Zero Emission Vehicles: Forecasting Fleet Scenarios and their Emissions Implications (2019)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

7 CHAPTER 2 BRIEF LITERATURE REVIEW TO SUPPORT MODELING SCENARIO DEVELOPMENT 2.1 METHODOLOGY FOR LITERATURE REVIEW The objective of the literature review was to quickly gain insight into key factors affecting ZEV adoption, and how those could be represented by MA3T in our analysis. The objective of this literature review is not to complete a comprehensive review of all literature available; based on collaborative input from the panel, the research team identified approximately 30 resources that were well-suited to developing variables required by the MA3T model. The review emphasized material produced by organizations that are active in ATV research or deployment, including the U.S. Energy Information Administration (EIA), the U.S. Department of Energy (DOE) national laboratories (e.g., the U.S. National Renewable Energy Laboratory [NREL], ORNL, energy producers (e.g., British Petroleum), automobile manufacturers (e.g., the Alliance of Automobile Manufacturers), the National Academy of Sciences, and non-governmental organizations (e.g., the Union of Concerned Scientists). Resources were evaluated based on whether they provided insight into the key factors leading towards increasing the market share of ZEVs. These factors were considered during the literature review: • EVSE initiatives. • Historical data on ZEV market share and projected adoption rates and sales trends. • Policies and programs understood to affect ZEV adoption rates (e.g., purchase rebates and high occupancy vehicle [HOV] lane access). • Vehicle technology characteristics (e.g., battery life, change in cost of technology). The approach for the literature review was to identify the important factors to inform the use of MA3T. The MA3T model structure provides a useful framework for linking policy shifts to ATV adoption, and helped structure the literature review. A list of the inputs used in the MA3T model are included in Table 1.

8 Table 1. Inputs in MA3T Model. Technology • Vehicle manufacturer cost • Fuel economy in charge depleting-mode • Electricity consumption • Range • Acceleration • Cargo/luggage space • Towing capability • Gas/fuel storage capacity • Passenger capacity • Year on market • Vehicle range utilization • Vehicle survival rate • Component share of drivetrain cost • Supply constraint parameters (i.e., a cap on the number of possible ATV sales because of a supply limitation) • Overseas sales • Overseas technology spillover (i.e., additional U.S. sales that occur as a result of overseas sales) • Technology experience (i.e., the relationship between increased investment and the cumulative installed capacity of the technology) • Fuel economy adjustment factor • National stock • Resale value at end of planned vehicle lifetime • Annual maintenance cost • Make and model availability parameters (i.e., as make and model availability increases, consumer utility increases) • Vehicle price markup • Vehicle fuel/electricity consumption rate Policy • American Recovery and Reinvestment Act (ARRA) • "Instant rebate" • Tax credit • Free parking • HOV access Consumer • Geographic regions (i.e., state) • Area type (i.e., Metropolitan Statistical Area or central city), suburb inside Metropolitan Statistical Area (or suburban) and outside Metropolitan Statistical Area (or rural) • Consumer attitude towards innovation • Driver type • Electric charging availability (home and workplace) Infrastructure • Refueling availability (by fuel type) • Public electric charging availability • Gasoline, diesel, electricity, hydrogen, and natural gas prices Focus on the MA3T model included two main areas: 1. Forecasting tools that could be used to estimate electric vehicle adoption in 2040. 2. Parameters needed to develop adoption scenarios. The necessary parameters for developing adoption scenarios included historical data on ZEV market share, policies influencing ZEV adoption, technology advances that lower capital/operating vehicle costs, behavioral incentives, fuel prices, and consumer financial incentives.

9 Various resources that included peer-reviewed journal articles and government and non-governmental reports and online databases were reviewed (see Appendix A). Table 2 summarizes the number of resources reviewed by topic area. Many of the resources cover multiple topic areas, especially the areas that represent the parameters used in MA3T (i.e., consumer preferences, policies, technology, and infrastructure). Because ZEV technology and adoption has changed rapidly during the last several years, the review focused on recent resources – a majority of which were published between 2016 and 2018. Those resources are listed in Table 3. The key findings from this literature review are discussed in the next section. Table 2. Number of Resources Reviewed by Topic Area. Forecast Resources Historical Data Consumer Preferences Policies Technology Infrastructure Other 15 16 23 20 22 25 16 Table 3. Literature and Data Sources Reviewed. Title Author Publication Year Advanced Technology Vehicle Sales Dashboard Alliance of Automobile Manufacturers 2019 Alternative Fuels Data Center U.S. DOE 2019 Annual Energy Outlook 2019 U.S. EIA 2019 National ZEV Investment Plan: Cycle 2 Electrify America 2019 Annual Energy Outlook 2018 U.S. EIA 2018 BP Energy Outlook – 2018 Edition British Petroleum 2018 California ZEV Investment Plan: Cycle 2 Electrify America 2018 Demonstrating Plug-in Electric Vehicles Smart Charging and Storage Supporting the Grid R. Gadh 2018 Global EV Outlook 2018 International Energy Agency 2018 PEV Policy Evaluation Rubric: A Methodology for Evaluating the Impact of State and Local Polices on Plug-in Electric Vehicle Adoption G. Morrison, N. Veilleux and C. Powers 2018 Sizing Up a Potential Fuel Economy Standards Freeze K. Larson, T. Houser, and S. Mohan 2018

