce, K. Shen, M. Kejriwal, Y. Yao, Y. Zhu, Equity and reliability of public electric vehicle charging stations in the United States. Nat. Commun. 16, 5291 (2025).
30
P. Halder, K. Azad, S. Shah, E. Sarker, “Prospects and technological advancement of cellulosic bioethanol ecofuel production” in Advances in Eco-Fuels for a Sustainable Environment, K, Azad, Ed. (Elsevier, 2019), pp. 211–236.
31
S. Mallick, A. Patel, M. P. Paranthaman, J. H. Mugumya, S. Kim, M. L. Rasche, M. Jiang, H. Lopez, R. B. Gupta, An overview of various critical aspects of low-cobalt/cobalt-free Li-ion battery cathodes. Sustain. Energy Fuels 9, 724–738 (2025).
32
F. Burlig, J. Bushnell, D. Rapson, C. Wolfram, “Low energy: Estimating electric vehicle electricity use,” AEA Papers and Proceedings, vol. 111 (American Economic Association, 2021), pp. 430–435.
33
B. L. Cox, C. L. Mutel, The environmental and cost performance of current and future motorcycles. Appl. Energy 212, 1013–1024 (2018).
35
L. A.-W. Ellingsen, C. R. Hung, A. H. Strømman, Identifying key assumptions and differences in life cycle assessment studies of lithium-ion traction batteries with focus on greenhouse gas emissions. Transp. Res. Part D Transp. Environ. 55, 82–90 (2017).
36
H. C. Kim, T. J. Wallington, R. Arsenault, C. Bae, S. Ahn, J. Lee, Cradle-to-gate emissions from a commercial electric vehicle Li-ion battery: A comparative analysis. Environ. Sci. Technol. 50, 7715–7722 (2016).
37
J. Linn, The rebound effect for passenger vehicles. Energy J. (Camb. Mass.) 37, 257–288 (2016).
38
X. Liu, A. Elgowainy, R. Vijayagopal, M. Wang, Well-to-Wheels Analysis of Zero-Emission Plug-In Battery Electric Vehicle Technology for Medium- and Heavy-Duty Trucks. Environ. Sci. Technol. 55, 538–546 (2021).
39
US Environmental Protection Agency, “Emissions & Generation Resource Integrated Database (eGRID),” 2023 data (EPA, 2026); https://www.epa.gov/egrid.
40
M. Wang, A. Elgowainy, U. Lee, P. Benavides, A. Burnham, H. Cai, Q. Dai, T. Hawkins, J. Kelly, H. Kwon, X. Liu, Z. Lu, L. Ou, P. Sun, O. Winjobi, H. Xu, “Summary of Expansions and Updates in GREET® 2019” (Argonne National Laboratory, Energy Systems Division, Systems Assessment Center, ANL/ESD–19/6, 2019); https://doi.org/10.2172/1569562.