國家可再生能源實驗室：2030年國家充電網絡估算美國對電動汽車充電基礎設施的輕稅需求（英文版）（81頁）.pdf

《國家可再生能源實驗室：2030年國家充電網絡估算美國對電動汽車充電基礎設施的輕稅需求（英文版）（81頁）.pdf》由會員分享，可在線閱讀，更多相關《國家可再生能源實驗室：2030年國家充電網絡估算美國對電動汽車充電基礎設施的輕稅需求（英文版）（81頁）.pdf（81頁珍藏版）》請在三個皮匠報告上搜索。

1、The 2030 National Charging Network:Estimating U.S.Light-Duty Demand for Electric Vehicle Charging Infrastructureii This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Acknowledgments The authors would like to acknowledge the Joint Office of

2、Energy and Transportation and the U.S.Department of Energys(DOEs)Vehicle Technologies Office for supporting this analysis.Specific thanks to DOE,U.S.Department of Transportation,and Joint Office staff for their ongoing guidance,including Jacob Ward,Raphael Isaac,Patrick Walsh,Wayne Killen,Rachael Ne

3、aler,Lissa Myers,Suraiya Motsinger,Alan Jenn,Noel Crisostomo,Kara Podkaminer,Alex Schroeder,Gabe Klein,Andrew Rodgers,Andrew Wishnia,and Michael Berube.Internal support at the National Renewable Energy Laboratory was critical to completion of this report,including from Jeff Gonder,Matteo Muratori,An

4、drew Meintz,Arthur Yip,Nick Reinicke,Justin Rickard,Elizabeth Stone,Michael Deneen,John Farrell,Chris Gearhart,and Johney Green.The authors would also like to thank colleagues at the California Energy Commission(Michael Nicholas and Adam Davis)and U.S.Environmental Protection Agency(Susan Burke and

5、Meredith Cleveland)for ongoing collaborations that have been synergistic toward the execution of this analysis,including support for EVI-Pro and EVI-RoadTrip.Timely contributions from Atlas Public Policy were necessary to accurately estimate the magnitude of charging infrastructure announcements fro

6、m the public and private sectors.Thanks to Spencer Burget,Noah Gabriel,and Lucy McKenzie.Special thanks to external reviewers who provided feedback during various phases of this work.While reviewers were critical to improving the quality of this analysis,the views expressed in this report are not ne

7、cessarily a reflection of their(or their organizations)opinions.External reviewers included:Charles Satterfield.Edison Electric Institute Jamie Dunckley Electric Power Research Institute Paul J.Allen Environmental Resources Management Colin Murchie and Alex Beaton.EVgo Jamie Hall,Alexander Keros,Mic

8、hael Potter,and Kelly Jezierski.General Motors Brian Wilkie,Christopher Coy,and Ryan Wheeler National Grid Jen Roberton New York State Department of Public Service Vincent Riscica.New York State Energy Research&Development Authority Erick Karlen.Shell Recharge Solutions Madhur Boloor and Michael Mac

9、hala.Toyota Research Institute Nikita Demidov Trillium Susan Burke.U.S.Environmental Protection Agency,Office of Transportation and Air Quality iii This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Authors The authors of this report are:Er

10、ic Wood,National Renewable Energy Laboratory(NREL)Brennan Borlaug,NREL Matt Moniot,NREL Dong-Yeon(D-Y)Lee,NREL Yanbo Ge,NREL Fan Yang,NREL Zhaocai Liu,NREL iv This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.List of Acronyms BEV battery-e

11、lectric vehicle CBSA core-based statistical area CCS Combined Charging System DC direct current DOE U.S.Department of Energy EV electric vehicle EVI-X electric vehicle infrastructure analysis tools EVSE electric vehicle supply equipment FHWA Federal Highway Administration ICCT International Council

12、on Clean Transportation Joint Office Joint Office of Energy and Transportation L1 Level 1 L2 Level 2 LDV light-duty vehicle NACS North American Charging Specification NHTS National Household Travel Survey PEV plug-in electric vehicle PHEV plug-in hybrid electric vehicle SFH single-family home SOC st

13、ate of charge TAF Traveler Analysis Framework TNC transportation network company VMT vehicle miles traveled ZEV zero-emission vehicle v This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Executive Summary U.S.climate goals for economywide n

14、et-zero greenhouse gas emissions by 2050 will require rapid decarbonization of the light-duty vehicle1 fleet,and plug-in electric vehicles(PEVs)are poised to become the preferred technology for achieving this end(U.S.Department of Energy 2023).The speed of this intended transition to PEVs is evident

15、 in actions taken by government and private industry,both in the United States and globally.New PEV sales have reached 7%10%of the U.S.light-duty market as of early 2023(Argonne National Laboratory 2023).Globally,PEV sales accounted for 14%of the light-duty market in 2022,with China and Europe at 29

16、%and 21%,respectively(IEA 2023).A 2021 executive order(Executive Office of the President 2021)targets 50%of U.S.passenger car and light truck sales as zero-emission vehicles(ZEVs)by 2030,and California has established requirements for 100%light-duty ZEV sales by 2035(California Air Resources Board 2

17、022),with many states adopting or considering similar regulations(Khatib 2022).These goals were set prior to passage of the landmark U.S.Bipartisan Infrastructure Law and Inflation Reduction Act,which provide substantial policy support through tax credits and investment grants(Electrification Coalit

18、ion 2023).Companies in the automotive industry have committed to this transition,with most companies rapidly expanding offerings(Bartlett and Preston 2023)and many pledging to become ZEV-only manufacturers.Tesla has been a ZEV-only company since its inception in 2003;Audi,Fiat,Volvo,and Mercedes-Ben

19、z are targeting ZEV-only sales by 2030;and General Motors and Honda are targeting ZEV-only sales by 2035 and 2040,respectively(Bloomberg New Energy Finance 2022).The combination of policy action and industry goal-setting has led analysts to project that by 2030,PEVs could account for 48%61%of the U.

20、S.light-duty market(Slowik et al.2023).This transition is unprecedented in the history of the automotive industry and will require support across multiple domains,including adequate supply chains,favorable public policy,broad consumer education,proactive grid integration,and(germane to this report)a

21、 national charging network.As established by the Infrastructure Investment and Jobs Act,also known as the Bipartisan Infrastructure Law,the Joint Office of Energy and Transportation(Joint Office)is setting the vision for a national charging network that is convenient,affordable,reliable,and equitabl

22、e to enable a future where everyone can ride and drive electric.This report supports the vision of the Joint Office by presenting a quantitative needs assessment2 for a national charging network capable of supporting 3042 million PEVs on the road by 2030.3 1 This study considers personally owned,lig

23、ht-duty vehicles with gross vehicle weight rating of 8,500 pounds or less.Importantly,this definition includes vehicles driven for transportation network companies(ride-hailing)but excludes motorcycles,light-duty commercial vehicles,and Class 2b and 3 work trucks,the implications of which are discus

24、sed in Section 4 of this report.2 This study is presented as a needs assessment where the national charging network is sized relative to simulated demand from a hypothetical PEV fleet.This is slightly different from an infrastructure forecast,which might make considerations for charging providers be

25、ing incentivized(by private investors or public funding)to future-proof investments,install charging in quantities far exceeding demand,or deploy charging as part of a larger business model that considers utilization as a secondary metric of success.3 National PEV fleet size scenarios have been deve

26、loped using the National Renewable Energy Laboratorys Transportation Energy&Mobility Pathway Options(TEMPO)model and are consistent with multiple 2030 scenarios developed by third parties.Please see Section 2.2.1 for additional details.vi This report is available at no cost from the National Renewab

27、le Energy Laboratory at www.nrel.gov/publications.Estimating infrastructure needs at the national level is a challenging analytic problem that requires quantifying the needs of future PEV drivers in various use cases,under region-specific environmental conditions,and with consideration for the built

28、 environment.This analysis leverages the National Renewable Energy Laboratorys suite of electric vehicle infrastructure analysis tools(EVI-X)and the best available real-world data describing PEV adoption patterns,vehicle technology,residential access,travel profiles,and charging behavior to estimate

29、 future charging needs.Multiple PEV charging use cases are considered,including typical needs to accommodate daily driving for those with and without residential access,corridor-based charging4 supporting long-distance road trips,and ride-hailing electrification.While the analysis is national in sco

30、pe,the simulation framework enables inspection of results by state and city,with parametric sensitivity analysis used to test a range of assumptions.This modeling approach is used to draw the following conclusions:Convenient and affordable charging at/near home is core to the ecosystem but must be c

31、omplemented by reliable public fast charging.Industry focus groups with prospective PEV buyers consistently reveal that consumers want charging that is as fast as possible.However,consumer preferences tend to shift after a PEV purchase is made and lived experience with charging is accumulated.Home c

32、harging has been shown to be the preference of many PEV owners due to its cost and convenience.This dichotomy suggests that reliable public fast charging is key to consumer confidence,but also that a successful charging ecosystem will provide the right balance of fast charging and convenient destina

33、tion charging in the appropriate locations.5 Using sophisticated planning tools,this analysis finds that a national network in 2030 could be composed of 2635 million ports to support 3042 million PEVs.For a mid-adoption scenario of 33 million PEVs,a national network of 28 million ports could consist

34、 of:o 26.8 million privately accessible Level 1 and Level 2 charging ports located at single-family homes,multifamily properties,and workplaces6 o 182,000 publicly accessible fast charging ports along highway corridors and in local communities o 1 million publicly accessible Level 2 charging ports p

35、rimarily located near homes and workplaces(including in high-density neighborhoods,at office buildings,and at retail outlets).In contrast to gas stations,which typically require dedicated stops to public locations,the PEV charging network has the potential to provide charging in locations that do no

36、t 4 This study defines corridors as all roads within the National Highway System(Federal Highway Administration 2017),including the Interstate Highway System,as well as other roads important to national transportation.5 This study considers Level 1 and Level 2 alternating-current(AC)chargers rated b

37、etween 1.4 and 19.2 kW as destination chargers for light-duty vehicles.Direct-current(DC)chargers with nominal power ratings between 150 and 350+kW are considered fast chargers for light-duty vehicles in this work.It is the opinion of the authors that referring to all DC charging as“DC fast charging

38、”(DCFC)(as is typically done)is inappropriate given that the use of“fast”as a descriptor ultimately depends on the capacity of the battery being charged.As larger capacity light-duty PEVs enter the market and medium-and heavy-duty model options emerge,it is likely the case that some DC chargers will

39、 actually be used to slowly charge PEVs.Thus,the common practice of referring to all DC charging as DCFC is noticeably absent from this report.6 This analysis employs a novel charging infrastructure taxonomy that considers workplace charging as a mix of publicly and privately accessible infrastructu