10 Title Author Publication Year The Continued Transition to Electric Vehicles in U.S. Cities P. Slowik and N. Lutsey 2018 The Road Ahead for Zero-Emission Vehicles in California Beacon Economics 2018 Towards Road Freight Decarbonisation - Trends, Measures and Policies International Transport Forum 2018 Zero-Emission Vehicles – Fact Sheet California Public Utilities Commission (CPUC) 2018 Analysis of the Effect of Zero-Emission Vehicle Policies: State-Level Incentives and the California Zero-Emission Vehicle Regulations U.S. EIA 2017 California ZEV Investment Plan: Cycle 1 Electrify America and Volkswagen Group of America 2017 Comparison of Vehicle Choice Models T.S. Stephens, R.S. Levinson, A. Brooker, C. Liu, Z. Lin, A. Birky, and E. Kontou 2017 Consumer preferences for electric vehicles: a literature review L. Fanchao, E. Molin, and B. van Wee 2017 How policy can build the plug-in electric vehicle market: Insights from the Respondent-based Preference and Constraints (REPAC) model M. Wolinetz and J. Axsen 2017 Literature Review of Electric Vehicle Consumer Awareness and Outreach Activities L. Jin and P. Slowik 2017 National Plug-In Electric Vehicle Infrastructure Analysis U.S. NREL 2017 National ZEV Investment Plan: Cycle 1 Volkswagen Group of America 2017 No free ride to zero-emissions: Simulating a region's need to implement its own zero- emissions vehicle (ZEV) mandate to achieve 2050 GHG targets M. Sykes and J. Axsen 2017 REV UP Electric Vehicles: Multi-State Study of the Electric Vehicle Shopping Experience Sierra Club (M. Lunetta and G. Coplon- Newfield) 2017 State Efforts to Promote Hybrid and Electric Vehicles K. Hartman and E. Dowd 2017

11 Title Author Publication Year Supplement to the California ZEV Investment Plan: Cycle 1 Electrify America and Volkswagen Group of America. 2017 Transportation Energy Demand Forecast 2018 – 2030 California Energy Commission 2017 Electric Vehicle Survey Methodology and Assumptions: Driving Habits, Vehicle Needs, and Attitude towards Electric Vehicles in the Northeast and California Consumers Union and the Union of Concerned Scientists 2016 Electric vehicles revisited: a review of factors that affect adoption M. Coffman, P. Bernstein, and S. Wee 2016 Electrifying the Vehicle Market: Evaluating Automaker Leaders and Laggards in the United States D. Reichmuth and D. Anair 2016 Fuel Cell Technologies Market Report 2016 U.S. DOE, Office of Energy Efficiency and Renewable Energy 2016 Learning from Norwegian Battery Electric and Plug-in Hybrid Vehicle users E. Figenbaum and M. Kolbenstvedt 2016 Advances in consumer electric vehicle adoption research: A review and research agenda Z. Rezvani, J. Jansson, and J. Bodin 2015 Feasibility and Implications of Electric Vehicle (EV) Deployment and Infrastructure Development D. Fordham, J. Norris, and J. Proudfoot 2015 Overcoming Barriers to Deployment of Plug-in Electric Vehicles The National Academies Press 2015 Accounting for Electric Vehicles in Air Quality Conformity R. Farzaneh, Y. Chen, J. Johnson, J. Zietsman, C. Gu, T. Ramani, L.D. White, M.K., and Y. Zhang 2014

