10
Electricity as a Vehicle Fuel

Plug-in electric vehicles (PEVs) use energy stored in an onboard battery for propulsion and charge the battery using electricity from the power grid. PEVs include battery electric vehicles (BEVs), which rely entirely on energy from the power grid, and plug-in hybrid electric vehicles (PHEVs), which can power the vehicle from a mix of energy from the battery and from another fuel source, typically gasoline.

Because PEVs require substantial battery capacity, vehicle production emissions can differ from those of liquid fuel vehicles. Vehicle production emissions are discussed in Chapter 6. Here the focus is on the fuel: electricity.

When PEVs are charged, they add load to the power grid. This demand is satisfied by increased output from power plants, which has implications for air emissions and other power system impacts. Additionally, broad adoption of PEVs could add a large enough load to the power grid to trigger construction of new power plants in some regions (known as capacity expansion). The potential for flexible PEV charging profiles acting as demand response units could plausibly change the economic and logistical factors associated with increasing penetration of intermittent, non-dispatchable generators, like wind and solar (Weis et al., 2014). The focus here is primarily on grid emissions from operation of existing or projected future generators.

In this chapter the focus is on the consequential effects of PEV charging for power grid emissions. There are also other consequential effects of electric vehicle (EV) adoption and PEV use beyond the power sector, similar to biofuels, such as rebound effects and fuel market effects as well as potential land use changes. The focus here is on power sector effects because they are specific to PEVs.

This chapter begins by (1) comparing attributional life-cycle assessment (ALCA) and consequential life-cycle assessment (CLCA) of air emissions from electricity consumption; (2) providing an overview of approaches to estimating consequential emissions of PEV charging, including regression, simulation, proxies and real time data; and (3) summarizing several key issues, including upstream emissions, uncertainty and dynamics, energy efficiency, effects of public policy, and data sources.

COMPARING ATTRIBUTIONAL AND CONSEQUENTIAL LIFE-CYCLE ASSESSMENT FOR ELECTRICITY

Past life-cycle studies of EVs have accounted for power sector emissions using attributional or consequential methods. As discussed in Chapter 2, the two approaches are intended to answer different questions. In the context of PEVs:

• ALCA seeks to answer what emissions a PEV is associated with or responsible for, given some judgments about how to assign power grid emissions to demand sources.
• CLCA seeks to answer how emissions will change if a technology or policy is adopted given some judgments about how to predict future counterfactual scenarios.

Figure 10-1 provides a conceptual illustration of the difference between these questions and approaches for PEV charging. In attributional approaches, a portion of total power grid greenhouse gas (GHG) emissions are assigned to PEV charging. In contrast, in consequential approaches power grid emissions are estimated in two scenarios—one without PEV charging and one with PEV charging— and the difference between emissions in the two scenarios is the consequential effect of PEV charging.

LCA GHG estimates of PEVs can vary substantially as a result of the type of LCA estimate used and regional boundaries used, so it is important to understand the differences in methods and regional boundary choices to identify which questions each approach can answer (Ryan et al., 2016).

The most common attributional approach for estimating emissions from PEV charging is the simplest approach: compute the average emissions per unit of energy produced in the power system and assign this rate of emissions to PEV charging. Results from this approach can vary widely depending on boundary definitions (Weber et al., 2010). Some past U.S. studies assigned to PEV charging the average emission rate from power plants located in the country (Miotti et al., 2016) or the state (Yawitz et al., 2013) where the vehicle is charged. These boundaries are easy to identify, but because the grid is not generally organized around political boundaries, in most cases political boundaries have little to do with the impact of load at a given location. Most attributional studies use the average emissions from generators in regions defined by the power grid (Yuksel et al., 2016), rather than political boundaries, with the idea that load within one of these grid regions is more likely to affect generation within the region than across the boundary. However, electricity is constantly being traded across boundaries, and attributional assessments can vary widely, depending on how boundaries for analysis are chosen (Ryan et al., 2016; Weber et al., 2010).

