5 Charging Infrastructure for Plug-In Electric Vehicles

July 22, 2016

Ms. Kavita Kale Michigan Public Service Commission 7109 W. Saginaw Hwy. P. O. Box 30221 Lansing, MI 48909

RE: MPSC Case No. U-17990

Dear Ms. Kale:

The following is attached for paperless electronic filing:

Testimony of Douglas Jester on behalf of Sierra Club, Natural Resources and Defense Council, Environmental Law and Policy Center and Michigan Environmental Council

Exhibits SC-1 through SC-9

Proof of Service

Sincerely,

Christopher M. Bzdok [email protected]

xc: Parties to Case No. U-17990 James Clift, MEC ([email protected]) Ariana Gonzalez, NRDC ([email protected]) Laurie Williams, Sierra Club ([email protected]) Shannon Fisk, Earthjustice ([email protected]) Michael Soules, Earthjustice ([email protected]) Robert Kelter ([email protected])

STATE OF MICHIGAN BEFORE THE MICHIGAN PUBLIC SERVICE COMMISSION

In the matter of the Application of CONSUMERS ENERGY COMPANY Case No. U-17990 for authority to increase its rate for the ALJ Hon. Dennis W. Mack generation and distribution of electricity and for other relief.

DIRECT TESTIMONY OF DOUGLAS JESTER

ON BEHALF OF SIERRA CLUB, NATURAL RESOURCES DEFENSE COUNCIL,

ENVIRONMENTAL LAW AND POLICY CENTER, AND MICHIGAN

ENVIRONMENTAL COUNCIL

July 22, 2016

TABLE OF CONTENTS

I. QUALIFICATIONS ...... 3 II. CONSUMERS’ ELECTRIC VEHICLE CHARGING PROPOSAL ...... 7 III. THE COMMISSION SHOULD ACT TO ACCELERATE EV ADOPTION ...... 9 IV. UTILITY EV CHARGING PROGRAM STRUCTURE ...... 25 V. INTEGRATING EV CHARGING WITH THE ELECTRIC POWER SYSTEM .. 33 VI. PROMOTING DEVELOPMENT OF A COMPETITIVE EV CHARGING MARKET...... 35 VII. SPECIFIC VEHICLE CHARGING RECOMMENDATIONS ...... 41 VIII. LINE LOSS REDUCTION AND DISTRIBUTION SYSTEM OPTIMIZATION... 44

July 22, 2016 - U-17990 Direct Testimony of Douglas Jester on behalf of Sierra Club, Natural Resources and Defense Council, Environmental Law and Policy Center, and Michigan Environmental Council Page 1 of 61

1 I. QUALIFICATIONS

2 Q. State your name, business name and address.

3 A. My name is Douglas B. Jester. I am a principal of 5 Lakes Energy LLC, a Michigan

4 limited liability corporation, located at Suite 710, 115 W Allegan Street, Lansing,

5 Michigan 48933.

6 Q. What is the purpose of your testimony?

7 A. I review Consumers Energy’s proposal to initiate an electric vehicle charging program

8 and recommend that:

9  the Commission act to accelerate adoption of electric vehicles through utility

10 engagement in electric vehicle charging infrastructure,

11  the electric vehicle charging program be structured to be strategic,

12 comprehensive, and equitable in the deployment of charging infrastructure,

13  the Commission take steps to ensure that vehicle charging will be well integrated

14 with the electric power system, and

15  the Commission take steps to enable development of a competitive vehicle

16 charging market, while supporting utility engagement in this market.

17 I conclude testimony about electric vehicle charging with specific recommendations as to

18 how the Commission should respond to Consumers Energy’s proposal and structure

19 tariffs for electric vehicle charging as well as how the Commission should provide for

20 evaluation and revision of vehicle charging programs in future cases.