12 Title Author Publication Year Exploring the Impact of High Occupancy Vehicle (HOV) Lane Access on Plug-in Vehicle Sales and Usage in California G. Tal and M.A. Nicholas 2014 How Much Do Electric Vehicles Matter to Future U.S. Emissions? S. Babaee, A.S. Nagpure, and J.F. DeCarolis 2014 The influence of financial incentives and other socio-economic factors on electric vehicle adoption W. Sierzchula, S. Bakker, K. Maat, and B. van Wee 2014 Review of Hybrid, Plug-In Hybrid, And Electric Vehicle Market Modeling Studies B.M. Al-Alawi and T.H. Bradley 2013 Final Rule for Model Year 2017 and Later Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Average Fuel Economy Standards (and accompanying Fact Sheet) U.S. EPA and U.S. DOT 2012 Promoting the Market for Plug-In Hybrid and Battery Electric Vehicles: Role of Recharge Availability Z. Lin and D.L. Greene 2011 A Plug-in Hybrid Consumer Choice Model with Detailed Market Segmentation Z. Lin and D.L. Greene 2009 2.2 LITERATURE REVIEW FINDINGS RELEVANT TO MODELING SCENARIO DEVELOPMENT Based on the literature listed in Table 3, the following key findings on programs related to ATV adoption were identified. The following review focuses on model inputs that would be necessary to prepare the MA3T model for simulating ZEV adoption in the analysis scenarios. 2.2.1 Historical ATV Sales The Alliance of Automobile Manufacturers has recorded ATV sales since 2011 in their Advanced Technology Vehicle Sales Dashboard (ATV Sales Dashboard), focusing on sales of HEVs, FCEVs, BEVs, and PHEVs. Figure 1 shows the annual sales of the four types of ATVs accounted for in the ATV Sales Dashboard and the total ATV sales (solid line). HEVs have had the greatest number of sales since 2011, but sales for HEVs have been declining for several years while sales of PHEVs, FCEVs, and BEVs have been rising. Although FCEV sales increased dramatically between 2015 and 2016 (by a factor of 14) and grew by 80% in 2017, their sales accounted for only 0.3% of total ATV sales in 2018. Meanwhile, sales of BEVs have steadily grown since 2015 and doubled between 2017 and 2018, reaching 31% of total ATV sales in 2018. PHEVs have had a slightly lower growth rate compared to BEVs, and account for 19% of total ATV sales in 2018. ATV sales slumped substantially in 2015. This is likely due to reductions in manufacture incentives, declining gasoline prices, and potential delays in consumer

13 purchases as a result of future electric vehicle releases (Shepardson 2017; UCLA 2017; Voelcker 2016). These factors are captured within the MA3T forecast of the future vehicle fleet. California leads the 50 states and Washington D.C. in terms of total ATV sales, with 1,269,877 ATV sales (60% HEV, 22% BEV, 18% PHEV, and 0.4% FCEV) between 2011 and 2018. That value is five times higher than ATV sales in Florida (231,360), which accounts for the second most ATV sales for a state over the time period. Six other states have exceeded 100,000 ATV sales between 2011 and 2018, including (in order of decreasing sales) Texas, New York, Illinois, Washington, Pennsylvania, and Virginia (Alliance of Automobile Manufacturers 2019). Figure 1. Annual ATV sales by ATV category. Data Source: Alliance of Automobile Manufacturers 2019. 2.2.2 ATV Market Share According to the ATV Sales Dashboard, in 2018, ATVs represented approximately 3.9% of annual market share of all registered cars in the country. Combined, BEVs and FCEVs represented 1.2% of annual market share in 2018. Nine states signed a memorandum of understanding (MOU) with California to implement a ZEV mandate, requiring manufacturers to earn a specified number of credits for selling electric vehicles (BEVs, FCEVs, and PHEVs), depending on the total number of vehicles sold. In this report, “MOU states” refers to California and the nine other states (Connecticut, Maine, Maryland, Massachusetts, New Jersey, New York, Oregon, Rhode Island, and Vermont). In the MOU states in 2018, 0 100,000 200,000 300,000 400,000 500,000 600,000 700,000 2011 2012 2013 2014 2015 2016 2017 2018 An nu al A TV S al es in th e U. S. HEV PHEV BEV FCEV Total

14 ATVs represented 7.0% of the annual market share, while BEVs and FCEVs combined represented 2.5% of the market share. ATV market share is greatest in California, where they represented approximately 9.8% of all vehicles registered in the state between 2013 and 2018, and approximately 12% in 2018 alone. All of the MOU states have an ATV market share of about 3% or greater between 2013 and 2018. The standard federal reference used to forecast the future ATV market share is the Annual Energy Outlook (AEO) prepared by the EIA. The EIA produces the AEO in accordance with the DOE Organization Act of 1977, which requires the EIA to prepare annual reports on trends and projections for energy use and supply. The AEO prepares future fleet forecasts for the AEO’s primary forecast for future energy demand (the “reference” case), which assumes that trend improvement in known technologies continue, along with best estimates of economic and demographic trends. The AEO’s forecasts assume that laws and regulations in effect at the time of reporting remain unchanged. Assumptions applied in the AEO forecast can be found in U.S. EIA 2019b and 2019c. According to the AEO estimate, approximately 41.0 million HEVs, PHEVs, BEVs, and FCEVs vehicles will be in use by 2040 (U.S. EIA 2019a). The AEO forecasts that, by 2040, the majority of ATVs are expected to be BEVs (56%), followed by HEVs (29%), and PHEVs (13%). BEVs, HEVs, and PHEVs are forecast to comprise 7.4%, 4.8%, and 2.0%, respectively, of the total light-duty vehicle population. Only 2% of ATVs are expected to be FCEVs, and less than half a percent of all light-duty vehicles (Figure 2) (U.S. EIA 2019a). The AEO represents the standard federal government reference in forecasting both energy demand and ATV vehicle population; the MA3T is also calibrated to the AEO. Two other private organizations prepare comprehensive national forecasts of light-duty ATV population and sales: Navigant Consulting and Bloomberg New Energy Finance. The Bloomberg forecast for ATV population, which covers a period that includes the analysis year used in this research effort (2040), is approximately 60 million BEV, PHEV, or FCEV vehicles by 2040 (Bloomberg New Energy Finance 2018). For contrast, the AEO forecasts 27 million BEV, PHEV, and FCEV by 2040—less than half the Bloomberg forecast (U.S. EIA 2019a). The Navigant forecast also suggests larger growth in the ATV population than is true of the AEO, with a forecast of approximately 1.4 million ATV sales in 2026 (Electrify America 2019). Electrify America relies on Navigant Consulting’s forecasts when determining the best use of a planned $2 billion nationwide investment.