40、re at a variety of location types as discussed in Section 2.3.2.vii This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.require an additional trip or stop.Charging at locations with long dwell times(at/near home,work,or other destinations)ha

41、s the potential to provide drivers with a more convenient experience.This network must include reliable fast charging solutions to support PEV use cases not easily enabled by destination charging,including long-distance travel and ride-hailing,and to make electric vehicle ownership attainable for th

42、ose without reliable access charging while at home or at work.Fast charging serves multiple use cases,and technology is evolving rapidly.The majority of the 182,000 fast charging ports(65%)simulated in the mid-adoption scenario meet the needs of those without access to reliable overnight residential

43、 charging(estimated as 3 million vehicles by 2030 in the mid-adoption scenario).Support for ride-hailing drivers and travelers making long-distance trips accounts for the remainder of simulated fast charging demand(21%and 14%,respectively).While most near-term fast charging demand is simulated as be

44、ing met by 150-kW DC chargers,advances in battery technology are expected to stimulate demand for higher-power charging.We estimate that by 2030,DC chargers rated for at least 350 kW will be the most prevalent technology across the national fast charging network.The size and composition of the 2030

45、national public charging network will ultimately depend on evolving consumer behavior and will vary by community.While growth in all types of charging is necessary,the eventual size and composition of the national public charging network will ultimately depend on the national rate of PEV adoption,PE

46、V preferences across urban,suburban,and rural locations,access to residential/overnight charging,and individual charging preferences.Sensitivity analysis suggests that the size(as measured by number of ports)of the 2030 national public charging network could vary by up to 50%(excluding privately acc

47、essible infrastructure)by varying the share of plug-in hybrids,driver charging etiquette,and access to private workplace charging(see alternate scenarios presented in Section 3.3).Additionally,the national network is expected to vary dramatically by community.For example,densely populated areas will

48、 require significant investments to support those without residential access and ride-hailing electrification,while more rural areas are expected to require fast charging along highways to support long-distance travel for those passing through.Continued investments in U.S.charging infrastructure are

49、 necessary.A cumulative national capital investment of$53$127 billion7 in charging infrastructure is needed by 2030(including private residential charging)to support 33 million PEVs.The large range of potential capital costs found in this study is a result of variable and evolving equipment and inst

50、allation costs observed within the industry across charging networks,locations,and site designs.The estimated cumulative capital investment includes:o$22$72 billion for privately accessible Level 1 and Level 2 charging ports o$27$44 billion for publicly accessible fast charging ports o$5$11 billion

51、for publicly accessible Level 2 charging ports.The cost of grid upgrades and distributed energy resources have been excluded from these estimates.While these excluded costs can be significant in many cases and will 7 The scope of cost estimates can be generally defined as capital expenses for equipm

52、ent and installation necessary to support vehicle charging.Please refer to Section 2.3.4 for additional detail.viii This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.ultimately be critical in building out the national charging network,they

53、 tend to be site-specific and have been deemed out of scope for this analysis.Existing announcements put the United States on a path to meet 2030 investment needs.This report estimates that a$31$55-billion cumulative capital investment in publicly accessible charging infrastructure is necessary to s

54、upport a mid-adoption scenario of 33 million PEVs on the road by 2030.As of March 2023,we estimate$23.7 billion of capital has been announced for publicly accessible light-duty PEV charging infrastructure through the end of the decade,8 including from private firms,the public sector(including federa

55、l,state,and local governments),and electric utilities.Public and private investments in publicly accessible charging infrastructure have accelerated in recent years.If sustained with long-term market certainty grounded in accelerating consumer demand,these public and private investments will put the

56、 United States on a path to meeting the infrastructure needs simulated in this report.Existing and future announcements may be able to leverage direct and indirect incentives to deploy charging infrastructure through a variety of programs,including from the Inflation Reduction Act and the Low Carbon

57、 Fuel Standard,ultimately extending the reach of announced investments.While this analysis presents a needs-based assessment where charging infrastructure is brought online simultaneous to growth in the vehicle fleet,actual charging infrastructure will likely be necessary before demand for charging

58、materializes.The position that infrastructure investment should“lead”vehicle deployment is based on the understanding that many drivers will need to see charging available at the locations they frequent and along the highways they travel before becoming confident in the purchase of an electric vehic

59、le(Muratori et al.2020).On the other hand,infrastructure investment should be careful not to lead vehicle deployment to the point of creating prolonged periods of poor utilization,thereby jeopardizing the financial viability of infrastructure operators.9 These considerations suggest the balance of s

60、upply and demand for charging should be closely monitored at the local level and that steps should be taken to enable the efficient deployment of charging(defined as minimizing soft costs Nelder and Rogers 2019),including streamlined permitting and utility service connection processes(Hernandez 2022

61、).While not the case today,an environment where infrastructure can be deployed efficiently enables the industry to responsively balance the supply of infrastructure subject to forecasts for unprecedented increases in demand.This study leads us to reflect on how charging infrastructure planning has o

62、ften been analogized to a pyramid,with charging at home as the foundation,public fast charging as the smallest part of the network at the tip of the pyramid,and destination charging away from home occupying the middle of the pyramid.While this concept has served a useful purpose over the years,we re

63、commend a new conceptual model.The balance of public versus private charging and fast 8 Based on investment tracking conducted by Atlas Public Policy.9 While utilization is a key metric to most station owners,it is not the only metric of success.Business models underlying charging networks are compl

64、ex and evolving,with some stations collocated with more lucrative retail activities(as is the case with most gas stations today offering fuel at lower margins than items in the convenience store)and some stations deployed at a loss to help“complete”the network in areas critical for enabling infreque

65、nt,long-distance travel.Business relationships between charging networks,automakers,advertisers,and site hosts also make it difficult to measure the success of an individual station from utilization alone.ix This report is available at no cost from the National Renewable Energy Laboratory at www.nre

66、l.gov/publications.charging versus destination charging suggests a planning philosophy akin to a tree,as shown in Figure ES-1.As with a tree,there are parts of the national charging network that are visible and those that are hidden.Public charging is the visible part of the network that can be seen

67、 along highways,at popular destinations,and through data accessible online.Private charging is the hidden part of the network tucked away in personal garages,at apartment complexes,and at certain types of workplaces.This private network is akin to the roots of a tree,as it is foundational to the res

68、t of the system and an enabler for growth in more visible locations.Figure ES-1.Conceptual illustration of national charging infrastructure needs If access to private charging are the roots of the system,a reliable public fast charging network is the trunk,as it benefits from access to charging at h

69、ome and other private locations(a key selling point of PEVs)and ultimately helps grow the system by making PEV ownership more convenient(enabling road trips and supporting those without residential access).While fast charging is estimated to be a relatively small part of the national network in term

70、s of number of total ports,it requires significant investment and is vital to enabling future growth by assuring drivers they will be able to charge quickly whenever they need or want.The last part of the system is a broad set of publicly accessible destination charging locations in dense neighborho

71、ods,office buildings,and retail outlets where the speed of charging can be designed to match typical parking times(“right-speeding”).This network is similar to the branches of a tree in that its existence is contingent on a broad private network and a reliable fast charging network.As with the branc

72、hes of a tree,the public destination charging network is ill-equipped to grow without the support of charging elsewhere.x This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.This analysis envisions a future national charging network that is

73、strategic in locating the right amount of charging,in the right locations,with appropriate charging power.Ensuring that this infrastructure is reliable will be essential to establishing driver confidence and accelerating widespread adoption of PEVs.A successful national charging network will positio

74、n PEVs to provide a superior driving experience,lower total cost of ownership for drivers,become profitable for industry participants,and enable grid integration,all while meeting U.S.climate goals.xi This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/p

75、ublications.Table of Contents Executive Summary.v 1.Introduction.1 1.1.Current State of U.S.PEV and EVSE Markets.2 1.2.Recent Charging Infrastructure Investment and Analysis Studies.3 1.3.Equity Considerations.4 1.4.Report Motivation and Structure.5 2.An Integrated Approach for Multiple LDV Use Case

76、s.6 2.1.Modeling Philosophy and Simulation Pipeline.8 2.1.1.EVI-Pro:Charging Demands for Daily Travel.9 2.1.2.EVI-RoadTrip:Charging Demands for Long-Distance Travel.10 2.1.3.EVI-OnDemand:Charging Demands for Ride-Hailing PEVs.11 2.1.4.Utilization-Based Network Sizing.12 2.2.Demand-Side Consideration

77、s:Defining PEV Use Case Scenarios.13 2.2.1.PEV Adoption and Fleet Composition.15 2.2.2.PEV Technology Attributes.18 2.2.3.Residential Charging Access(Theres No Place Like Home).20 2.2.4.Driving Patterns.23 2.2.5.Charging Behavior.27 2.3.Supply-Side Considerations:Charging Network Terminology,Taxonom

78、y,Utilization,and Cost.28 2.3.1.EVSE Terminology.28 2.3.2.EVSE Taxonomy.29 2.3.3.Network Utilization.30 2.3.4.Cost.33 3.The National Charging Network of 2030.35 3.1.2030 Results by EVSE Taxonomy,PEV Use Case,and Region.35 3.1.1.Results by EVSE Taxonomy.35 3.1.2.Results by PEV Use Case.37 3.1.3.Resul

79、ts by Region.40 3.2.Network Growth From 2022 to 2030.49 3.3.Alternate Scenarios.51 4.Discussion.56 4.1.Philosophical Contribution.56 4.2.Modeling Uncertainty.57 4.3.Cost Estimate Considerations.58 4.4.Critical Topics for Future Research.59 4.5.Accessing EVI-X Capabilities.60 References.61 Appendix:2

80、022 Modeling Comparison.67 xii This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.List of Figures Figure ES-1.Conceptual illustration of national charging infrastructure needs.ix Figure 1.Shared simulation pipeline integrating EVI-Pro,EVI-R

81、oadTrip,and EVI-OnDemand.9 Figure 2.EVI-Pro block diagram for charging behavior simulations and network design.10 Figure 3.EVI-RoadTrip block diagram for traffic generation,charging behavior simulations,and network design.11 Figure 4.EVI-OnDemand block diagram for driver simulations and related assu

82、mptions.12 Figure 5.Conceptual diagram illustrating independent demand estimations,demand aggregation,and integrated network design.12 Figure 6.Composite hourly demand for DC charging by use case for an illustrative region.13 Figure 7.U.S.national light-duty PEV stock under three adoption scenarios.