Figure 10-3 provides a simplified conceptual illustration of why consequential emissions from PEV charging can differ from average grid emission rates. In this example, Region 1 has both coal and nuclear power plants, and Region 2 has only nuclear plants. In Region 1 the nuclear plants are fully utilized to meet existing load, and the coal plants are partly utilized (represented as height in red). If consumption is increased in Region 1 to charge a PEV, the nuclear plants cannot increase generation to supply that load, so the coal plants will increase generation. Although the average emissions in Region 1 are those associated with a mix of nuclear and coal generation, the effect of adding load to that region is the emissions associated with increasing coal generation. Although Region 2 contains only nuclear generators, they are already fully utilized and cannot increase power generation. If new PEV charging load is added to Region 2, it will need to increase trade with Region 1 in order to meet its demand. So, even though power generation in Region 2 is entirely from a zero-carbon source, the effect of charging a PEV in Region 2 may be to increase emissions from coal generators.

A related example raised frequently is PEV owners who have rooftop solar generation at their homes. Rooftop solar generation reduces emissions by displacing fossil fuel generation. However, whether or not a household has rooftop solar generation, adding a PEV does not (usually) increase the amount of solar power generated. Instead, adding PEV load will increase the amount of energy that the household demands from the power grid (or reduce the amount of rooftop solar sold to the grid), triggering increased generation from plants on the grid. Like Region 2 in the example above, even in a household with rooftop solar the effect of charging a PEV may be to increase generation at a fossil fuel plant.¹

These examples show why the emissions associated with a technology change (such as a household purchasing a PEV instead of a gasoline vehicle) or a policy change (e.g., a policy encouraging or mandating PEV adoption) can look quite different from the average power generation emissions in a region.

___________________

¹ It is worth noting that when decisions are coupled, such as comparing the adoption of PEV with rooftop solar to adoption of a gasoline vehicle and no solar, the difference in consequential emissions between the two scenarios involves both the effects of adding rooftop solar and the effects of replacing gasoline demand with electricity demand.

TABLE 10-1 Sample of Published Studies Assessing PEV Emissions in the United States

NOTE: Vehicle types: BEV = battery electric vehicle; CV= commercial vehicle; HEV = hybrid electric vehicle; ICEV= internal combustion engine vehicle; ICV = internal combustion vehicle; PHEV = plug-in-hybrid-electric vehicle. Regional resolution: NERC = North American Electric Reliability Corporation; eGRID = EPA’s Emissions & Generation Resource Integrated Database; US-REGEN = U.S. Regional Economy, Greenhouse Gas, and Energy; RTO = regional transmission organization.

Because the power grid is highly interconnected in many regions, with many generators adding to the system and many demand sources drawing from the system, it is generally not possible to know precisely which power plants increase generation in response to increased load, especially when projecting into the future, such as over the life of a PEV. When assigning emissions to a PEV, many researchers and advocates with an attributional view find that it seems unfair to assign certain grid emissions to existing loads and different grid emissions to new loads, so it is common in ALCA to assign equal emission rates to all loads. Such allocations do not estimate how a technology or policy will change net emissions. To answer these questions, CLCA is needed.

With CLCA no assignment of emissions is made. Rather, the question of interest is how emissions will change if more PEVs are adopted (and charged). The effect of adding PEV charging load to the power grid depends on when and where the load is added. Figure 10-4 shows a hypothetical dispatch curve, which orders power plants available for dispatch based on their marginal generation cost. In the early morning hours in this example, the existing load is 67 GW. Adding new load at this time will increase generation from the specific generators on the margin (located just above the 67 GW load line): the natural gas combined cycle (red). In the afternoon on a hot day, the existing load is 114 GW, so adding new load at this time will increase generation from plants next in the dispatch order: natural gas and other (yellow). If fossil generation assets were to be replaced by dispatchable low-carbon energy sources, the addition of PEV charging demand would not create additional fossil fuel emissions from the electricity system.

APPROACHES TO CONSEQUENTIAL LIFE-CYCLE ASSESSMENT FOR PLUG-IN VEHICLE CHARGING EMISSIONS

CLCA approaches attempt to estimate emissions from the plants that will change generation in response to a change in load. A dispatch curve like the one in Figure 10-4 helps to visualize why this answer may change with time and location. In practice there are a number of factors that make operations more complicated than a simple dispatch curve, including transmission constraints, regulations, ramp rate limits, and other factors.

In broad terms, CLCA approaches to estimating PEV emissions fall into four main categories: regression, simulation, proxies, and real-time data.

Regression

Regression approaches use past data on power grid operations and emissions to statistically estimate how a marginal change in load affects emissions output across different operating conditions (Ryan et al., 2016). Figure 10-5 provides an example—for the Midwest Reliability Organization (MRO), which covers the Midwest region² in which marginal generation is estimated to be primarily from coal plants during low demand hours but with more from gas plants at high demand hours, with implications for emissions.