15 Figure 2. Light-Duty ATV Stock by Vehicle Type. Data Source: EIA 2019 Annual Energy Outlook (U.S. EIA 2019a). 2.3 FACTORS AFFECTING ATV ADOPTION 2.3.1 Consumer Preferences Consumer uptake of ATVs is influenced by a number of factors. Purchasing a vehicle represents one of the top expenditures made by individuals or households. The purchase is expected to last many years, and in the face of uncertainty, consumers are more likely to “gravitate to the known and familiar” (National Academies Press 2015). Regardless of the type of vehicle purchased, the top five influences of any car purchase decision include reliability, durability, quality of workmanship, value for the money, and manufacturer’s reputation (National Academies Press, 2015). Assuming these influences are satisfied, consumers will consider ease of charging and battery performance when evaluating whether to purchase a plug-in electric vehicle (PEV) (Fordham et al., 2015). However, based on a nationwide telephone survey of adult vehicle owners, only 42% of drivers (45 million households) are strong candidates for PHEV ownership or lease (National Academies Press 2015). Factors affecting adoption of ATVs have been widely studied (e.g., see Rezvani et al. 2015). Based on the literature review, key drivers encouraging ATV adoption, as well as key barriers inhibiting ATV adoption (particularly BEVs), are identified in Table 4. These drivers and barriers are features of ATVs that make them more or less appealing to consumers and that can be improved upon to increase their attractiveness to consumers. 0 5 10 15 20 25 30 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2020 2025 2030 2035 2040 N um be r o f V eh ic le s ( M ill io ns ) Pe rc en ta ge o f T ot al A TV s HEV PHEV BEV FCEV All ATVs

16 Table 4. Drivers and Barriers to ATV Adoption. Primary Driver of ATV Adoption Primary Barrier to ATV Adoption • Greater vehicle model choice and availability (outside of MOU states, primarily CA) • High vehicle performance and reliability • HOV lane access (such as California metropolitan areas) • Long driving range • Lower fuel cost (combined with lower purchase cost) • Purchase cost parity with ICEVs due to federal and state incentives • Higher purchase cost • Lack of home charging availability • Lack of knowledge about BEVs • Limited vehicle range • Slow charging time Secondary Driver of ATV Adoption Secondary Barrier to ATV Adoption • Environmental benefits • Free parking • Averse attitude towards new technology • Battery concerns (reliability, safety) • Limited charging infrastructure availability (if home charging is available) Each driver and barrier is discussed in more detail below. Drivers • Greater vehicle choice and availability is a driver of ATV adoption. A study by the Sierra Club found that “the average number of EVs on lots in California was nearly twice the average number on lots in the other nine MOU states” (Sierra Club 2017). The availability of ZEVs in California may be an important factor in explaining why that state has the highest market share of ZEVs of any state. Vehicle choice for all areas is likely to improve in the future, as many automakers have set aggressive goals to increase the number of EVs in their fleet (including BMW, GM, Volvo, and Volkswagen) (California Energy Commission 2017). The availability of more EV models on the market has been found to increase the probability of consumers choosing an electric vehicle (EV) (Liao et al. 2017). • Vehicle performance and reliability are foundational to an individual’s decision to consider an EV. A Norwegian study found that drivers of all vehicle types (ICEV, PHEV, and BEV) indicated that reliability and “best for my need” were among their top five considerations when purchasing the vehicle. However, anecdotal evidence indicates that, at least for some consumers who might be characterized as “early adopters” of ATVs, the metrics influencing traditional car purchase decisions do not fully apply, or at least the relative importance of each metric may be different. For example, despite widely reported reliability issues with some Tesla vehicles,