83、16 Figure 8.Assumed spatial distribution of 33 million PEVs in 2030 by CBSA and state.17 Figure 9.Spatial distribution of new(20192022)LDV registrations by body type.18 Figure 10.Residential charging accessibility scenarios as a function of PEV stock share.In the boxplot figure,the box reflects the

84、inner quartile range(25%75%),with the horizontal line reflecting the median value.Whiskers represent the 5th and 95th percentile values,respectively.21 Figure 11.Likelihood of overnight charging access for ride-hailing drivers for the baseline scenario across all metropolitan CBSAs.22 Figure 12.2017

85、 NHTS auto weekday trip distribution by hour of day and activity type(other”activities include general errands,buy services,exercise,recreational activities,health care visits,religious or community activities,work-related meetings,volunteer activities,paid work from home,attending school as a stude

86、nt,changing type of transportation,attending childcare,and attending adult care).23 Figure 13.National origin-destination data set from Jan.Feb.2020(licensed from INRIX).24 Figure 14.County-to-country origin-destination flows visualized from the FHWA TAF data set.25 Figure 15.Assumed national compos

87、ition of ride-hailing drivers by shift type and residential charging access.26 Figure 16.PEV charging infrastructure hierarchy.29 Figure 17.EVSE taxonomy employed by this analysis.30 Figure 18.Average network utilization across 24,637 ports from December 2021 by location and EVSE type.31 Figure 19.D

88、istribution of average daily port utilization and average peak hour port utilization by location and EVSE type.32 Figure 20.Simulated national DC charging network sized individually by use case and sized by consolidating demand.38 Figure 21.Average daily charging demand simulated by EVI-Pro for typi

89、cal daily travel,broken out by powertrain type,body style,and residential access.39 Figure 22.Average daily charging demand simulated by EVI-OnDemand for ride-hailing use cases,broken out by shift duration and residential access.40 Figure 23.Example charging demand from EVI-RoadTrip overlaid with lo

90、cations of existing DC stations,including those part of the Tesla Supercharger and Electrify America networks.47 Figure 24.Distribution of peak hourly utilization across corridor stations as simulated by EVI-RoadTrip.48 Figure 25.Normalized DC charging demand across CBSAs as a function of worst-case

91、 ambient conditions.49 Figure 26.Simulated cumulative network size(left column)and cumulative investment(right column)between 2022 and 2030.Both private and public infrastructure estimates are shown in the top row,while the bottom row isolates the public network result.50 xiii This report is availab

92、le at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Figure 27.Simulated annual network growth(left column)and investment need(right column)between 2023 and 2030.Both private and public infrastructure estimates are shown in the top row,while the bottom row isolate

93、s the public network result.51 Figure 28.Conceptual illustration of national charging infrastructure needs.56 Figure A-1.Size of the 2022 national charging network as simulated in the national pipeline compared to the actual network as measured by the Alternative Fuels Data Center.67 List of Tables

94、Table 1.Foundational Studies Underlying National Analysis.7 Table 2.Demand-Side Assumptions Used in the Mid-Adoption Scenario.14 Table 3.Description of Select Plausible Alternates to the Baseline Scenario.15 Table 4.Vehicle Model Attributes Used in the Baseline Scenario.19 Table 5.EVSE Capital Cost

95、Assumptions.33 Table 6.Simulated Cumulative National Network Size Through 2030 by Access,EVSE,and Location Types(includes a total of 28 million ports).36 Table 7.Simulated Cumulative National Infrastructure Investment Need Through 2030 by Access,EVSE,and Location Types(a total of$53$127 billion).Exc

96、ludes cost of utility upgrades,distributed energy resources,operating costs,and maintenance costs.37 Table 8.State-Level Port Count Summary for the Simulated 2030 Private Network.41 Table 9.State-Level Port Count Summary for the Simulated 2030 Public L2 Network.42 Table 10.State-Level Port Count Sum

97、mary for the Simulated 2030 Public DC Network.43 Table 11.Port Count Summary for the Simulated Private Network in the Top 10 CBSAs in Terms of Assumed PEV Adoption.44 Table 12.Port Count Summary for the Simulated Public L2 Network in the Top 10 CBSAs in Terms of Assumed PEV Adoption.45 Table 13.Port

98、 Count Summary for the Simulated Public DC Network in the Top 10 CBSAs in Terms of Assumed PEV Adoption.45 Table 14.Top 10 CBSAs by Simulated DC Ports per 1,000 PEVs.46 Table 15.Description of Select Plausible Alternates to the Baseline Scenario.52 Table 16.Relative Port Counts Resulting from Parame

99、tric Sensitivity Analysis.53 Table 17.Relative Infrastructure Costs Resulting from Parametric Sensitivity Analysis.54 Table 18.Summary of Recent 2030 U.S.Charging Infrastructure Assessments.58 1 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publica

100、tions.1.Introduction U.S.climate goals for economywide net-zero greenhouse gas emissions by 2050 will require rapid decarbonization of the light-duty vehicle(LDV)fleet,and plug-in electric vehicles(PEVs)are poised to become the preferred technology for achieving this end(U.S.Department of Energy 202

101、3).The speed of this intended transition to PEVs is evident in actions taken by government and private industry,both in the United States and globally.New PEV sales have reached 7%10%of the U.S.light-duty market as of early 2023(Argonne National Laboratory 2023).Globally,PEV sales accounted for 14%o

102、f the light-duty market in 2022,with China and Europe at 29%and 21%,respectively(IEA 2023).A 2021 executive order(Executive Office of the President 2021)targets 50%of U.S.passenger car and light truck sales as zero-emission vehicles(ZEVs)by 2030,and California has established requirements for 100%li

103、ght-duty ZEV sales by 2035(California Air Resources Board 2022),with many states adopting or considering similar regulations(Khatib 2022).These goals were set prior to passage of the landmark U.S.Bipartisan Infrastructure Law and Inflation Reduction Act,which provide substantial policy support throu

104、gh tax credits and investment grants(Electrification Coalition 2023).Companies in the automotive industry have committed to this transition,with most companies rapidly expanding offerings(Bartlett and Preston 2023)and many pledging to become ZEV-only manufacturers.Tesla has been a ZEV-only company s

105、ince its inception in 2003;Audi,Fiat,Volvo,and Mercedes-Benz are targeting ZEV-only sales by 2030;and General Motors and Honda are targeting ZEV-only sales by 2035 and 2040,respectively(Bloomberg New Energy Finance 2022).The combination of policy action and industry goal-setting has led analysts to

106、project that by 2030,PEVs could account for 48%61%of the U.S.light-duty market(Slowik et al.2023).This transition is unprecedented in the history of the automotive industry and will require support across multiple domains,including adequate supply chains,favorable public policy,broad consumer educat

107、ion,proactive grid integration,and(germane to this report)a national charging network.As established by the 2021 Bipartisan Infrastructure Law,the U.S.Joint Office of Energy and Transportation(Joint Office)is setting the vision for a national charging network that is convenient,affordable,reliable,a

108、nd equitable to enable a future where everyone can ride and drive electric.This report supports the vision of the Joint Office by presenting a quantitative needs assessment for a national charging network capable of supporting 3042 million PEVs on the road by 2030.Estimating infrastructure needs at

109、the national level is a challenging analytic problem that requires quantifying the needs of future PEV drivers in various use cases,under region-specific environmental conditions,and with consideration for the built environment.This analysis leverages the National Renewable Energy Laboratorys(NRELs)

110、suite of electric vehicle infrastructure analysis tools(EVI-X)and the best available real-world data describing PEV adoption patterns,vehicle technology,residential access,travel profiles,and charging behavior to estimate future charging needs.Multiple PEV charging use cases are considered,including

111、 typical needs to accommodate daily driving for those with and without residential access,corridor-based charging supporting long-distance road trips,and ride-hailing electrification.While the analysis is national in scope,the simulation framework enables inspection of results by state and city,with

112、 parametric sensitivity analysis used to test a range of assumptions.2 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.The remainder of Section 1 reviews the current state of the U.S.PEV and electric vehicle supply equipment(EVSE)markets

113、,discusses recent EVSE initiatives and analysis studies,highlights equity considerations in the deployment of charging infrastructure,and outlines the structure used for the remainder of the report.1.1.Current State of U.S.PEV and EVSE Markets Mass-market PEV sales began in the United States at the

114、end of 2010 with just a few models available to consumers.As new plug-in models have been introduced and production volumes have increased,sales have accelerated accordingly.It took nearly 8 years to reach 1 million cumulative sales,but just 2 1/2 more years to reach 2 million cumulative sales in Ju

115、ne 2021.As of February 2023,U.S.cumulative PEV sales have surpassed 3.4 million,with PEV sales at 7%10%of all LDVs in early 2023(Argonne National Laboratory 2023).The growth in PEV sales has been accompanied by a similar growth in PEV capabilities,with electric driving range and maximum charging pow

116、er improving dramatically in recent years.The U.S.Department of Energys(DOEs)Alternative Fueling Station Locator contains information on public and private nonresidential alternative fueling stations in the United States and Canada,including PEV charging infrastructure.PEV charging continues to expe

117、rience rapidly changing technology and growing infrastructure.According to the Station Locator,as of March 2023,about 132,000 publicly accessible charging ports are currently installed in the United States.This includes about 29,000 direct-current(DC)charging ports and 103,000 Level 2(L2)ports.While

118、 strides have been made in recent years to improve interoperability10 of PEV charging,the U.S.network remains fragmented.Today,nearly all U.S.PEV manufacturers equip their new battery-electric vehicles(BEVs)with DC charging inlets compatible with the SAE standard Type 1 Combined Charging System(CCS-

119、1).Tesla,the largest PEV manufacturer in the U.S.and operator of the largest U.S.DC charging network,11 does not follow this standard.Tesla BEVs sold in the U.S.have historically been equipped with a proprietary inlet type exclusive to Tesla with compatible DC chargers available through the Tesla Su

120、percharger network.However,Tesla has recently taken steps to open their charging network.In a November 2022 release,Tesla announced they are opening their connector design to other charging providers and vehicles manufacturers(Tesla 2022).Teslas North American Charging Specification(NACS)is currentl

121、y available at select third-party charging stations,including some locations on EVgos network(EVgo 2023).Tesla has also recently taken steps to open their Supercharger network to other vehicles(Tesla 2023).A small number of Superchargers in New York and California have recently been retrofitted to s

122、upport charging vehicles with CCS-1 inlets relying on activation through the Tesla mobile app.Tesla has announced plans to make 7,500 chargers publicly accessible to non-Tesla PEVs by the end of 2024(including 3,500 Superchargers)(The White House 2023).Finally,Tesla has recently reached agreements t