The main advantage of regression-based approaches is that they are based on real data about how the power system has operated. The main limitations are that they typically can only model the effect of marginal changes in load and they only look backward at how the power grid worked when the data were collected, making it difficult to predict how future technologies or policies might affect future loads and emissions. Regression approaches vary, and each has its advantages and limitations. There are two main approaches, one that uses total generation in a region and one that uses total consumption.

The approach originally proposed by Siler-Evans et al. (2012) uses total generation in a region as the independent variable and considers only dispatchable plants (fossil fuel plants that can change generation on demand). An advantage of this approach is that it avoids a potential correlation/causality confusion of counting temporary changes in generation timing from hydroelectric plants as consequential emissions. (Unlike fossil plants, a hydroelectric plant has finite supply limited by lake levels, so increasing generation at one time to satisfy demand reduces the supply available to generate electricity at a future time). A disadvantage is that it ignores marginal trade across regions by focusing on regional generation rather than regional consumption, and assumptions are needed to translate marginal consumption to marginal generation.

___________________

² The region covers the provinces of Saskatchewan and Manitoba, and all or parts of the states of Arkansas, Illinois, Iowa, Kansas, Louisiana, Michigan, Minnesota, Missouri, Montana, Nebraska, New Mexico, North Dakota, Oklahoma, South Dakota, Texas, and Wisconsin.

In contrast, the approach originally proposed by Graff Zivin et al. (2014) uses total consumption in a region as the independent variable and considers all plants in the broader interconnect. An advantage of this approach is that it measures the relationship between consumption and generation directly and allows for trade across regions at the margin within an interconnected area. A disadvantage is that it can conflate increased generation with shifted generation timing for hydroelectric plants.

Simulation

Simulation approaches model the power grid mathematically. Figure 10-6 show simulations of operations under scenarios that include and exclude PEV load, observing the difference in emissions across the two scenarios. These approaches typically model the power system as optimally satisfying load at minimum cost subject to practical constraints, such as transmission constraints, ramp rate limits, and capacity constraints. The main advantage of simulation-based approaches is that they can be used to study future scenarios or large changes in load. The main limitation is that it is difficult for a model to capture all of the factors that might affect grid operations in practice, and therefore there is generally some expected deviation between what an idealized model predicts and what would happen in practice. In particular, models that make more simplifying assumptions, such as using simple dispatch ordering without constraints on transmission or generation, can typically model larger systems at some expense of fidelity, while models that include detailed operational constraints typically limit scope to a particular region and may therefore miss effects of PEV load on marginal trade with other regions.

Proxies

Some analyses use proxies to approximate marginal emissions. For example, the U.S. Environmental Protection Agency’s (EPA) Emissions & Generation Resource Integrated Database (eGRID) data³ provide estimates of non-baseload generation by region, and non-baseload generation is sometimes used as a proxy for marginal generation.

___________________

³ See https://www.epa.gov/egrid.

Real-Time Data

Regression and simulation models as well as non-baseload eGRID data are ultimately based on actual generation resources dispatch and interchange data. This is the function that the regional transmission operators (RTOs) and independent system operators (ISOs) are performing (Greer, 2012):

Each of the RTOs/ISOs has its own control area where they are responsible for operating the electric grid reliably. There are currently seven RTOs or ISOs in the United States. Their names imply but are not limited to their general regional coverage:

• Pennsylvania New Jersey Maryland Interconnection (PJM).
• Midcontinent Independent System Operator (formerly, Midwest ISO) (MISO).
• Electric Reliability Council of Texas (ERCOT).
• California ISO (CAISO).
• New York ISO (NYISO).
• Southwest Power Pool (SPP).
• ISO New England (ISO-NE).

Several RTOs and ISOs publish real-time market and dispatch data as well as real-time electricity interchanges with other control areas. If not already publicly available, RTOs and ISOs can provide real-time marginal data, which can potentially be combined with data on EV charging patterns to estimate consequential emissions of EV charging.