17 demand was expected to remain high for Tesla vehicles as of late 2018 (Consumer Reports 2019, Linnane 2018). • Free or reduced-fee HOV lane access is expected to be a significant factor influencing ATV uptake in metropolitan areas with heavy traffic and high tolling costs. In 2013, HOV lane access in California was the primary motivation for the purchase of 34% of plug-in Toyota Priuses, 20% of Chevy Volts, and 38% of Nissan Leafs (U.S. EIA 2017). BEV buyers in Norway were similarly motivated by the free toll-road access. Free or reduced parking benefits are less of a driver. However, in markets without a large metropolitan area, HOV lane access is not valued highly (U.S. EIA 2017). • Extending the range of ZEVs drives adoption rates. Reducing so-called “range anxiety” with longer-range ZEVs is a driver of ZEV adoption (Electrify America 2018). In addition to vehicle price, long range (e.g., 200-mile electric range) has been identified as a key attribute that consumers consider when deciding whether to purchase an EV (Consumers Union and Union of Concerned Scientists 2016). • Lower ATV fuel costs (e.g., electricity) promote ATV adoption to a point. The expectation of lower fuel costs is a key driver of ATV adoption. Babaee et al. (2014) similarly find that gasoline price and battery cost are the most significant factors to influence ATV deployment. The on- board battery of PHEVs and BEVs allows vehicle owners to either offset or replace gasoline purchases with cheaper home electricity. However, relatively low U.S. gasoline fuel costs in have reduced the significance of this factor as of 2019. PHEV and BEV sales have been found to be correlated with gasoline prices (National Academies Press, 2015). • State and federal incentives are critical to promoting ATV adoption. Without incentives, ATVs are generally more expensive compared to ICEVs. While battery technology has improved and costs have declined, the initial vehicle purchase cost remains one of the top barriers to ATV adoption, and incentives reduce the effective cost of ATVs (e.g., Slowik and Lutsey 2018, Wolinetz and Axsen. 2017). • Environmental benefits are important to early adopters, but are not primary drivers for most consumers. Communities with higher percentages of residents that value the environment see higher EV adoption and investment in EV infrastructure, which drives adoption rates (Fordham et al., 2015). A pro-environment attitude influencing EV preference has been identified by researchers, as listed in Liao et al. (2017) and cited by Rezvani et al. (2015). Barriers • For consumers without access to home charging, limited availability and speed of public charging is a barrier to ATV adoption. However, the availability of public charging infrastructure is generally not a significant concern for drivers with access to home charging, particularly consumers in urban areas. Most PHEV/BEV owners have access to home charging, and primarily charge their vehicles at home. For these EV owners, public charging is only necessary for infrequent longer trips. While the availability of public charging infrastructure— particularly fast chargers—reduces range anxiety, it is generally not a primary barrier to EV adoption, especially for PHEVs. However, lower PEV adoption outside of rural areas is partially

18 driven by lack of charging infrastructure and additional requirements for long-distance trips (U.S. DOE 2017). • Consumers and auto dealers are not well educated about ATVs. Several studies discuss the importance of increasing consumer knowledge about ATVs in driving adoption (see review article by L. Jin and P. Slowik 2017). Perhaps equally important is the need to better educate auto dealers and sales personnel so they are prepared to educate consumers. However, without policy incentives aimed specifically at auto dealers, they may not promote sales of ATVs. This is in part due to the large percentage of auto dealer profits that come from service and maintenance, as ATVs require less service and maintenance than ICEVs (Automotive News 2017). • BEVs with vehicle ranges less than 200 miles on a single charge present a key barrier. A web-based survey of vehicle drivers in ten states (California and nine northeastern states) found that a driving range of 200 miles was a top attribute (Consumers Union & Union of Concerned Scientists, 2016). • Aversion to new technology has been a significant barrier for potential consumers. Lin and Greene found that consumers with low tolerance for technology risk enter the PHEV market more slowly but gain interest as the number of PHEVs achieve higher market penetration (Lin and Greene 2009). Others have found that technological risk reduces the probability of choosing an EV (Liao et al. 2017). In Overcoming Barriers to Deployment of Plug-in Electric Vehicles (National Academies Press 2015), the estimated timeframe for normal market penetration for new technologies is approximately 10 to 15 years. • Consumers are sensitive not only to the individual product, but also to the supporting infrastructure (e.g., charging infrastructure). Market penetration of new technologies takes many years, and is affected not only by the technology, but the supporting and complementary infrastructure (The National Academies Press, 2015). This includes not only charging infrastructure, but also factors such as amenable zoning ordinances and knowledgeable mechanics. Coffman et al. (2016) cite many sources that found adequate charging infrastructure to be critically important to EV adoption. 2.3.2 State and Federal Policies Influencing BEV/FCEV Adoption ATV adoption is strongly correlated with supportive state and federal policies. Polices can be categorized as demand-focused (provides incentives to consumers) or supply-focused (provides incentives/mandates to automakers/dealers). Both types of incentives are important in influencing ATV adoption. Until ATVs reach price parity with ICEVs, demand-focused policies will continue to be a key factor in ATV uptake (Wolinetz and Axsen, 2017; Sierra Club 2017). The Alternative Fuels Data Center provides a summary of all federal and state laws and incentives for alternative fuels and vehicles, air quality, vehicle efficiency, and other transportation-related topics (U.S. DOE 2019a). Similarly, the National Conference of State Legislatures tracks state incentives used to promote hybrid and electric vehicles in a database (Hartman and Dowd 2017). This database shows that all but four states provide support for ATVs through grants for state agencies, alternative fuel tax exemption, rebates for EVSE, HOV lane access, tax credits for ATV infrastructure, use-tax exemptions, purchase rebates, emission exemptions, tax exclusions, and other incentives (U.S. DOE 2019b).