21 In addition, I am testifying concerning Consumers Energy’s proposals for various

22 expenditures on its distribution system. I have testified in case U-17735 that Consumers

1 Energy should be making specific efforts to reduce losses in its distribution system and in

2 case U-17317 that Consumers Energy should be adopting dynamic volt-VAR control and

3 the practice Conservation Voltage Reduction. With respect to both issues, progress can be

4 greatly facilitated by using data from advanced meters, which Consumers Energy will

5 have virtually completed deployment by the conclusion of this case. In this case,

6 Consumers Energy represents that it is advancing on both loss reduction and voltage

7 regulation. I testify in support of those efforts but encourage the Commission to begin

8 expecting Consumers Energy to produce concrete results, by requiring the Company to

9 report its results and by requiring those results to be reflected in the loss factor used to

10 project Power Supply Cost Recovery (PSCR) expenses.

11 Q. On whose behalf are you appearing in this case?

12 A. I am testifying on behalf of Sierra Club, Natural Resources Defense Council,

13 Environmental Law and Policy Center, and Michigan Environmental Council concerning

14 electric vehicle charging. I am testifying on behalf of Michigan Environmental Council

15 and Natural Resources Defense Council concerning Consumers Energy’s proposed

16 distribution system expenditures.

17 Q. Summarize your experience in the field of electric utility regulation.

18 A. I have worked for more than 20 years in regulating the electricity industry and in related

19 fields. My work experience is summarized in my resume, attached as Exhibit SC-1.

20 Q. Have you testified before this Commission or as an expert in any other proceeding?

21 A. Yes. I have testified in:

22 • Case U-17473 (Consumers Energy Plant Retirement Securitization)

1 • Case U-17096-R (Indiana Michigan 2013 PSCR Reconciliation)

2 • Case U-17301 (Consumers Energy Renewable Energy Plan 2013 Biennial Review);

3 • Case U-17302 (DTE Energy Renewable Energy Plan 2013 Biennial Review);

4 • Case U-17317 (Consumers Energy 2014 PSCR Plan);

5 • Case U-17319 (DTE Electric 2014 PSCR Plan);

6 • Case U-17674 (WEPCO 2015 PSCR Plan);

7 • Case U-17679 (Indiana-Michigan 2015 PSCR Plan);

8 • Case U-17689 (DTE Electric Cost of Service and Rate Design);

9 • Case U-17688 (Consumers Energy Cost of Service and Rate Design);

10 • Case U-17698 (Indiana-Michigan Cost of Service and Rate Design);

11 • Case U-17762 (DTE Electric Energy Optimization Plan);

12 • Case U-17752 (Consumers Energy Community Solar);

13 • Case U-17735 (Consumers Energy General Rates);

14 • Case U-17767 (DTE General Rates);

15 • Case U-17792 (Consumers Energy Renewable Energy Plan Revision);

16 • Case U-17895 (UPPCO General Rates);

17 • Case U-17911 (UPPCO 2016 PSCR Plan); and

18 • Case U-18014 (DTE General Rates).

19 In the past, I have also testified as an expert witness on behalf of the State of Michigan

20 before the Federal Energy Regulatory Commission in cases relating to the relicensing of

21 hydro-electric generation. I also have been listed as a witness on behalf of the State of

1 Michigan, prepared case files and submissions, and been deposed in cases before the

2 United States District Court for the Western District of Michigan and the Ingham County

3 Circuit Court of the State of Michigan, concerning electricity generation matters in which

4 the cases were settled before trial.

5 Q. Do you have specific qualifications in relation to electric vehicle charging

6 infrastructure?

7 A. In 2010, I served as an active member of the Commission’s electric vehicle charging

8 collaborative.

9 In 2012, my colleagues and I at 5 Lakes Energy, on behalf of the Pew Charitable Trusts,

10 engaged stakeholders in a number of States in roundtable discussions about the

11 development of electric vehicle infrastructure and drafted a report about best practices,

12 which informed Pew’s subsequent work in this field.