Table 10-2 summarizes the advantages and disadvantages of regression-based, simulation-based, proxy, and real-time CLCA approaches and compares them with ALCA approaches that use average grid emissions. While regression-based approaches have advantages for shorter term analyses with incremental changes to operations, simulation-based approaches are likely needed to understand the impacts of PEVs over typical vehicle lifetimes, given the substantial changes to the grid that would be required to achieve climate stabilization targets and the potential for future feedstock prices and policy to change dispatch order. Proxies can be easy to use, though how well they estimate marginal emissions can vary, and real-time marginal emission estimates can provide observed dispatch information but may not, on their own, provide a basis for projecting future scenarios. Ryan et al. (2016) summarize additional recommendations for appropriate models estimating power-sector emissions.

TABLE 10-2 Comparison of Approaches to Estimate Grid Emissions from PEV Charging

NOTES: ALCA = attributional life-cycle assessment; CLCA = consequential life-cycle assessment; ISOs = independent system operators; RTOs = regional transmission operators.

In the context of a low-carbon fuel standard (LCFS) policy, for regulatory impact assessment (see Chapter 3), consequential grid emissions in future grid scenarios are needed, so simulation-based CLCA approaches may be most appropriate. For verification (see Chapter 5), real-time marginal emission factors from RTOs and ISOs have the potential to provide useful data. For carbon intensities assigned to fuels, there is no single agreed-upon estimate for PEVs. The choice depends on a policymaker’s goals, and there are different views, both in the research community broadly and among the members of this committee.

UPSTREAM EMISSIONS

The emissions consequences of increasing electricity demand are not limited to emissions from combustion at the power plants that increase generation to serve that load. Power plants also have upstream emissions from feedstock production, processing, and transport. For example, increasing generation from coal-fired power plants implies increased coal consumption, which triggers additional emissions from coal mining, coal transportation, and other supply chain activities. Attributional estimates of these upstream emissions have been used in ALCA studies (Anair and Mahmassani, 2012; Elgowainy et al., 2018; Kawamoto et al., 2019; MacPherson et al., 2012; Miotti et al., 2016; Nealer et al., 2015; Tong et al., 2020; Yawitz et al., 2013) and in policy.⁴ However, consequential or marginal emissions from changes in these supply chain activities are not well characterized in the literature, and these emissions are typically ignored or average attributional emissions estimates are used as a proxy. Upstream emissions can vary, but some studies have estimated upstream GHG emissions as 5–10 percent of electricity GHG emissions (Michalek et al., 2011). In addition, land use implications have not been well characterized.

Emissions from sources upstream of power plants are omitted in some electricity LCA studies and included with ALCA estimates in other LCA studies. CLCA estimates of how emissions upstream of power plants change with generation are generally not available in the literature.

UNCERTAINTY AND DYNAMICS

Consistent with the general findings about uncertainty in LCA (see Chapter 4), all approaches to estimating the power grid emissions consequences of PEV charging involve some degree of uncertainty. Furthermore, because the power grid will change over time in ways that cannot be fully predicted, including during the life of a vehicle, consequential emissions have important dynamic sources of uncertainty. For example, future marginal emissions may look different if the price of coal drops and the price of natural gas increases than they do if the price of natural gas drops and the price of coal increases (Weis et al., 2016). Many such factors can affect the emissions consequences of future PEV charging.

In sum, the emissions consequences of PEV charging are inherently uncertain, and effects of future PEV charging depend on future factors that cannot now be known.

One key potential benefit of PEVs is the potential to make electricity from low emissions sources, such as wind, solar, hydro, or nuclear power. Such sources typically do not operate on the margin in the United States today, but if the capacity of these sources increases in the future such that renewable sources are routinely curtailed, renewable generators could be on the margin in a future power grid, implying low

___________________

⁴ See https://ww2.arb.ca.gov/our-work/programs/low-carbon-fuel-standard.

consequential emissions from increasing the EV charging load. Several governments have announced targets to substantially increase renewable generation in the coming decades. The California LCFS policy provides incentives and alternative crediting for smart charging timed to coincide with low-emission grid composition, though low emission average composition does not necessarily imply low marginal emissions. Changes in power grid emissions caused by PEV charging could be low if PEV charging coincides with times when renewables would otherwise be curtailed.

The committee notes that studies that examine GHG emissions in isolation may miss co-benefits or tradeoffs with other externalities that can be larger in magnitude. For example, studies have found that the change in external costs of health effects from conventional air pollution can be larger than the change in external costs from GHG emissions when gasoline vehicles are replaced by PEVs (Michalek et al., 2011; Tessum et al., 2014; Weiss et al., 2016).