123、hat will soon give all Ford and 10 While interoperability related to connector compatibility is discussed in the body of the report,interoperability of competing charging networks to allow for roaming is another important dimension.Absence of network-to-network interoperability forces drivers to mai

124、ntain multiple sets of apps and credentials in order to access individual charging networks(a substandard experience relative to the convenience of legacy fueling infrastructure).11 As of March 2023.3 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/p

125、ublications.General Motors customers access to the majority of Teslas North American Supercharger network via adapters,with new Ford and General Motors BEVs being equipped with NACS inlets starting in 2025(Ford Motor Company 2023;General Motors 2023).The U.S.L2 network also remains fragmented,but to

126、 a lesser extent.There are two L2 connectors used in the United States:the SAE J1772 connector(used by all PEV manufacturers except Tesla)and the Tesla NACS connector.The NACS connector is natively only compatible with Tesla vehicles;however,an adapter is available that allows Tesla vehicles to char

127、ge using J1772 connectors.L2 NACS connectors are currently available as part of Teslas network of Destination Chargers and account for 12%of all publicly accessible L2 charging ports.Despite the fragmented nature of todays charging ecosystem,this analysis makes no attempt to develop charging infrast

128、ructure scenarios by connector.Such scenarios would require estimating future market shares and corporate strategies for different light-duty PEV manufacturers to project the future interoperability of charging networks,which is beyond the purview of this analysis.The remainder of this report will n

129、ot address interoperability challenges or fragmentation between connector types.Additional information on PEV charging infrastructure trends can be found on DOEs Alternative Fuels Data Center(2023b).1.2.Recent Charging Infrastructure Investment and Analysis Studies Significant investments are being

130、made in U.S.charging infrastructure for PEVs.At the forefront of these investments is the federal governments commitment to invest up to$7.5 billion into publicly accessible PEV charging infrastructure through the Bipartisan Infrastructure Law.This consists of the$5.0-billion National Electric Vehic

131、le Infrastructure(NEVI)Formula Program administered by the U.S.Department of Transportation through the states,District of Columbia,and Puerto Rico and the$2.5-billion Charging and Fueling Infrastructure Discretionary Grant Program being administered through the U.S.Department of Transportation(the

132、latter including eligibility for all alternative fuel infrastructure).An additional$3.0 billion in public investment has been made across all levels of government,led by programs from the state of California.Atlas Public Policys EV Hub tracks domestic investments in PEV charging infrastructure.As of

133、 April 1,2023,EV Hub reports a cumulative total of$11.2 billion in charging infrastructure announcements from the private sector,led by companies including Tesla,Electrify America,BP,General Motors,Daimler,and Mercedes.This excludes an estimated$3.0 billion in capital raised by charging companies(in

134、cluding ChargePoint,EVgo,Blink,and Volta),some percentage of which is expected to be invested in EVSE hardware and installation.EV Hub reports an additional$2.0 billion in approved utility filings,led by utilities including Southern California Edison,Consolidated Edison,and Pacific Gas&Electric.As o

135、f March 2023,we estimate$23.7 billion has been announced for publicly accessible light-duty PEV charging infrastructure through the end of the decade.12 Importantly,this estimate excludes financial incentives to deploy charging infrastructure through a variety of programs,12 While based on data prov

136、ided by Atlas Public Policy,NRELs estimate deviates from a recent Atlas Public Policy assessment(Nigro 2023),which reports cumulative U.S.public charging infrastructure funding at$19.9 billion.This discrepancy is primarily due to NRELs inclusion of funding assumed to primarily(though not exclusively

137、)support deployment of public charging infrastructure(most notably the Charging and Fueling Infrastructure Discretionary Grant Program,which includes eligibility for all alternative fuel infrastructure).4 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.g

138、ov/publications.including from the Inflation Reduction Act and the Low Carbon Fuel Standard in place in California,Oregon,and Washington.While these incentives are significant and will ultimately extend the reach of announced investments,their value is dependent on factors outside the purview of thi

139、s analysis and are thus excluded from this reports estimate of announced charging infrastructure investments.At least four existing studies have attempted to estimate the national charging infrastructure investment need for light-duty PEVs.The International Council on Clean Transportations(ICCTs)202

140、1 white paper“Charging Up America:Assessing the Growing Need for U.S.Charging Infrastructure Through 2030”estimates that 26 million light-duty PEVs would require a total of 2.4 million workplace and public charging ports(Bauer et al.2021).This results in an estimated$28-billion investment for nonres

141、idential charging infrastructure(including installation labor costs but excluding utility upgrades).When accounting for private-access charging at single-family and multifamily residences(estimated at$20.5 billion),ICCT finds a total of$48.5 billion in cumulative investment will be needed by the end

142、 of the decade.Atlas Public Policys 2021 U.S.Passenger Vehicle Electrification Infrastructure Assessment examined the charging infrastructure investment necessary through 2030 to put the United States on a path to 100%light-duty PEV sales by 2035(McKenzie and Nigro 2021).Atlas finds that$39 billion

143、in public charging infrastructure will be necessary by 2030(including installation labor costs but excluding utility upgrades).When accounting for private-access charging at single-family and multifamily residences and private depot charging,Atlas finds a total need of$87 billion in cumulative inves

144、tment by 2030.McKinsey&Companys 2022 article“Building the electric-vehicle charging infrastructure America needs”examines a scenario with 50%of LDV sales as PEVs by 2030(Kampshoff et al.2022).This analysis estimates 1.2 million public chargers and 28 million private chargers will be necessary by 203

145、0(a 20 x increase over todays network).S&P Global Mobilitys 2023 report EV Chargers:How many do we need?finds that U.S.PEV charging infrastructure will need to quadruple by 2025 and grow by a factor of 8 by 2030(S&P Global Mobility 2023).Assuming 28 million PEVs on the road by 2030,this report estim

146、ates 2.13 million Level 2 and 172,000 DC chargers in public locations will be necessary.These estimates are in addition to privately accessible residential chargers.These findings are all consistent in showing that continued investment in U.S.charging infrastructure is necessary to support the elect

147、rification of the light-duty fleet.A comparison of these findings with this report is included in the discussion section.1.3.Equity Considerations Equitable deployment of charging infrastructure for all populations is of critical importance as investments accelerate.This analysis indirectly addresse

148、s equitable infrastructure deployment by considering the needs of individuals without reliable access to residential charging,drivers for ride-hailing platforms,and(in some cases)ride-hailing drivers without access to residential charging.These individuals are more likely to be from low-income house

149、holds,renters,and those without access to off-street parking.As discussed later in this report,charging infrastructure supporting these populations is explicitly considered in this study.5 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.

150、A broader set of analytic tools that directly address equitable charging infrastructure deployment is being developed by the Joint Office United Support for Transportation(JUST)Lab Consortium with leadership from Argonne National Laboratory,Lawrence Berkeley National Laboratory,and NREL(Joint Office

151、 of Energy and Transportation 2023).The JUST Lab Consortium is conducting actionable research on integrating equity into federally funded PEV infrastructure deployment efforts.This consortium builds on prior efforts at each lab that have developed foundational capabilities,including launch of an Ele

152、ctric Vehicle Charging Justice40 Map(Argonne National Laboratory 2022),application of geospatial analysis to prioritize charging deployments for underserved communities(Zhou et al.2022),and development of the Electric Vehicle Infrastructure for Equity(EVI-Equity)model for quantifying equity metrics

153、of proposed charging network designs(Lee et al.2022).Embedding these tools within the national framework presented in this report is a key objective for future research.1.4.Report Motivation and Structure This report is being published at a unique time in the evolution of the national charging netwo

154、rk.In September 2022,the U.S.Department of Transportation,in consultation and coordination with the new Joint Office,approved Year 1 NEVI plans for all 50 states(plus Washington,D.C.,and Puerto Rico)as part of a$5-billion investment funded by the Bipartisan Infrastructure Law(U.S.Department of Trans

155、portation 2022).In March 2023,the U.S.Department of Transportation opened applications for the first round of funding under the$2.5-billion Charging and Fueling Infrastructure Discretionary Grant Program,also funded by the Bipartisan Infrastructure Law(U.S.Department of Transportation 2023).In the p

156、rivate sector,Tesla continues its trajectory of expanding the countrys largest DC network(including opening some Superchargers to non-Tesla vehicles),Electrify America is halfway through its 10-year,$2-billion mandatory investment period,and many other charging networks are entering the market and e

157、xpanding their footprint.Amidst these ongoing investments,this work aims to provide a shared point of reference for the near-term(through 2030)charging infrastructure needs of U.S.light-duty PEVs.Given the broad coalition of stakeholders dependent on and investing in charging infrastructure(includin

158、g automotive manufacturers,charging network providers,electric utilities,and governments at every level),a public document of this nature can serve as a common reference for the industry.The remainder of this report describes the integrated approach used for estimating needs of multiple LDV use case

159、s(including typical driving needs,long-distance travel,and ride-hailing electrification),introduces and justifies modeling assumptions,describes potential alternate futures,and presents results over time at various levels of geographic resolution.6 This report is available at no cost from the Nation

160、al Renewable Energy Laboratory at www.nrel.gov/publications.2.An Integrated Approach for Multiple LDV Use Cases This report builds on the foundation of years of research and collaboration at NREL and beyond.Several recent analytic works serve as the basis for this study and will be referenced throug

161、hout the remainder of the report(see Table 1).The building blocks of this report include development and ongoing refinement of models used to estimate charging infrastructure needs for light-duty PEVs in multiple use cases.The core tools used in this study are:EVI-Pro:For typical daily charging need

162、s EVI-RoadTrip:For fast charging along highways supporting long-distance travel EVI-OnDemand:For electrification of transportation network companies(TNCs).Each of these models is described in more detail in Section 2.1.In addition to modeling tools,several assumptions must be made to define vehicle

163、use scenarios and estimate the corresponding charging demands.These include scenario-specific assumptions on vehicle adoption(number of PEVs with regional variation),fleet composition(PEV chassis types and preference for BEVs/plug-in hybrid electric vehicles PHEVs),technology attributes(e.g.,vehicle

164、 efficiency/range,charging efficiency/speed),and driving/charging behavior.A key determinant of charging behaviorparticularly the demand for public chargingis the share of PEV owners able to access charging at their primary residence.Home charging is typically the most convenient and affordable char

165、ging location for those that have access,but many do notas discussed at length by Ge et al.(2021).Assumptions for each of these“demand-side”considerations are discussed in Section 2.2.This section concludes by establishing charging network terminology(with help from DOEs Alternative Fuels Data Cente