13 More recently, my colleagues and I at 5 Lakes Energy have produced integrated resource

14 planning tools for least-cost compliance with the Clean Power Plan in ten states. These

15 tools incorporate means to model the potential effects of various levels of electric vehicle

16 market penetration on the electricity system.

17 Q. Are you sponsoring any exhibits?

18 A. SC-1 Resume of Douglas B. Jester

19 SC-2 NRDC Report on Driving Out Pollution

20 SC-3 NRC on Overcoming Barriers to Deployment of Plugin EVs

21 SC-4 Richardson Paper on Electric Vehicles and the Grid

22 SC-5 Li et al paper on Indirect Network Effects in Vehicle Charging

1 SC-6 Consumers Energy Response to Discovery 17990-ELPC-CE-98

2 SC-7 Consumers Energy Response to Discovery 17990-ELPC-CE-93

3 SC-8 Lefkowitz Testimony (partial) in Maryland PSC Case 9418

4 SC-9 Consumers Energy Response to Discovery 17990-MEC-CE-49

5 Q. What materials have you reviewed in preparation for your testimony?

6 A. I reviewed Consumers Energy’s application in this case. In addition, there is a substantial

7 literature on electric vehicles and electrical vehicle charging that I have routinely read

8 over the last several years. For a summary of that literature that is most relevant in this

1 9 case, I recommend a recent report by staff of the Natural Resources Defense Council0F0F0F

10 that I sponsor as Exhibit SC-2. I have also reviewed the Commission’s orders and

11 selected portions of the record in cases U-17317 and U-17735. In addition, I cite sources

12 from my accumulated personal library on relevant subjects.

13 II. CONSUMERS’ ELECTRIC VEHICLE CHARGING PROPOSAL

14 Q. Please summarize Consumers Energy’s proposal to develop electric vehicle charging

15 infrastructure?

16 A. Consumers’ proposal is described in the direct testimony of witness Julio Morales, with

17 proposed capital expenditures shown in Exhibit A-49, proposed operations and

18 maintenance expenditures in Exhibit A-50, and locations of proposed fast-charging

19 stations shown in Exhibit A-51. In addition, witness Anne K. Rogus testifies as to the

2 20 revenue requirements1F1F1F for the expenditures described by witness Morales and witness

1 NRDC, 2016. Driving Out Pollution: How Utilities Can Accelerate the Market for Electric Vehicles. 2 Direct testimony of Anne K. Rogus, p 26, lines 3 through 12 and Exhibit A-54.

1 Rachel L. Brege proposes to delete reimbursement for electric vehicle charging

3 2 equipment from the Company’s tariff sheets2F2F2F .

3 Mr. Morales explains that of the proposed approximately $15 million capital expenditure

4 through the end of 2018, about $6 million is for 60 direct current fast-charge stations and

5 about $9 million is for 750 240V alternating current charging stations. He also proposes

6 $150,000 annual operations and maintenance expenditures in the test year and

7 (presumably) thereafter.

8 The fast charging stations will be located two each at 30 locations illustrated in Exhibit

9 A-51, located at approximately 50 mile intervals along major highways. Consumers

10 Energy provided no specific plan for locations of 240V alternating current stations, but

4 11 witness Morales represents3F3F3F that

12 “Installation of the 240V AC chargers will begin in higher- 13 populated metropolitan areas and expand to smaller cities. These 14 stations will be strategically placed in public and private areas to 15 increase visibility of the stations. By identifying locations in 16 Michigan with limited or no charging access, we can help 17 Michigan residents be more comfortable with their decision to 18 purchase a PEV when they see widespread availability of charging 19 stations. We would partner with businesses in high-traffic areas to 20 install one or two charging stations per site, with no installation 21 costs incurred by the host. These locations would include places 22 where people stay for a considerable amount of time, such as 23 restaurants, malls, movie theaters, hospitals, hotels, airports, and 24 large workplaces. These stations will be placed in areas of the 25 community to service customers of all income and socioeconomic 26 levels.”