ENERGY EFFICIENCY

As discussed in Chapter 6, the life-cycle emissions of transportation fuels cannot be fully understood in isolation from the vehicles that use them. In particular, vehicle efficiency affects how much fuel must be consumed to serve a given travel need. PEV efficiency can be substantially affected by a number of factors including:

1. Which specific PEV design is being studied: BEVs in 2021 ranged in efficiency from 24 kWh/100 mi for the Tesla Model 3 to 50 kWh/100 mi for the Porsche Taycan Turbo S.⁵
2. Driving conditions: In city driving conditions with frequent stops, PEVs have substantial efficiency benefits over gasoline vehicles, but for highway cruising, PEVs are more comparable to gasoline vehicles (Karabasoglu and Michalek, 2013; Lee et al., 2017).
3. Climate: Both BEVs and gasoline vehicles are less efficient in cold weather, but BEVs typically experience a greater efficiency loss, in part because, unlike gasoline vehicles that use waste heat from the engine to heat the cabin, BEVs must use energy from the battery to heat the cabin instead of propelling the vehicle. BEVs can lose half of their range in extreme hot and cold weather climates (Lee and Thomas, 2017; Yuksel et al., 2016; Yuksel and Michalek, 2015). Other climate-related factors, such as humidity and precipitation, can also have substantial effects on vehicle efficiency.

The effects of these sources of heterogeneity, in addition to the effects of location and charge timing on grid emissions, can be larger than the differences in emissions among vehicle technologies. Therefore, studies that select a single vehicle design to represent each technology, a single estimate of grid emissions, and a single set of assumptions about charging, driving, and climate conditions have significant limitations. Variation in these factors can qualitatively affect the outcome of an LCA, as illustrated in Figure 10-7.

DATA SOURCES FOR RESEARCH

There are various data sources that researchers can use to estimate marginal emissions and externalities of electricity consumption (to estimate consequential emissions from PEV charging): two key ones come from EPA and the Center for Climate and Energy Decision Making at Carnegie Mellon University.

___________________

⁵ See https://www.fueleconomy.gov/.

• EPA maintains estimates from two models: eGRID⁶ and AVERT.⁷ AVERT, in particular, supports calculation of marginal emissions factors on a regional, state, and county-levels. The user manual of AVERT (EPA, 2020) states: “within each region across the country, system operators decide when, how, and in what order to dispatch generation from each power plant in response to customer demand for electricity in each moment and the variable cost of production at each plant.” AVERT analyzes how hourly changes in demand change the output of fossil generators and, with that, their hourly generation, heat input, and emissions of PM_2.5, SO₂, NO_x, and CO₂.” The carbon intensity returned from marginal models such as AVERT can differ significantly from average eGRID data. Table 10-3 shows that, for selected states, this difference can be greater than 60 percent (Mueller and Unnasch, 2021).

___________________

⁶ See https://www.epa.gov/egrid.

⁷ See https://www.epa.gov/avert.

• The Center for Climate and Energy Decision Making at Carnegie Mellon University maintains a database of marginal emission factors that include GHG emissions and air pollutants:⁸ These estimates are summarized in Table 10-4: they include regression-based marginal emission factors, simulation-based marginal emission factors, and average emissions from fossil generators (a potential proxy for marginal emissions).

___________________

⁸ See https://cedm.shinyapps.io/MarginalFactors/.

EFFECTS OF PUBLIC POLICY ON CONSEQUENTIAL PLUG-IN ELECTRIC VEHICLE EMISSIONS

The consequential GHG emissions implications of vehicle electrification can be substantially affected by policy. In the United States, emission rates of new light-duty vehicles are capped by corporate average fuel economy (CAFE) standards (regulated by the National Highway Traffic Safety Administration) and light-duty vehicle GHG emissions standards (regulated by EPA).⁹ Both agencies treat some alternative fuels, including electricity, favorably in compliance calculations; effectively counting some alternative fuel vehicles as lower emitting than they actually are and therefore permitting higher emissions from other vehicles to meet the same standard, resulting in increased overall permitted fleet emissions when alternative fuel vehicles are sold (Gan et al., 2021; Jenn et al., 2016, 2019). For example, in 2021 each BEV sold counts in compliance calculations as though 1.5 BEVs were sold, and BEV charging emissions are multiplied by zero in compliance calculations (Jenn et al., 2016) For this reason, the consequential implications of policies that increase BEV market share, such as some LCFS policies or California’s “zero emission vehicle” policy, is to increase permitted fleet emissions, at least as long as these incentives are in place in national fleet standards (Choi et al., 2013).