166、r)and proposes a new charging infrastructure taxonomy that explicitly decouples location type(e.g.,home,work,retail)from access type(e.g.,public,private).Finally,real-world observations of public charging utilization(Borlaug et al.2023)and installed cost(Borlaug et al.2020)are presented as“supply-si

167、de”considerations in Section 2.3.7 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Table 1.Foundational Studies Underlying National Analysis Citation Title Venue Technical Contribution Wood et al.2017 National Plug-In Electric Vehicle In

168、frastructure Analysis DOE Office of Energy Efficiency and Renewable Energy technical report Introduced coverage vs.capacity concept;first national instance of EVI-Pro Wood et al.2018 Charging Electric Vehicles in Smart Cities:An EVI-Pro Analysis of Columbus,Ohio NREL technical report Initial use of

169、large-scale telematics data within EVI-Pro Moniot,Rames,and Wood 2019 Meeting 2025 Zero Emission Vehicle Goals:An Assessment of Electric Vehicle Charging Infrastructure in Maryland NREL technical report Piloted use of EVI-Pro for scenarios with low levels of residential access Borlaug et al.2020 Lev

170、elized Cost of Charging Electric Vehicles in the United States Joule article Compiled public data on installed cost of charging(updated on rolling basis)Alexander et al.2021 Assembly Bill 2127:Electric Vehicle Charging Infrastructure Assessment:Analyzing Charging Needs to Support Zero-Emission Vehic

171、les in 2030 California Energy Commission report Revised EVI-Pro methodology to account for emerging charging behavior observations and implemented demand-based network sizing;introduced EVI-RoadTrip for corridor-based analysis Ge et al.2021 Theres No Place Like Home:Residential Parking,Electrical Ac

172、cess,and Implications for the Future of Electric Vehicle Charging Infrastructure NREL technical report Collected novel survey data on residential parking and electrical access;proposed likely adopter model for estimating evolution of residential access as a function of PEV fleet size Moniot,Ge,and W

173、ood 2022 Estimating Fast Charging Infrastructure Requirements to Fully Electrify Ride-Hailing Fleets Across the United States IEEE Transactions on Transportation Electrification article Developed and applied EVI-OnDemand model for quantifying national infrastructure needs of ride-hailing electrifica

174、tion Alexander and Lee 2023 California Electric Vehicle Infrastructure for Road Trips:Direct Current Fast Charging Needs to Enable Interregional Long-Distance Travel for Electric Vehicles California Energy Commission staff report,forthcoming Technical documentation for EVI-RoadTrip methodology Borla

175、ug et al.2023 Public Electric Vehicle Charging Station Utilization in the United States Transportation Research Part D:Transport and Environment article Quantitative analysis of real-world infrastructure utilization;used as basis for network sizing approach 8 This report is available at no cost from

176、 the National Renewable Energy Laboratory at www.nrel.gov/publications.2.1.Modeling Philosophy and Simulation Pipeline The core tools used in this study are EVI-Pro(for typical daily charging needs),EVI-RoadTrip(for fast charging along highways supporting long-distance travel),and EVI-OnDemand(for r

177、ide-hailing electrification).The development and application of individual models dedicated to specific use cases provides at least two benefits:(1)increased modularity maximizes the flexibility in our modeling;namely,models may be combined or run in isolation(where appropriate),as demonstrated in m

178、any of the studies listed in Table 1;and(2)each model can be tailored to the unique driving and charging behaviors of their associated use case.The models used in this study are a subset of the larger EVI-X modeling suite maintained by NREL for network planning,site design,and financial analysis acr

179、oss light-,medium-,and heavy-duty vehicles(National Renewable Energy Laboratory 2023).LDV use cases vary widely and have unique infrastructure requirements that must be accommodated to facilitate a seamless transition to PEVs.Typical daily use of LDVs tends to be characterized by short trips with lo

180、ng dwell periods(e.g.,70%of daily driving under 40 miles and 95%under 100 miles with vehicles typically parked 95%of their lifetime).These periods present ample opportunities for destination charging(most notably at home and workplace locations)that is“right-speeded”to match typical dwell times.EVI-

181、Pro assumes such an opportunistic approach to charging,attempting to make use of low-cost destination charging where convenient and rely on fast charging only when necessary.13 In contrast,the use of PEVs for long-distance travel and in ride-hailing applications requires that they can pull over in c

182、onvenient locations and charge quickly to either resume a road trip or return to service.EVI-RoadTrip and EVI-OnDemand both employ this charging behavior philosophy but rely on distinct data sets to describe the geographic footprint of long-distance vs.ride-hailing travel patterns.Long-distance trav

183、el requires a network of fast charging stations along highways(including urban and rural areas that these highways pass through),while ride-hailing electrification necessitates access to fast charging within the urban areas where such services are most common(such as near urban centers and airport l

184、ocations).Additional details of each model will be discussed in the following subsections of this report.Each of these individual models is integrated into a shared simulation pipeline,as shown in Figure 1.Models are provided with a self-consistent set of exogenous inputs that prescribe the size,com

185、position,and geographic distribution of the national PEV fleet;technology attributes of vehicles and charging infrastructure;assumed levels of residential/overnight charging access;and regional environmental conditions.Each model uses these inputs in bottom-up simulations of charging behavior by sup

186、erimposing the use of a PEV over travel data from internal combustion engine vehicles.By relying on historical travel data from conventional vehicles,these models implicitly design infrastructure networks capable of making PEVs a one-to-one 13 EVI-Pro assumes fast charging as being necessary only wh

187、en long dwell time opportunities to charge slowly are not present in the detailed driving pattern data sets used as inputs.In reality,charging preferences will be dictated by myriad conditions that are challenging to anticipate in a model.For this reason,EVI-Pro has been configured in this analysis

188、to simulate a minority of BEV drivers(10%)as preferring fast charging over slower alternatives,including opportunities to charge at home.The size of this behavior cohort is believed to be consistent with the limited set of real-world charging behavior observations available in the literature.BEV man

189、ufacturers are arguably in the best position to observe actual charging behavior in the field and are encouraged to consider publishing aggregated charging behavior statistics to inform the efficient deployment of charging infrastructure.9 This report is available at no cost from the National Renewa

190、ble Energy Laboratory at www.nrel.gov/publications.replacement for internal combustion engine vehicles,effectively minimizing impacts to existing driving behavior and identifying the most convenient network of charging infrastructure capable of meeting driver needs.Figure 1.Shared simulation pipelin

191、e integrating EVI-Pro,EVI-RoadTrip,and EVI-OnDemand The independent(but coordinated)simulations produce a set of intermediate outputs estimating daily charging demands for typical PEV use,long-distance travel,and ride-hailing electrification.These intermediate outputs are indexed in time(hourly over

192、 a representative 24-hour period)and space(core-based statistical area CBSA or county level)such that they can be aggregated into a composite set of charging demands across multiple use cases.Once combined,the peak hour for every combination of charging type(e.g.,Level 1 L1,L2,DC),location type(e.g.

193、,home,work,retail),and geography(e.g.,CBSA)is identified for the purpose of network sizing.Rather than sizing the simulated charging network to precisely meet the peak hourly demand in all situations,the simulation pipeline uses an assumed networkwide utilization rate in the peak hour to“oversize”th

194、e network by some margin.This sizing margin accounts for the fact that charging demand tends to vary seasonally and around holidays.As the EVI-X modeling ensemble simulates demand on a typical day,the network sizing approach attempts to account for periods of peak demand,which could far exceed what

195、is experienced on a typical day.This margin is calibrated based on analysis of real-world utilization data,as described later in this section.The resulting final output of the pipeline is a set of charging infrastructure port counts by region,location type,and charging type that can be aggregated up

196、 to the national level or reported out for individual states or CBSAs.The remainder of Section 2.1 will be used to briefly describe the simulation models and data used as the justification for future utilization assumptions.2.1.1.EVI-Pro:Charging Demands for Daily Travel EVI-Pro is a tool for projec

197、ting consumer demand for PEV charging infrastructure under typical daily conditions.EVI-Pro uses detailed data on personal vehicle travel patterns,vehicle attributes,and charging station characteristics in bottom-up simulations to estimate the quantity and type of charging infrastructure necessary t

198、o support regional adoption of PEVs.A block 10 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.diagram of data flows within EVI-Pro is shown in Figure 2.EVI-Pro has been used in multiple detailed planning studies including Wood et al.(20

199、17,2018),Moniot et al.(2019),and Alexander et al.(2021).Figure 2.EVI-Pro block diagram for charging behavior simulations and network design 2.1.2.EVI-RoadTrip:Charging Demands for Long-Distance Travel EVI-RoadTrip projects the amount and locations of DC charging infrastructure needed for BEVs long-d

200、istance travel needs(i.e.,100 miles).This model addresses an under-researched but increasingly important use case for vehicle electrification:long-distance road trips.A fast charging network connecting regions across the nation is critical to accelerate the transition to electric vehicles(EVs)by ena

201、bling timely interregional travel and reducing range anxiety.The model follows three key steps within the context of this analysis(as shown in Figure 3):trip data generation,driving/charging simulation,and station siting/sizing.The model simulates interregional road trips by BEVs(including across st

202、ate lines),estimates energy use and charging demand along the road trip routes,calculates geographic clusters of charging demand,and simulates the existence of charging stations to serve those clusters,typically locating them in locations zoned for retail activity.EVI-RoadTrip was introduced by Alex

203、ander et al.(2021)and is documented in Alexander et al.(2023).11 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Figure 3.EVI-RoadTrip block diagram for traffic generation,charging behavior simulations,and network design 2.1.3.EVI-OnDema

204、nd:Charging Demands for Ride-Hailing PEVs The charging demands from ride-hailing fleets are given unique attention within this study given the aggressive rate of fleet electrification pledged by major ride-hailing companies(Uber 2020;Lyft 2020)and the likely reliance on public infrastructure for man

205、y of these ride-hailing vehicles(Jenn 2020;Moniot et al.2022).Further,ride-hailing vehicles operate distinctly from vehicles used for personal travel and are not comprehensively characterized in travel surveys.These factors motivated the use of EVI-OnDemand for estimating ride-hailing charging deman

206、d.EVI-OnDemand simulates ride-hailing fleets operating in urban areas in a spatially implicit manner given the lack of data made available by prominent ride-hailing companies.The model estimates charging infrastructure necessary to support all-electric ride-hailing fleets with market shares consiste

207、nt with present-day operations.Fleetwide charging demand for each geography is obtained through repeated simulations of heterogeneous drivers,until the total mileage across all drivers matches the projected total within the urban area being evaluated.As shown in Figure 4,drivers are uniquely modeled