19 One of the most expansive levers in use as of 2019 is the California Zero-Emission Vehicle Regulations (ZEVR). Under Section 209 of the Clean Air Act, California has obtained a federal waiver to enact vehicle emission standards that are stricter than federal standards. The ZEVR encourage BEV, FCEV, and PHEV sales in the form of monetary penalties for manufacturers that fail to obtain the required level of sales credits (U.S. EIA 2017). The MOU states set a collective target of at least 3.3 million BEVs, FCEVs, and PHEVs on the road by 2025. In 2018, California set a new target of 5 million by 2030.i High-level observations about the impact of state and federal policies are included below: • Research suggests that state and federal policies play a substantial role in encouraging ATV adoption. In California, with the state’s strict ZEVR, ATVs represent approximately 12% of all ATV sales as of 2018. The annual ATV market share for new vehicle purchases across the rest of the country and Washington D.C. was approximately 4% in 2018 (Alliance of Automobile Manufacturers 2019). Sierzchula et al. (2014) found that financial incentives positively and significantly predicted EV adoption rates, and Rezvani et al. (2015) indicate the importance of financial incentives in motivating consumers to purchase EVs. • To effectively accelerate ATV adoption, states may need to consider a broader mix of incentives. While consumer tax rebates or discounts are favored in 2019, survey findings suggest that they may not be sufficient to encourage ATV adoption. Based on survey participants’ lack of familiarity with consumer-oriented incentives, the Sierra Club encourages policy makers to consider grants and incentives for businesses (and particularly dealerships), municipalities, and government agencies in order to encourage more widespread adoption (Sierra Club 2017). • A combination of supply- and demand-focused policies are needed to substantially increase EV market penetration. Wolinetz and Axsen (2017) emphasize that a combination of strong demand and supply-focused policies are necessary to dramatically increase PEV adoption. Demand-focused policies are important to reducing the upfront purchase cost until EVs reach price parity with ICEVs and consumers become more familiar with the technology. However, supply-focused polices are key to increasing the variety and availability of PEVs. Based on their model, strong demand-focused policies will not result in new market penetration beyond 17 to 28% by 2030, while a combination of demand- and supply-focused policies will result in a market penetration between 38 to 49%. • Individual state markets may be slow to react to new rebate programs and other state incentives. In Analysis of the Effect of Zero-Emission Vehicle Policies: State-Level Incentives and the California Zero-Emission Vehicle Regulations, monthly sales of ATVs in certain states show no clear response to the introduction of a rebate program (U.S. EIA 2017). • Dealers may not be well-informed about the incentives available at the state level. A survey conducted by Sierra Club volunteers revealed that about 33% of the time, the salesperson did not discuss federal and state tax credits and rebates available to reduce the cost of a ZEV or PHEV (Sierra Club 2017). • An ATV mandate does not necessarily correspond to availability for ATVs. The survey conducted by Sierra Club volunteers found ZEVs were 2.5 times more likely to be found on dealership lots in California compared with dealership lots in the eight other MOU states. (Sierra Club 2017).

20 • States that offer the largest tax credits to offset the price of ATVs in 2019 do not have the largest market share of ATVs. For example, as of 2019, Colorado offered the highest incentives for ATV purchases of any state, with incentives valued at approximately $7,500 for a BEV, but ATVs have a relatively low market share in that state (about 5% in 2018). Conversely, California has by far the highest ATV market share (12% in 2018) but not the highest valued incentives (valued at approximately $3,000 for a BEV). Sierzchula et al. (2014) found that financial incentives do not guarantee high adoption of EVs. The valuation of state-level ATV incentives was not well correlated with market share of ATVs (U.S. EIA 2017). No information was available on how a mix of incentives would affect ATV adoption in specific states, and this is an uncertainty that could be investigated. 2.3.3 Technology Characteristics Two of the most critical factors for ATV market penetration are the cost considerations and risk associated with new technology. Because purchasing a vehicle is one of the most expensive purchases made by individuals and households, the research suggests that concerns about technology present a substantial barrier for adoption (National Academies Press 2015). Key barriers to adoption are technology-driven, and include: • Battery safety concerns • Consumer lack of knowledge and uncertainty • High cost • Lack of charging infrastructure • Limited driving range • Limited model availability • Long charging time • Uncertain battery life Key factors to consider with regard to the role of technology in ATV adoption include the following: • The cost of the battery is the single largest barrier to ATV adoption. EV batteries make up a large portion of an EV’s total value (e.g., see ICCT 2019, Environmental and Energy Study Institute 2017). In Promoting the Market for Plug-In Hybrid and Battery Electric Vehicles: Role of Recharge Availability (2012), Lin and Greene argue that of all factors considered in the MA3T model V20190404, technological advancement that reduces battery cost would have the largest impact on increasing ATV sales. This is particularly true when access to home recharging is improved. No Free Ride to Zero-Emissions (Sykes and Axsen 2017) observes that three of the most critical factors that lead to a region’s ability to achieve GHG targets through ATV purchases include (1) the degree to which consumers perceive varying lifecycle costs for the same technology (i.e., the likelihood that consumers will select the vehicle with the lowest lifecycle costs); (2) BEV capital cost progress ratio (i.e., the rate at which capital costs decline every time a technology’s cumulative production increases); and (3) exogenous capital cost decline rate (i.e., the rate at which capital costs decline regardless of cumulative production). In The Road Ahead for Zero-Emission Vehicles in California, price was also considered a critical factor (Beacon Economics 2018). In Review of Hybrid, Plug-In Hybrid, and Electric Vehicle Market Modeling Studies, the authors note that ATV adoption was found to be sensitive to technology and ownership cost, particularly when ICEV operation cost was high (Al-Alawi and Bradley 2013).