Every PEV sold in the United States increases permitted fleet emissions in federal standards, and because the auto industry is constrained by fleet GHG standards, they tend to sell fleets that emit as much as permitted by the standards. Studies find that each time an EV is sold, permitted fleet GHG emissions increase by up to 60 tons, depending on the year and vehicle type (Jenn et al., 2016), and policies like California’s zero emission vehicle mandate result in increased emissions due to these fleet standards (Jenn et al., 2019). It is possible that such policies may also induce innovation, trigger increased adoption, or enable stricter future standards that may reduce long-run emissions, though these factors are more difficult to quantify.

The emissions implications of policies like an LCFS that encourage fuel and vehicle technology switching can be substantially affected by interactions with other policies, including fleet vehicle emission standards.

REFERENCES

___________________

⁹ See https://www.govinfo.gov/content/pkg/FR-2021-12-30/pdf/2021-27854.pdf.

Choi, D.-G., F. Kreikebaum, V. M. Thomas, and D. Divan. 2013. Coordinated EV adoption: Double digit reductions in emissions and fuel use for $40/vehicle-year. Environmental Science & Technology 47(18):10703−10707. http://dx.doi.org/10.1021/es4016926.

Desai, R. R., R. B. Chen, E. Hittinger, and E. Williams. 2019. Heterogeneity in economic and carbon benefits of electric technology vehicles in the US. Environmental Science & Technology 54(2): 1136−1146.

EIA (Energy Information Administration). 2012. Electric generator dispatch depends on system demand and the relative cost of operation. https://www.eia.gov/todayinenergy/detail.php?id=7590.

Elgowainy, A., J. Han, L. Poch, M. Wang, A. Vyas, M. Mahalik, and A. Rousseau, A. 2010. Well-to-Wheels Analysis of Energy Use and Greenhouse Gas Emissions of Plug-in Hybrid Electric Vehicles. ANL/ESD/10-1. Argonne National Lab. (ANL), Argonne, IL (United States). https://doi.org/10.2172/982352.

Elgowainy, A., J. Han, J. Ward, F. Joseck, D. Gohlke, A. Lindauer, T. Ramsden, M. Biddy, M. Alexander, M., S. Barnhart, and I. Sutherland. 2018. Current and future United States light-duty vehicle pathways: Cradle-to-grave lifecycle greenhouse gas emissions and economic assessment. Environmental Science & Technology 52(4):2392−2399.

EPA (US Environmental Protection Agency). 2020. AVoided Emissions and geneRation Tool AVERT). User Manual Version 3.0 https://www.epa.gov/sites/default/files/2020-09/documents/avert_user_manual_09-12-20_508.pdf. (accessed February 4, 2022).

EPA. 40 CFR Parts 86 and 600. Federal Register / Vol. 86, No. 248 / Thursday, December 30, 2021. https://www.govinfo.gov/content/pkg/FR-2021-12-30/pdf/2021-27854.pdf.

EPA. n.d. Emissions & Generation Resource Integrated Database (eGRID). https://www.epa.gov/egrid (accessed February 4, 2022).

Gan, Y., M. Wang, Z. Lu, and J. Kelly, J. 2021. Taking into account greenhouse gas emissions of electric vehicles for transportation de-carbonization. Energy Policy 155(2021):112353. https://doi.org/10.1016/j.enpol.2021.112353.

Graff, J. S. G. Z., M. J. Kotchen, and E. T. Mansur. 2014. Spatial and temporal heterogeneity of marginal emissions: Implications for electric cars and other electricity-shifting policies. Journal of Economic Behavior & Organization 107:248–268.

Greer, M., ed. 2012. Chapter 3 – U.S. Electric Markets, Structure, and Regulations. Pages 39−100 in Electricity Marginal Cost Pricing. Butterworth-Heinemann, ISBN 9780123851345. https://doi.org/10.1016/B978-0-12-385134-5.00003-X.

Hadley, S. W., and A. A. Tsvetkova. 2009. Potential impacts of plug-in hybrid electric vehicles on regional power generation. The Electricity Journal 22(10):56−68.