208、 based on probabilistic sampling of driver shift length and the likelihood of overnight charging access.These factors influence the demand for fast charging mid-shift,modeled as time-sensitive en route charging.For instance,drivers with short shifts and access to overnight charging are unlikely to r

209、equire access to fast charging infrastructure.In contrast,drivers with longer shifts and no access to overnight charging will depend more heavily on public-access DC charging.The model also considers local driving speeds and ambient conditions to produce plausible energy consumption rates while driv

210、ers are on shift.12 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Figure 4.EVI-OnDemand block diagram for driver simulations and related assumptions The key output from EVI-OnDemand for this study is the aggregate fleetwide demand for

211、DC charging by city to support drivers mid-shift when needed.The aggregate demand for DC charging is disaggregated by time of day by leveraging emerging empirical data in the literature characterizing when ride-hailing vehicles frequent DC chargers(Jenn 2020).Additional documentation of the EVI-OnDe

212、mand simulation model can be found in Moniot,Ge,and Wood(2022)and the model source code(GitHub 2023).2.1.4.Utilization-Based Network Sizing Following independent use case simulations,charging demand from each model is aggregated in time and space to form a composite estimate of demand for each geogr

213、aphy.The peak hourly demand from the composite profile is used to size each component of the network,represented as a combination of location type and charger type(e.g.,public office L2,public retail 150-kW DC).This process is conceptually illustrated in Figure 5.Figure 5.Conceptual diagram illustra

214、ting independent demand estimations,demand aggregation,and integrated network design Demand aggregation allows for the resultant simulated charging network to incorporate resource sharing across different use cases,as is common in the real world(e.g.,ride-hailing PEVs charging alongside road tripper

215、s or employees charging alongside shoppers).This effectively 13 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.reduces the modeled network requirements when contrasted with a counterfactual where the network is synthesized for each use

216、case independently and then summed,since the spatiotemporal charging demands for the different use cases may not necessarily align.An example of this occurrence is shown in Figure 6 for a simulated fast charging network in an illustrative region.Figure 6.Composite hourly demand for DC charging by us

217、e case for an illustrative region 2.2.Demand-Side Considerations:Defining PEV Use Case Scenarios Several input parameters must be specified and synchronized across the three EVI-X models used in this report to estimate comprehensive charging infrastructure needs for light-duty PEVs in the United Sta

218、tes by 2030.This study considers multiple PEV use case scenarios relying on“demand-side”input assumptions,including fleet size,geographic distribution,vehicle and infrastructure technology attributes,residential charging access,and driving/charging behavior.To assess potential futures,a baseline sce

219、nario is first presented using demand-side assumptions shown in Table 2.Plausible alternatives to the baseline scenario are explored using parametric sensitivity analysis as defined by Table 3.These scenarios are not intended to be exhaustive in terms of the potential evolution pathways for the nati

220、onal charging network of 2030,but rather informative of the impacts of various considerations that will be important for charging infrastructure stakeholders to consider.14 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Table 2.Demand-S

221、ide Assumptions Used in the Mid-Adoption Scenario Modeling Parameter 2030 Nominal Assumption PEV fleet size(LDV only)33 million(2.7 million registered as of 2022)PEV powertrain shares BEV=90%(2022:72%)PHEV=10%(2022:28%)PEV body type distribution Sedan=24%(2022:58%)C/SUV=56%(2022:40%)Pickup=17%(2022:

222、0%)Van=3%(2022:2%)Average PEV electric range(model year 2030)BEV=280 miles PHEV=45 miles BEV minimum DC charge time(model year 2030;20%80%state of charge SOC)20 minutes a Maximum DC power rating(per port)350+kW Geographical distribution Scaled proportional to existing PEV and gasoline-hybrid registr

223、ations with a ceiling of 35%of LDVs on the road in 2030 as PEVs in high adoption areas and a floor of 3%in low adoption areas PEVs with reliable access to residential charging 90%Weather conditions Typical ambient conditions are used for each simulated region,impacting electric range accordingly Dri

224、ving behavior EVI-Pro:Consistent with Federal Highway Administration(FHWA)2017 National Household Travel Survey(NHTS)EVI-RoadTrip:Directly applies FHWA Traveler Analysis Framework(TAF)EVI-On Demand:Consistent with Balding et al.(2019)Charging behavior All models attempt to maximize use of home charg

225、ing(when available)and utilize charging away from home only as necessary.When fast charging is necessary,BEVs prefer the fastest option compatible with their vehicle,up to 350+kW.a Tesla recently reported an average charge duration of 27.5 minutes on their Supercharger network(Kane 2023),and a media

226、n duration of 36 minutes has been calculated from public 50-kW DC chargers as part of the EV WATTS program(Energetics 2023).These estimates are provided as context for the 2030 modeling assumption,despite the fact neither statistic necessarily aligns with 20%80%SOC events in all cases.15 This report

227、 is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Table 3.Description of Select Plausible Alternates to the Baseline Scenario Scenario Description High Adoption PEV fleet size growth to 42 million PEVs on the road by 2030(baseline:33 million PEVs by

228、2030)Low Adoption PEV fleet size growth to 30 million PEVs on the road by 2030(baseline:33 million PEVs by 2030)Low Home Charging Access Assumes 85%of PEV drivers with residential access based on the“existing electrical access”scenario from Ge et.al(2021)(baseline:90%residential access)High Home Cha

229、rging Access Assumes 98%of PEV drivers with residential access based on the“potential electrical access”scenario from Ge et.al(2021)(baseline:90%residential access)Reduced Daily Travel PEVs are driven 60%of days,25%less than the baseline(80%of days)Bad Charging Etiquette PEVs are not unplugged durin

230、g public destination L2 charging until the drivers activity at the destination is complete and the vehicle departs(baseline:PEVs are capable of being unplugged when they are finished charging and made available for another PEV)PHEV Success PHEVs retain 2022 PEV market share(28%)through 2030(baseline

231、:PHEVs have 10%PEV market share in 2030)Alternate PEV Adoption PEV adoption is geographically uniform in 2030 with no urban early adopter preference(baseline:geographic distribution of PEVs in 2030 reflects 2022 distribution of PEVs and hybrid electric vehicles)Extreme Weather EVSE network designed

232、for extreme(95th percentile)weather conditions affecting PEV range and increasing charging demand(baseline:EVSE network designed for average weather conditions)Slow TNC Electrification TNC fleets are only 50%PEVs by 2030(baseline:100%TNC PEVs by 2030)Private Workplace Charging 100%of workplace charg

233、ing at private EVSE through 2030(baseline:100%in 2022,decreasing to 50%by 2030)The remainder of this subsection reviews demand-side assumptions in greater detail,including assumptions for fleet size/composition,technology attributes,residential charging access,and driving/charging behavior.2.2.1.PEV

234、 Adoption and Fleet Composition National PEV adoption scenarios were developed using NRELs Transportation Energy&Mobility Pathway Options(TEMPO)model,an all-inclusive transportation demand model that covers the entire United States(Muratori et al.2021).This study examines three TEMPO PEV adoption sc

235、enarios(shown in Figure 7),each of which implicitly assumes the shape of the sales curve between 2022 and 2030.The low adoption scenario assumes 30 million light-duty PEVs on the road by 2030(correlating with 43%of light-duty sales as PEVs by 2030);the mid-adoption scenario assumes 33 million(correl

236、ating with 50%of sales);and the high adoption scenario assumes 42 million(correlating with 68%of sales).This reports baseline scenario uses the mid-adoption national fleet size scenario of 33 million light-duty PEVs on the road by 2030.16 This report is available at no cost from the National Renewab

237、le Energy Laboratory at www.nrel.gov/publications.The TEMPO PEV adoption scenarios are largely consistent with scenarios developed as part of infrastructure analysis studies conducted by ICCT,Atlas Public Policy,McKinsey&Company,and S&P Global Mobility(as described in Section 1.2).These studies cons

238、ider national 2030 PEV fleet sizes between 26 and 48 million.Figure 7.U.S.national light-duty PEV stock under three adoption scenarios As of 2022,PHEVs accounted for 28%of total PEV stock.Recent sales trends and manufacturer announcements suggest the industry is trending toward increased shares of B

239、EVs.The baseline scenario assumes 90%of 2030 PEVs are BEVs,with the remainder of the PEV fleet consisting of PHEVs.The“PHEV Success”scenario is provided to consider potential impacts to the national charging network resulting from PHEVs holding constant at 28%of the growing PEV fleet.Regarding body

240、type,PEV sales to date have been dominated by sedans,accounting for 58%of all PEV registrations in 2022.However,this trend is expected to shift in coming years as the supply of C/SUV and pickup PEVs increases.The baseline scenario assumes the 2030 PEV fleet mirrors the body type distribution of new(

241、2 years old)vehicle registrations in 2022 with 24%sedan,56%C/SUV,17%pickup,and 3%van.The spatial distribution of the 2030 PEV fleet is assumed to be proportional to existing PEV and gasoline-hybrid registrations.As visualized in Figure 8,this approach results in the greatest PEV adoption occurring i

242、n urban areas with up to 35%of LDVs on the road as PEVs in 2030,and the lowest levels of PEV adoption in the rural areas with as low as 3%of LDVs on the road as PEVs in 2030.This assumption is tested using the“Alternate PEV Adoption”scenario,in which PEV adoption in 2030 is assumed uniform across al

243、l states and CBSAs.While this alternate adoption 17 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.scenario is not intended as a projection,it is useful in illustrating the impact of more homogeneous PEV adoption across urban and rural

244、areas.Figure 8.Assumed spatial distribution of 33 million PEVs in 2030 by CBSA and state In addition to modeling regional preferences for PEVs,the baseline scenario also considers regional preferences for body types,as shown in Figure 9.Using 2022 LDV registration data,we find that:Sedans tend to be

245、 most popular in urban areas and rural parts of the Southeast.C/SUVs tend to be most popular in Colorado,Michigan,and the Northeast.Pickups tend to be most popular in rural areas west of the Mississippi River.Vans tend to be most popular in urban and rural areas around the Great Lakes.These trends a

246、re reflected in the adoption scenarios,with the 2030 PEV fleet disaggregated independently by body type using regional preferences reflected in the 2022 LDV registration data for all fuel types.18 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publi

247、cations.Figure 9.Spatial distribution of new(20192022)LDV registrations by body type.Source:Experian LDV registrations 2.2.2.PEV Technology Attributes Eight PEV types are represented in this study,resulting from the combination of two powertrain types(BEV and PHEV)and four body types(sedan,C/SUV,pic