21 • Battery costs continue to decline at a fast rate. Researchers have found that between 2010 and 2016, battery cost declined by 19.5% per year, and that costs will continue to decline by 9.7% per year between 2016 and 2025, and by 7.7% per year between 2025 and 2030 (Beacon Economics 2018). Coffman et al. (2016) also cite a large drop in battery costs. However, supply chain issues with battery inputs (such as lithium, cobalt, and graphite) could lead to a slowdown in battery price decline (Beacon Economics 2018). 2.3.4 Infrastructure Initiatives EVSE is no longer a rarity in the United States. The Alternative Fuels Data Center compiles a database that lists 33,615 stations in all 50 states, the District of Columbia, American Samoa, and Puerto Rico in 2019. Fuels available include biodiesel, compressed natural gas, ethanol, electricity, hydrogen, liquefied natural gas, and liquefied petroleum gas (U.S. DOE 2019a). The ten states in the database with the largest number of electric stations and chargers are listed in Table 5. Together, the top ten states represent approximately 59% of all electric charging stations in the 50 states and the District of Columbia. The database indicates that only one state has more than five hydrogen fueling stations: California, with 44 hydrogen refueling stations (both private and public). Table 5. Electric Stations and Public Electric Charger Counts by State (2019). State Public Stations Public Chargers Private Stations Private Chargers California 5,057 19,555 668 2,739 Florida 1,138 2,921 175 398 New York 1,167 2,686 100 181 Texas 1,114 3,045 124 279 Washington 873 2,356 102 247 Georgia 765 2,316 69 104 Colorado 676 1,814 58 145 Oregon 618 1,490 70 125 Virginia 571 1,346 98 245 Maryland 586 1,578 82 192 Total (all 50 states and D.C.) 12,565 39,107 1,546 4,655 Source: U.S. DOE 2019c The Alternative Fuels Data Center also provides an extensive list of federal and state laws and incentives, including nearly 300 incentives/regulations that support truck stop electrification or fueling infrastructure (U.S. DOE 2019b). Several high-profile future or ongoing investments include the following: • As of 2019, the three Investor Owned Utilities in California (Pacific Gas and Electric [PG&E], Southern California Edison [SCE], and San Diego Gas & Electric [SDG&E]) are in the process of implementing pilot programs to support EVSE at multi-unit dwellings, workplaces, and public interest destinations pursuant to California SB 350. The utility pilots will install infrastructure for

22 up to 12,500 stations at a cost of $197 M. This pilot program places particular emphasis on disadvantaged communities, and up to 10% of total funding will be allocated to these communities (California Public Utilities Commission 2018, Beacon Economics 2018). • Electrify America will spend $800 million in California and $1.2 billion outside of California as part of the Partial Consent Decree entered by the U.S. District Court for the Northern District of California in 2016 (note that Electrify America was established by VW in the wake of the VW “dieselgate” scandal and its subsequent settlement agreements). Through Electrify America’s second investment cycle (July 2019 through December 2021), Electrify America plans to deploy approximately 215 electric charging facilities outside of California, and at least 2,610 electric charging facilities within California. This effort will be coordinated with state agencies to supplement, not duplicate, other EVSE initiatives underway, and the stations will be “future- proofed” to the greatest extent possible to ensure that stations can be converted from 150 kW to 320 kW by the end of the 4th cycle (i.e., the end of quarter four of 2026) (Electrify America 2018, 2019). • Assembly Bill 118 in California created the Alternative and Renewable Fuel and Vehicle Technology Program (ARFVTP), which has funded 39% of statewide total public charging sites and 38% of charging outlets (Beacon Economics 2018). • The FHWA has unveiled the National Alternative Fuel and Electric Charging Network which, as of 2019, included 100 interstate corridors across 46 states where electric, hydrogen, propane, and natural gas fueling stations are available (“corridor ready”) or in progress (“corridor pending”) (U.S. FHWA 2019). • The West Coast Electric Highway (alternately, the West Coast Green Highway) is an initiative to establish a network of EV direct current (DC) fast charging stations (DCFC) located every 25 to 50 miles along Interstate 5, Hwy 99, and other major roadways in British Columbia, Washington, Oregon, and California. The initiative is a collection of projects, funding sources, and partners with the same vision—to provide a network of fast charging stations enabling electric vehicle drivers to make longer trips and travel between cities (Washington State Department of Transportation 2014). • Regional Electric Vehicle Plan for the West (REV West) is an initiative by the States of Colorado, Idaho, Montana, Nevada, New Mexico, Utah and Wyoming to make it possible to drive an electric vehicle across the signatory states’ major transportation corridors. The initial transportation corridors include Interstates 25, 70 and 76 in Colorado; Interstates 15, 84, 86, and 90 in Idaho; Interstates 15, 90 and 94 in Montana; Interstates 15 and 80 in Nevada; Interstates 10, 25 and 40 in New Mexico; Interstates 15, 70, 80 and 84 in Utah; and Interstates 25, 80 and 90 in Wyoming (State of Colorado et al. 2017). The relative importance of these infrastructure initiatives varies depending on the type of vehicle and length of the driver’s typical trip. Some key observations include the following: • Public charging may not be valued highly by PHEV owners. Lin and Greene (2009) observe in A Plug-in Hybrid Consumer Choice Model with Detailed Market Segmentation that the value of public charging to PHEV owners is difficult to quantify because recharging is not required for use of the vehicle. However, the authors found that the overall market success of PHEVs appears