27 In addition, witness Morales represents that Consumers Energy will provide up to 2500

28 of its customers who purchase an electric vehicle a reimbursement incentive of $1,000

3 Direct testimony of Rachel L. Brege, p 6, line 21 through page 7, line 2 and Exhibit A-11 Schedule F-5, page 42 of 91. 4 Direct testimony of Julio Morales, p 12, line 18 through p 13, line 5

1 toward the installation of a 240V home charging station. However, there is no indication

2 in any of the Company Testimony or Exhibits that Consumers Energy is proposing

3 recovery of the costs of such reimbursement incentives, and reference to past similar

4 incentives is deleted from the tariff sheets rather than amended.

5 According to witness Morales, Consumers Energy proposes that the costs of the fast

5 6 chargers and the 240V alternative current chargers be recovered through base rates4F4F4F .

7 Electricity used at the 240V charging stations would be metered and charged to the site

6 8 host5F5F5F , but does not indicate under what tariff. Consumers Energy offers no indication of

9 how it intends to recover any costs for the fast charging stations.

10 III. THE COMMISSION SHOULD ACT TO ACCELERATE EV ADOPTION

11 Q. Why should the Commission act to accelerate electric vehicle adoption?

12 A. Vehicle electrification will produce a number of general societal benefits, including

13 reductions in air pollution that will benefit public health, mitigation of climate change,

14 improvements in national energy security, and increases in macroeconomic stability. In

15 addition to these general societal benefits, accelerating electric vehicle adoption in

16 Michigan will help to maintain Michigan’s leadership in the automobile industry and will

17 potentially provide substantial benefits to all electric utility customers of Consumers

18 Energy, whether or not they own electric vehicles.

19 Reliable access to electric vehicle charging infrastructure is critical to the growth of the

7 20 electric vehicle market.6F6F6F However, electric vehicle adoption and electric vehicle charging

5 Direct testimony of Julio Morales, p 19, lines 20-21. 6 Direct testimony of Julio Morales, p. 19, lines 1-2. 7 National Research Council, 2015. Overcoming Barriers to Deployment of Plug-In Electric Vehicles. Available from http://www.nap.edu/catalog/21725/overcoming-barriers-to-deployment-of-plug-in-electric-vehicles

1 infrastructure suffer a “chicken-or-egg” market coordination problem that is best

2 addressed through utility engagement in accelerated development of charging

3 infrastructure. Michigan utility engagement can only occur with the support of the

4 Commission, so the Commission should act in this case to accelerate electric vehicle

5 adoption.

6 Q. How does vehicle electrification reduce air pollution and benefit public health?

8 7 A. Michigan Department of Environmental Quality’s 2014 Annual Air Quality Report7F7F7F

8 shows that mobile sources (principally on-road vehicles) are the source of 78.5% of

9 carbon monoxide emissions, 57.9% of lead emissions, 61.8% of nitrous oxide emissions,

10 23% of volatile organic compounds (and 51% of controllable, non-biogenic emissions of

11 volatile organic compounds), and 12.6% of fine particulate matter (PM2.5) emissions.

12 Carbon monoxide interferes with oxygen uptake and transport in all animals and can

13 impair vision, motor function, mental acuity, and work performance. Lead adversely

14 affects kidneys, liver, nervous system, and other organs and is particularly of concern for

15 effects of fetal and childhood exposure on central nervous system development, hence

16 intelligence and behavior. Nitrous oxide causes respiratory distress including asthma

17 exacerbations and may cause structural alteration of lungs. Fine particulate matter

18 aggravates respiratory and cardiovascular problems and has been implicated in heart

9 19 disease, lung disease, and miscarriages. National studies8F8F8F suggest that these are

20 substantial, with premature deaths due to vehicle emissions exceeding those due to

8 DEQ, 2015. Michigan 2014 Annual Air Quality Report. Available from http://www.michigan.gov/documents/deq/deq-aqd-amu-2014_Annual_Air_Quality_Report_492732_7.pdf 9 See Caiazzo, Fabio et al. 2013. Air Pollution and Early Deaths in the United States. Atmospheric Environment 79: 198-208.