248、kup,and van).Each PEV type includes up to three vintages,referred to as model year groups.The 2020 model year group is meant to capture PEVs sold up to 2020,the 2025 model year group captures PEVs sold between 20212025,and the 2030 model year group captures 20262030.While the actual PEV market is fa

249、r more diverse than this simple representation,the vehicles used in this study are meant to serve as exemplars of the larger market and believed to provide a sufficient level of detail for analysis of 2030 charging infrastructure needs.Table 4 provides a summary of vehicle attributes used in the bas

250、eline scenario.19 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Table 4.Vehicle Model Attributes Used in the Baseline Scenario Vehicle Model Model Year Group Energy Consumption Rate,Wh/mi a Nominal Electric Driving Range,mi Peak DC Cha

251、rge Power,kW Minimum DC Charge Time,minutes b BEV sedan 2020 2025 2030 320 300 300 190 260 290 150 150 250 26 24 20 PHEV sedan 2020 2025 2030 290 290 290 45 50 55 N/A N/A N/A N/A N/A N/A BEV C/SUV 2020 2025 2030 390 430 420 190 240 280 150 150 350 30 30 20 PHEV C/SUV 2020 2025 2030 370 380 370 35 40

252、 40 N/A N/A N/A N/A N/A N/A BEV pickup 2020 2025 2030 570 500 280 300 250 350+24 20 PHEV pickup 2020 2025 2030 440 420 35 35 N/A N/A N/A N/A BEV van 2020 2025 2030 460 440 240 280 150 350 30 20 PHEV van 2020 2025 2030 390 380 35 40 N/A N/A N/A N/A a Excludes charging efficiency losses.Alternating-cu

253、rrent(AC)charging assumed as 90%efficient in all cases.b Assumes 20%to 80%SOC under ideal conditions(preconditioned pack,moderate ambient temperature,no power derating,etc.).Given the adoption trajectory assumed in the baseline scenario,the 2030 PEV fleet in this analysis is dominated by the 2030 mo

254、del year group.Stock turnover and a dramatic increase in projected PEV sales toward the end of the decade result in the 2020,2025,and 2030 model year groups representing 5%,20%,and 75%of the 2030 on-road fleet,respectively.PEV technology is assumed to improve over the period of this analysis,most dr

255、amatically with respect to DC charge acceptance increasing from peak power ratings of 150 kW in the 2020 model year group to 250350 kW in the 2030 model year group.14 Most modern BEVs are capable of relatively high DC charging rates under low-SOC conditions,but as SOC increases during a charging eve

256、nt,a vehicles battery management system begins to taper its charge rate to protect the pack from overvoltage and thermal abuse.14 PHEVs are assumed to be incapable of DC charging in this analysis.20 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/pub

257、lications.This analysis assumes that advances in battery technology(potentially including prevalence of 800-V packs,multilayer cathodes,electrolyte improvements,and advanced charge protocols)will not only enable higher peak power levels at low SOC,but also decrease overall DC charge times.All BEVs s

258、old after 2025 are assumed to be capable of 20-minute DC charge times assuming 20%to 80%state of charge under ideal conditions(preconditioned pack,moderate ambient temperature,no power derating,etc.).In the real world,actual DC charging times will vary based on arrival and departure SOC,pack thermal

259、 conditions(temperatures that are too high or too low will result in power derating),the vehicles battery management system,and the capabilities of the charging station.2.2.3.Residential Charging Access(Theres No Place Like Home)The key enabler for early adoption of PEVs has been home charging at re

260、sidential locations,where vehicles tend to remain parked for long durations overnight.Going forward,there is uncertainty around how effectively home charging can scale as the primary charging location for PEV owners.As the PEV market expands beyond early adopters(typically high-income single-family

261、homes SFHs that have access to off-street parking)to mainstream consumers,planners must consider developing charging infrastructure solutions for households without consistent access to overnight home charging.This includes,but may not be limited to,renters,residents of apartment buildings(and other

262、 multifamily dwellings),and individuals in SFHs without access to off-street parking.In situations where residential off-street charging access is unattainable,a portfolio of solutions may be possible,including providing access to public charging in residential neighborhoods(on street),at workplaces

263、,at commonly visited public locations,and(when necessary)at centralized locations via high-power fast charging infrastructure(similar to existing gas stations).The future of U.S.residential charging access was explored in depth by Ge et al.s(2021)report Theres No Place Like Home.This research review

264、ed public information on residential housing attributes with implicit relation to home charging access,including national data on vehicle ownership,residence type,housing density,and housing tenure(i.e.,rent or own).These public data were complemented by a panel survey sample of 3,772 U.S.individual

265、s to uncover previously unknown distributions of residential parking availability,parking behavior,existing electrical access,and perceived potential for new electrical access by parking location.These responses connected parking availability and existing or potential electrical access to residence

266、type to inform charging access scenarios that were incorporated into the final projection framework.Charging access trends with respect to residence type were identified and coupled with a PEV likely adopter model to infer national residential charging access scenarios as a function of the national

267、PEV fleet size.This work serves as the basis of residential charging access assumptions in this report,which assumes 90%of PEVs have reliable access to overnight charging in a scenario with 33 million PEVs nationwide.Alternate 2030 scenarios for residential access explore home charging as low as 85%

268、and as high as 98%.The distribution of residential access across CBSAs is shown in Figure 10.Note that residential access and fleet size are coupled within the national framework,such that locations with high PEV adoption tend to be estimated with lower levels of residential access,as can be seen fo

269、r CBSAs in California and the Pacific Northwest where residential access decreases over time as the size of the PEV fleet increases.21 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Figure 10.Residential charging accessibility scenarios

270、 as a function of PEV stock share.In the boxplot figure,the box reflects the inner quartile range(25%75%),with the horizontal line reflecting the median value.Whiskers represent the 5th and 95th percentile values,respectively.This analysis pays special attention to the demographics of ride-hailing d

271、rivers,who(consistent with industry goals)are assumed to achieve 100%adoption of PEVs by 2030.Drivers for ride-hailing services are disproportionately lower income,complicating opportunities to leverage data sources representative of the general population.This analysis introduces a means of charact

272、erizing the likelihood of access to overnight charging for ride-hailing drivers.Note that emerging business models,such as leased vehicles with overnight charging at a depot location or leases where public charging is included in the lease of the vehicle,are not explicitly considered.However,such mo

273、dels could be evaluated in the future by assuming greater rates of overnight charging access irrespective of driver housing status or through a driver preference for midday fast charging.Consistent with the approach outlined by Moniot,Ge,and Wood(2022),Ge et al.s(2021)report is once again leveraged

274、for estimating residential access among ride-hailing drivers.Although this survey was intended to be representative of the broader population,the survey produced relationships between demographic descriptorstenure,housing type,and incomeand overnight charging access,which allows for the estimation o

275、f ride-hailing drivers residential 22 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.charging access if their income distribution is known.Ride-hailing driver income data15(Benenson Strategy Group 2020)were combined with demographic dat

276、a from the U.S.Census and information from Ge et al.(2021)to estimate regional-specific residential access rates among ride-hailing drivers.This approach enables differentiation across geographies by accounting for variability in housing stock and household income,leading to consideration of lower o

277、vernight charging access in dense CBSAs(such as New York City)versus more sprawling CBSAs with a greater availability of more affordable housing options with more favorable rates of overnight charging(such as Houston).The baseline scenario distribution of residential access across CBSAs is shown in

278、Figure 11.This distribution results in a national average of 60%for residential charging access among ride-hailing drivers(significantly lower than the 90%assumed for the overall PEV fleet).These CBSA-specific residential access rates are used by EVI-OnDemand when simulating charging behavior among

279、ride-hailing drivers.Figure 11.Likelihood of overnight charging access for ride-hailing drivers for the baseline scenario across all metropolitan CBSAs 15 Driver household income data are used instead of the income obtained exclusively from ride-hailing services.Household income includes additional

280、revenue from separate forms of employment and across all household members.This value is considered to be a more accurate indicator of the type of housing the driver lives in,and also enables direct comparison against household-level census data.23 This report is available at no cost from the Nation

281、al Renewable Energy Laboratory at www.nrel.gov/publications.2.2.4.Driving Patterns PEV driving patterns in this analysis are represented by an ensemble of data sets from conventional vehicles,which are simulated as PEVs to estimate the charging infrastructure necessary for supporting electrification

282、 of LDVs in multiple use cases.EVI-Pro simulations rely on FHWAs 2017 NHTS and a national data set licensed from INRIX.EVI-RoadTrip utilizes FHWAs TAF to describe long-distance driving trends,and EVI-OnDemand employs observations from a Fehr&Peers analysis of the ride-hailing industry in select U.S.

283、markets(Balding et al.2019).As each of these datasets were developed prior to the onset of the COVID-19 pandemic in March 2020,their use within this study imply an assumption that mobility patterns have fully returned to the pre-pandemic state by 2030.Estimating the near-term evolution of personal m

284、obility in the United States was deemed out of scope for this analysis.Driving pattern inputs to EVI-Pro are derived from the 2017 NHTS.The NHTS is a national travel survey conducted every 68 years to describe travel activity at the household level across all transportation modes(e.g.,walk,bike,driv

285、e,ride-hail,transit,air).In addition to being publicly accessible,the NHTS enables“trip chaining,”or the linking of automobile trips in a sequential manner.This is a key feature for PEV charging simulations in EVI-Pro,as it enables battery SOC to be estimated over a 24-hour period.A visualization of

286、 2017 NHTS auto weekday trip distribution by hour of day and activity type is shown in Figure 12 for illustrative purposes.Figure 12.2017 NHTS auto weekday trip distribution by hour of day and activity type(other”activities include general errands,buy services,exercise,recreational activities,health

287、 care visits,religious or community activities,work-related meetings,volunteer activities,paid work from home,attending school as a student,changing type of transportation,attending childcare,and attending adult care)24 This report is available at no cost from the National Renewable Energy Laborator

288、y at www.nrel.gov/publications.While the NHTS data include data points for hundreds of thousands of household vehicles,select cities and states are intentionally oversampled,leaving many geographies with sparse samples.To derive trip chains from all CBSAs and rural counties,a procedure for drawing w

289、eighted samples from the NHTS that are representative of any target geography was developed.This method relies on broadly accessible demographic variables from the U.S.Census to sample household vehicles from the NHTS that are representative of a particular census tract in question.This approach was

290、 calibrated using standard in-sample linear regression techniques and independently validated using out-of-sample travel survey data from the 2012 California Household Travel Survey.One limitation of the NHTS is a lack of spatial information regarding trip destinations.Use of NHTS driving data in EV