Hoehne, C. G., and M. V. Chester. 2016. Optimizing plug-in electric vehicle and vehicle-to-grid charge scheduling to minimize carbon emissions. Energy 115:646−657.

Holland, S. P., E. T. Mansur, N. Z. Muller, and A. J. Yates. 2016. Are there environmental benefits from driving electric vehicles? The importance of local factors. American Economic Review 106(12):3700−3729.

Holland, S. P., E. T. Mansur, N. Z. Muller, and A. J. Yates. 2019a. Distributional effects of air pollution from electric vehicle adoption. Journal of the Association of Environmental and Resource Economists 6(S1):S65−S94.

Holland, R.A., K. Scott, P. Agnolucci, C. Rapti, F. Eigenbrod, and G. Taylor. 2019b. The influence of the global electric power system on terrestrial biodiversity. Proceedings of the National Academy of Sciences 116(51):26078−26084.

Jenn, A., I. L. Azevedo, and J. J. Michalek. 2016. Alternative fuel vehicle adoption increases fleet gasoline consumption and greenhouse gas emissions under United States corporate average fuel economy policy and greenhouse gas emissions standards. Environmental Science & Technology 50(5): 2165−2174.

Jenn, A., I. L. Azevedo, and J. J. Michalek. 2019. Alternative-fuel-vehicle policy interactions increase U.S. greenhouse gas emissions. Transportation Research Part A: Policy and Practice 124:397−407.

Jenn, A., J. H. Lee, S. Hardman, and G. Tal. 2020. An in-depth examination of electric vehicle incentives: Consumer heterogeneity and changing response over time. Transportation Research Part A: Policy and Practice 132:97−109.

Karabasoglu, O., and J. J. Michalek. 2013. Influence of driving patterns on lifetime cost and life cycle emissions of hybrid and plug-in electric vehicle powertrains. Energy Policy 60:445−461.

Kawamoto, R., H. Mochizuki, Y. Moriguchi, T. Nakano, M. Motohashi, Y. Sakai, and A. Inaba. 2019. Estimation of CO₂ emissions of internal combustion engine vehicle and battery electric vehicle using LCA. Sustainability 11(9):2690.

Lee, D.-Y., and V.M. Thomas. 2017. Parametric modeling approach for economic and environmental life cycle assessment of medium-duty trucks. Journal of Cleaner Production 142(4):3300–3321. http://dx.doi.org/10.1016/j.jclepro.2016.10.139.

MacPherson, N. D., G. A. Keoleian, and J. C. Kelly. 2012. Fuel economy and greenhouse gas emissions labeling for plug‐in hybrid vehicles from a life cycle perspective. Journal of Industrial Ecology 16(5):761−773.

Michalek, J. J., M. Chester, P. Jaramillo, C. Samaras, C. N. Shiau, and L. B. Lave. 2011. “Valuation of plug-in vehicle life-cycle air emissions and oil displacement benefits.” PNAS 108(40):16554-16558. https://www.pnas.org/content/108/40/16554.

Michalek, J. J., M. Chester., P. Jaramillo, C. Samaras, C.-S.N. Shiau, and L. B. Lave. 2011. Valuation of plug-in vehicle life-cycle air emissions and oil displacement benefits. PNAS 108(40):16554−16558. https://doi.org/10.1073/pnas.1104473108.

Miotti, M., G. J. Supran, E. J. Kim, and J. E. Trancik. 2016. Personal vehicles evaluated against climate change mitigation targets. Environmental Science & Technology 50(20):10795−10804. https://doi.org/10.1021/acs.est.6b00177.

MISO (Midcontinent Independent System Operator). n.d. https://www.misoenergy.org/markets-and-operations/real-time-market-data/real-time-displays/.

Mueller, S., and S. Unnasch. 2021. High octane low carbon fuels: the bridge to improve both gasoline and electric vehicles. University of Illinois at Chicago. https://erc.uic.edu/wp-content/uploads/sites/633/2021/03/UIC-Marginal-EV-HOF-Analysis-DRAFT-3_22_2021_UPDATE.pdf (accessed February 4, 2022).

Nealer, R., D. Reichmuth, and D. Anair. 2015. Cleaner Cars from Cradle to Grave: How Electric Cars Beat Gasoline Cars on Lifetime Global Warming Emissions. Union of Concerned Scientists.