10 1 vehicle crashes by more than 50%. Caiazzo et al.9F9F9F estimate that Michigan annually

2 suffers 2484 premature deaths due to PM2.5 from vehicles.

3 Vehicle electrification along with cleaner electricity generation can clearly reduce these

4 emissions and their health effects. While no specific study has determined the effects in

5 Michigan of a transition to electric vehicles given forecasted generation sources, the

6 concentration of vehicle-related emissions in urban and high traffic areas with high

7 human exposure, in contrast to most power plants, suggests that the benefits would be

8 substantial.

9 Q. How does vehicle electrification mitigate climate change?

10 A. The combustion of fossil fuel in vehicles produces carbon dioxide and nitrous oxide, two

11 11 important greenhouse gases. The US DOE Energy Information Administration10F10F10F found

12 that 34% of greenhouse gas emissions in the US in 2009 were from transportation fuels.

12 13 A Center for Climate Strategies report1F11F11F , prepared for the Michigan Department of

14 Environmental Quality, found that on-road vehicles in Michigan produced 56.8 million

15 metric tons CO2-equivalent greenhouse gas emissions, or about 23.6% of Michigan’s net

16 emissions. Thus, any comprehensive effort to mitigate climate change requires significant

17 reductions in fossil fuel use in electric vehicles.

18 All analyses of strategies to mitigate climate change that I have read conclude that

13 19 substantial reduction of greenhouse gas emissions from vehicles is a necessary step12F12F12F ,

10 Ibid., Table 5. 11 Available from https://www.eia.gov/environment/emissions/ghg_report/images/figure3-lg.jpg 12 Center for Climate Strategies, 2008. Final Michigan Greenhouse Gas Inventory and Reference Case Projections 1990-2025. Available from http://pbadupws.nrc.gov/docs/ML1126/ML112650011.pdf. 13 E.g., Williams, J.H. et al. 2012. The Technology Path to Deep Greenhouse Gas Emissions Cuts by 2050: The Pivotal Role of Electricity. Science 335: no 6064, pp 53-59.

14 1 and that the most likely path to do so is vehicle electrification13F13F13F in combination with

15 2 reductions in the carbon intensity of electric power production.14F14F14F Moreover, multiple

3 studies have shown that vehicle electrification reduces greenhouse gas emissions even

16 4 with current generation portfolios. For example, a recent report15F15F15F by the Union of

5 Concerned Scientists illustrates in the following map that electric vehicles charged in

6 Michigan produce greenhouse gasses equivalent to those from a gasoline vehicle that

7 averages 38 miles per gallon, which is higher than the vast majority of gasoline-powered

17 8 vehicles16F16F16F :

14 On-board energy storage can be in the form of voltaic energy in batteries or hydrogen for use in fuel cells, either of which would be charged using electric power. 15 See for example, http://unsdsn.org/wp-content/uploads/2014/09/US-Deep-Decarbonization-Report.pdf, which concludes that, in concert with other power sector trends, 80-95% of all passenger vehicle miles traveled must come from vehicles that use primarily electricity. 16 Union of Concerned Scientists, 2015. Cleaner Cars from Cradle to Grave. Available from http://www.ucsusa.org/clean-vehicles/electric-vehicles/life-cycle-ev-emissions#.V4vXAI-cFJ8. 17 DOE also has a calculator at http://www.afdc.energy.gov/vehicles/electric_emissions.php that compares emissions from powering an electric vehicle to emissions from a comparable internal combustion vehicle. For Michigan, this calculator shows that EVs pollute about 50% less CO2.