291、I-Pro requires that attention be paid to appropriately defining geographies.While geographic precision is often desired,small geographies run the risk of vehicles crossing boundaries during normal operation and placing demand for charging outside the geography in which their“home”is located.To ensur

292、e appropriate spatial resolutions are considered when using NHTS data for EVI-Pro simulations,a spatially explicit analysis was required.For this analysis,we relied on a large,national data set of real-world travel patterns with geocoded trip origins and destinations.The data provider for this analy

293、sis was INRIX,and the data included millions of trips from Jan.Feb.2020(data during the COVID-19 lockdown were intentionally excluded).This data set is visualized in Figure 13.Figure 13.National origin-destination data set from Jan.Feb.2020(licensed from INRIX)Multiple geographies were evaluated usi

294、ng this data set,including counties,census urbanized areas,and CBSAs(including metropolitan and micropolitan statistical areas).For each geography,the frequency of interregional travel was tested and evaluated for suitability of a net-zero charging demand difference in EVI-Pro.This analysis revealed

295、 that CBSAs were the smallest geography with national coverage for which a modeling assumption of net-zero flow in charging demand could be considered valid.Consequently,CBSAs are the default geography for 25 This report is available at no cost from the National Renewable Energy Laboratory at www.nr

296、el.gov/publications.aggregating the individual EVI-Pro simulations that depend on the weighted sampling of NHTS driving days.EVI-RoadTrip relies on long-distance travel data from the TAF.Since long-distance travel tends to be underrepresented in travel surveys and often crosses political boundaries,

297、FHWA developed a synthetic data set with national coverage to estimate long-distance passenger travel.FHWAs TAF was modeled using a variety of predictors,such as population and economic activity,and calibrated to a large travel survey(Federal Highway Administration 2018).TAF consists of a set of cou

298、nty-to-county trip tables for long-distance passenger trips(defined as trips longer than 100 miles)by automobile,bus,air,and rail.The TAF projects person-trip flows for auto travel in 2008 and for 2040,the latter of which is shown in Figure 14.Figure 14.County-to-country origin-destination flows vis

299、ualized from the FHWA TAF data set EVI-OnDemand requires the total passenger miles served by PEVs in ride-hailing fleets in order to estimate charging demands.Few data are available in the literature regarding the share of miles affiliated with ride-hailing fleets outside of an analysis performed by

300、 Fehr&Peers.In the analysis,the authors aggregated real-world ride-hailing miles provided by Uber and Lyft from September 2018 across the six metropolitan areas of Seattle,San Francisco,Los Angeles,Chicago,Washington,D.C.,and Boston.Moniot,Ge,and Wood(2022)compared the total miles across the ride-ha

301、iling fleets for each region against the overall number of vehicle miles traveled(VMT)for the month as reported by the local metropolitan planning organization.It found that ride-hailing fleets comprise between 2%and 3%of VMT within the six regions analyzed,with greater rates of penetration within t

302、he urban cores of each region.The VMT shares found by Fehr&Peers are used for the six regions provided,and a VMT share of 1.5%is assumed for all other regions in lieu of more granular data.The VMT shares reported by Fehr&Peers are assumed to have above-average rates of VMT penetration given the high

303、 household incomes and prominence of technology and information workers in the regions 26 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.analyzed.VMT penetrations for each CBSA were multiplied by the inferred number of vehicle miles tra

304、veled in each CBSA.Total VMT values were obtained at the CBSA level by disaggregating state-level VMT values reported in Table VM-2 of the 2019 Highway Statistics Report(U.S.Department of Transportation 2020)based on vehicle registrations,which were separately sourced from IHS Markit(2017)at the ZIP

305、 code level and aggregated to CBSA and state levels.A key variable influencing the charging demands of ride-hailing vehicles is the time vehicles are assumed to be spent on shift.Full-time drivers operating vehicles for ride-hailing services accrue significantly more miles than part-time drivers and

306、 will thus induce greater demand for charging.However,a greater share of full-time drivers may also reduce the total population of vehicles given the fleet sizing procedure introduced previously.Accurately characterizing drivers based on hours driving per shift or shifts per week is difficult given

307、the lack of publicly available data pertaining to ride-hailing drivers.One study from 2019 found 11%of drivers to be full time using data from RideAustin(Wenzel et al.2019).More recently,a blog post published by an Uber economist(Mishkin 2020)suggested that the vast majority of drivers are part time

308、 through analysis of proprietary driver data sourced from all Uber drivers in California.The assumed national composition of ride-hailing drivers by shift type and residential charging access is shown in Figure 15.Figure 15.Assumed national composition of ride-hailing drivers by shift type and resid

309、ential charging access 27 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.2.2.5.Charging Behavior The final demand-side input into the national framework is assumed PEV charging behavior.Charging behavior assumptions embedded in EVI-Road

310、Trip and EVI-OnDemand are relatively straightforward.In these models,BEVs operate for as long as possible before crossing some range or SOC threshold,then seek out DC charging at the highest possible rate and return to their long-distance trip or ride-hail shift once sufficiently charged.The more co

311、mplicated charging decisions are addressed by EVI-Pro during typical daily driving,particularly for those without residential access.In support of this analysis,many informal conversations with industry stakeholders were conducted.Over these conversations,a consensus emerged on several key points,in

312、cluding:Home is likely the most convenient and cost-effective charging location(for those with access).The industry should take measured steps toward improving access to charging at or near home locations.For those with residential access,PEV technology is progressing in such a way(longer electric d

313、riving ranges)that home is likely the only place that most people will need to charge on a regular basis.For those without residential access,some drivers will find L2 charging away from home to be an effective solution,but only when appropriately collocated with activities with long dwell times(e.g

314、.,8+hours).An interesting point of discussion in these interviews involved the design of fast charging installations,the primary question being“How fast is fast enough?”Historically,a significant share of the publicly accessible DC charging network has been rated at 50 kW.However,there is a recent t

315、rend toward“future proofing”DC stations,with a greater share of new installations at higher power ratings,including up to 350 kW.This trend is motivated by driver preferences for faster charging;however,battery technology tends to be the limiting factor on DC charging times.As previously discussed,m

316、odern BEVs have a maximum DC acceptance rating,which tends to decrease throughout the course of a fast charge event and can further be derated under adverse thermal conditions.Additionally,some destination charging locations may feature typical dwells of over an hour,providing ample opportunity for

317、charging on units rated for 50150 kW.Ultimately,this study elected to employ a baseline charging behavior approach within EVI-Pro that attempts to maximize the use of residential charging as a first priority,then takes advantage of L2 charging away from home at locations with sufficiently long dwell

318、s(typically workplaces),and finally relies on fast charging to meet the needs of drivers that dont have access to home charging and dont exhibit dwell time away from home compatible with L2 charging speeds.16 16 EVI-Pro assumes fast charging as being necessary only when long dwell time opportunities

319、 to charge slowly are not present in the detailed driving pattern datasets used as inputs.In reality,charging preferences will be dictated by a myriad of conditions that are challenging to anticipate in a model.For this reason,EVI-Pro has been configured in this analysis to simulate a minority of BE

320、V drivers(10%)as preferring fast charging over slower alternatives,including opportunities to charge at home.The size of this behavior cohort is believed to be consistent with the limited set of real-world charging behavior observations available in the literature.BEV manufacturers are arguably in t

321、he best position to observe actual charging behavior in the field and are encouraged to consider publishing aggregated charging behavior statistics to inform the efficient deployment of charging infrastructure.28 This report is available at no cost from the National Renewable Energy Laboratory at ww

322、w.nrel.gov/publications.When fast charging is employed within EVI-Pro,the highest rated power unit is selected among the set of 50-,150-,250-,and 350-kW charging so long as the selected charger does not exceed the maximum DC acceptance rate of the vehicle being simulated.The decision to employ charg

323、ing behavior that prioritizes the fastest possible DC charging(when other options have been exhausted)is based on several considerations.First,stakeholder feedback is consistent that when drivers seek fast charging,they prefer fast charging that is at least as fast as what their vehicle is rated for

324、.Second,the industry(to this point)has largely stayed away from pricing models that incentivize fast charging that is only“as fast as necessary.”While there is theoretically potential to optimize installation and operating costs by incentivizing drivers to charge only as fast as necessary,consensus

325、is that such a sophisticated pricing model is inappropriate for this nascent industry.As of 2022,the general population has relatively minimal exposure to PEV charging.Overly complicated pricing models run the risk of introducing detrimental consumer experiences and slowing consumer acceptance of th

326、is new technology.The baseline scenario assumes drivers prefer DC charging that is“as fast as possible.”2.3.Supply-Side Considerations:Charging Network Terminology,Taxonomy,Utilization,and Cost Multiple input parameters must be specified across the three EVI-X models used in this report to estimate

327、the charging infrastructure needs for 33 million light-duty PEVs in the United States by 2030.This subsection reviews critical“supply-side”input assumptions,including EVSE terminology,EVSE taxonomy,network utilization,and infrastructure costs.2.3.1.EVSE Terminology Charging infrastructure terminolog

328、y in this report is consistent with definitions used by the Federal Highway Administration(2023)and is aligned with Open Charge Point Interface(OCPI)terminology for the hierarchy of PEV charging stations,as shown in Figure 16(adapted from DOEs Alternative Fuel Data Center):Station location:A site wi

329、th one or more EVSE ports at the same address.Examples include a parking garage or a mall parking lot.EVSE port:Provides power to charge only one vehicle at a time,even though it may have multiple connectors.The unit that houses EVSE ports is sometimes called a charging post,which can have one or mo

330、re EVSE ports.Connector:What is plugged into a vehicle to charge it.Multiple connectors and connector types(e.g.,Tesla,CCS,CHAdeMO)can be available on one EVSE port,but only one vehicle will charge at a time.Connectors are sometimes called plugs.As discussed in Wood et al.(2017),charging infrastruct

331、ure needs can be thought of in terms of coverage and capacity,wherein coverage needs tend to be defined in terms of number of stations and capacity needs tend to be defined in terms of number of ports.This analysis is primarily concerned with estimating future demand for charging,and thus presents r

332、esults in terms of port counts(as opposed to stations).29 This report is available at no cost from the National Renewable Energy Laboratory at www.nrel.gov/publications.Figure 16.PEV charging infrastructure hierarchy.Source:Alternative Fuels Data Center(2023a)2.3.2.EVSE Taxonomy Traditional EVSE tax

333、onomy approaches adopt a pyramid concept that communicates charging needs in terms of home,workplace,and public charging.This legacy approach has the potential to confuse access type(e.g.,public,private)and location type(e.g.,home,office,retail).Further,the legacy pyramid concept is particularly ambiguous with respect to workplace charging.Work is commonly described as an activity type in travel s