Nopmongcol, U., J. Grant, E. Knipping, M. Alexander, R. Schurhoff, D. Young, J. Jung, T. Shah, and G. Yarwood. 2017. Air quality impacts of electrifying vehicles and equipment across the United States. Environmental Science & Technology 51(5):2830−2837.

Onat, N. C., M. Kucukvar, and O. Tatari, O. 2015. Conventional, hybrid, plug-in hybrid or electric vehicles? State-based comparative carbon and energy footprint analysis in the United States. Applied Energy 150:36−49.

Ryan, N. A., J. X. Johnson, and G. A. Keoleian. 2016. Comparative assessment of models and methods to calculate grid electricity emissions. Environmental Science & Technology 50:8937−8953. https://doi.org/10.1021/acs.est.5b05216.

Sheppard, C. J., A. T. Jenn, J. B. Greenblatt, G. S. Bauer, and B. F. Gerke. 2021. Private versus shared, automated electric vehicles for US personal mobility: Energy use, greenhouse gas emissions, grid integration, and cost impacts. Environmental Science & Technology 55(5):3229−3239.

Siler-Evans, K., I. L. Azevedo, and M. G. Morgan. 2012. Marginal emissions factors for the U.S. electricity system. Environmental Science & Technology 46(9):4742−4748.

Tamayao, M. A. M., J. J. Michalek, C. Hendrickson, and I. M. Azevedo. 2015. Regional variability and uncertainty of electric vehicle life cycle CO₂ emissions across the United States. Environmental Science & Technology 49(14):8844−8855.

Tamayao, M-A. M., J. J. Michalek, C. Hendrickson, and I. M. L. Azevedo. 2015. Regional Variability and Uncertainty of Electric Vehicle Life Cycle CO₂ Emissions across the United States. Environmental Science & Technology 49(14):8844−8855. https://doi.org/10.1021/acs.est.5b00815.

Tessum, C. W., J. D. Hill, and J. D. Marshall. 2014. Life cycle air quality impacts of conventional and alternative light-duty transportation in the United States. Proceedings of the National Academy of Sciences 111(52):18490−18495.

Thomas, C. S. 2012. How green are electric vehicles? International Journal of Hydrogen Energy 37(7):6053−6062.

Tong, X., S. Ovtar, K. Brodersen, P. V. Hendriksen, and M. Chen. 2020. Large-area solid oxide cells with La0. 6Sr0. 4CoO3-δ infiltrated oxygen electrodes for electricity generation and hydrogen production. Journal of Power Sources 451:227742.

Weber, C. L., P. Jaramillo, J. Marriott, and C. Samaras. 2010. Life cycle assessment and grid electricity: what do we know and what can we know? Environmental Science & Technology 44(6):1895−1901. https://doi.org/10.1021/es9017909.

Weis, A., P. Jaramillo, and J. Michalek. 2014. Estimating the potential of controlled plug-in hybrid electric vehicle charging to reduce operational and capacity expansion costs for electric power systems with high wind penetration. Applied Energy 115(190−204):0306−2619. https://doi.org/10.1016/j.apenergy.2013.10.017.

Weis, A., P. Jaramillo, and J. Michalek. 2016. Consequential life cycle air emissions externalities for plug-in electric vehicles in the PJM interconnection. Environmental Research Letters 11(2). https://iopscience.iop.org/article/10.1088/1748-9326/11/2/024009.

Weis, A., J. J. Michalek, P. Jaramillo, and R. Lueken. 2015. Emissions and cost implications of controlled electric vehicle charging in the U.S. PJM interconnection. Environmental Science & Technology 49(9):5813−5819. https://doi.org/10.1021/es505822f.

Yawitz, D., A. Kenward, and E. D. Larson. 2013. A Roadmap to Climate-Friendly Cars. http://assets.climatecentral.org/pdfs/ClimateFriendlyCarsReport_Final.pdf.

Yuksel, T., and J. J. Michalek. 2015. Effects of regional temperature on electric vehicle efficiency, range, and emissions in the United States. Environmental Science & Technology 49(6):3974−3980.

Yuksel, T., M. Tamayao, C. Hendrickson, I. Azevedo, and J. J. Michalek. 2016. Effect of regional grid mix, driving patterns and climate on the comparative carbon footprint of electric and gasoline vehicles. Environmental Research Letters 11(4):044007.