23 to be driven by recharging availability, consumers’ attitudes towards novel technologies, and vehicle-usage intensity. • Satisfying the demand for long-distance BEV charging facilities may not require substantial additional investment. National Plug-In Electric Vehicle Infrastructure Analysis (Electrify America 2017) suggests that (1) relatively few direct current fast charging (DCFC) stations would be required to satisfy long-distance BEV travel between U.S. cities, and (2) that number is similar to the number of corridor-located DCFC stations already established by Tesla or planned by Electrify America. • The availability of chargers appears to be a stronger driver for ATV adoption along the west coast, which is also an area with a relatively developed network of charging stations (U.S. EIA, 2017). 2.4 SUMMARY OF LITERATURE REVIEW This literature review summarizes historical ATV data, ATV adoption forecasts, and available information and data on the key factors (consumer preferences, technology, policy, and infrastructure) affecting adoption of ATVs. Examples of the key factors identified in the literature review are summarized in Table 6. The table indicates the relative importance of each factor example, and whether it is an input for the MA3T model. The majority of example factors with high importance are included in the MA3T model, which will be useful for establishing ATV adoption rates under various scenarios. An important finding in this literature review was the significance of policy incentives that can be considered as either supply- or demand-focused. State and federal-level incentives provide strong support for ATVs, with all but four states providing at least some support for ATVs through grants for state agencies, alternative fuel tax exemption, rebates for EVSE, HOV lane access, tax credits for ATV infrastructure, use-tax exemptions, purchase rebates, emission exemptions, tax exclusions, and other incentives. These incentives are demand-focused, and though they may be highly valued, there is mixed evidence that consumers are well-informed of their existence. On the whole, the literature suggests that for ATV adoption to play a large role in emission reduction, cost will play the largest role. Supply-focused polices are key to increasing the variety and availability of ATVs. While ZEV regulations are in place in California and other states that require manufacturers to achieve certain sales quotas on ATVs, grants and incentives for businesses (and particularly dealerships), municipalities, and government agencies may be necessary to encourage more widespread adoption. As consumers become better informed of available incentives and the technology behind ATVs, market share is expected to continue to increase. The key factors presented in this literature review and summarized in Table 6 present the basic inputs for quantifying future ATV adoption. The forecasts for ATV sales and the trends in technology advancements, policies and programs, and EVSE identified in this literature review were used to develop inputs for the MA3T model. The impacts on emissions reductions is covered in detail in Chapter 4.

24 Table 6. Summary of key factors affecting ATV adoption and their relative importance. Driver of ATV Adoption Relative Importance MA3T Parameter MA3T Factor High Vehicle Performance and Reliability Higha No Consumer HOV Lane Access High Yes Policy Long Driving Range High Yes Technology Purchase Cost Parity with ICEVs (due to federal and state incentives) High Yes Policy, Technology Free Parking Medium Yes Policy Lower Fuel (e.g., Electricity) Cost (combined with lower purchase cost) Medium Yes Infrastructure Environmental Benefits Low No NAb Barrier to ATV Adoption Relative Importance MA3T Parameter MA3T Factor Higher Purchase Cost High Yes Technology Lack of Home Charging High Yes Consumer, Infrastructure Lack of Knowledge about BEVs High No Consumer Limited Vehicle Model Choice and Availability High No Technology Limited Vehicle Range High Yes Technology Negative Attitude Towards New Technology High Yes Consumer Battery Concerns (reliability, safety) Medium No Consumer Limited Charging Infrastructure Availability (if home charging is available) Lowc Yes Infrastructure Slow Charging Time Low No Infrastructure a HOV lane access is not a primary driver for markets without a large metropolitan area. b NA = not applicable. c This factor has high relative importance for those without access to home chargers.