2 With announced coal plant retirements and replacement generation coming from a

3 mixture of renewable and natural gas generation, the benefits of vehicle electrification in

4 Michigan will accelerate fairly rapidly.

5 Because only 15 to 17 million passenger vehicles are sold each year nationally, it will

6 take about 15 years of exclusively electric vehicle purchases to largely replace the fleet

7 with electric vehicles. Ramping electric vehicle penetration of new sales to 100% by

8 2035 will require that the annual increment of electric vehicle share of sales average

9 almost 5% per year beginning immediately. Thus, if vehicle electrification is necessary

10 for mitigating climate change, then near-term acceleration of electric vehicle adoption is

11 necessary.

1 Q. How does vehicle electrification improve energy security?

2 A. Despite the effects of fuel efficiency standards and recent increases in US oil production,

3 the United States still imports approximately 25% of our oil consumption and is not

18 4 currently projected to ever reach oil self-sufficiency.17F17F17F Because of the potential

5 disruption to the US economy due to international oil supply interruptions, the US invests

6 substantially in a strategic oil reserve and large military presence in oil-producing

19 7 regions.18F18F18F

8 Since electricity can be produced using a wide variety of technologies and fuels, and in

9 practice all of these are largely domestic, vehicle electrification will reduce the United

20 10 States’ exposure to oil-related risks. As a result, the US Department of Energy found19F19F19F

11 that “reliance on oil is the greatest immediate threat to US economic and national

12 security.... Vehicle efficiency has the greatest short- to mid-term impact on oil

13 consumption. Electrification will play a growing role in both efficiency and fuel

21 14 diversification.”20F20F20F

15 Q. How does vehicle electrification increase macroeconomic stability?

16 A. Oil price and supply shocks have been a significant contributing factor to economic

17 recessions. “All but one of the 11 postwar recessions were associated with an increase in

18 the price of oil, the single exception being the recession of 1960. Likewise, all but one of

19 the 12 oil price episodes listed in Table 1 were accompanied by US recessions, the single

18 EIA, 2016. Annual Energy Outlook 2016. Available from http://www.eia.gov/forecasts/aeo/. 19 DOD, 2014. 2014 Quadrennial Defense Review. Available from http://archive.defense.gov/pubs/2014_Quadrennial_Defense_Review.pdf. 20 DOE, 2011. Report on the First Quadrennial Technology Review. Available from http://cms.doe.gov/sites/prod/files/ReportOnTheFirstQTR.pdf. 21 M.R. Copulos, and A.J. Liska & R.K. Perrin (2010) The Hidden Cost of Oil Securing Foreign Oil: A Case for Including Military Operations in the Climate Change Impact of Fuels

1 exception being the 2003 oil price increase associated with the Venezuelan unrest and

22 2 second Persian Gulf War.”21F21F21F Further, these episodes have particularly acute effects on the

3 automobile industry (hence Michigan) as is suggested by the following table of real GDP

4 growth (annual rate) and contribution of autos to the overall GDP growth rate in five

23 5 historical oil shock episodes.2F22F22F

24 7 Since the auto industry has accounted for 4.5% to 2.8% of GDP23F23F23F during this period,

8 contributions of this magnitude to GDP change by the auto industry illustrates substantial

9 auto industry recessions, and in some cases the recession was entirely in the auto industry

10 while the rest of the economy grew, as indicated by an auto industry contribution to the

11 recession that is larger than the size of the recession itself.

12 The principal mechanisms by which oil shocks cause recessions are through large shifts

13 in balance of payments for oil imports and large shifts in automobile product mix demand

25 14 that cannot be satisfied with existing capacity24F24F24F . Vehicle electrification will contribute to

15 reduced oil imports, weakening the transmission of oil shocks to aggregate demand.

16 Electricity prices are more stable than oil prices, so vehicle electrification will reduce or

22 Hamilton, J. 2013. Historical Oil Shocks. In Parker, R. E..and R. Whaples, 2013. Handbook of Major Events in Economic History. Preprint available from http://econweb.ucsd.edu/~jhamilton/oil_history.pdf. 23 Ibid. 24 Bureau of Economic Analysis, from http://bea.gov/industry/gdpbyind_data.htm. 25 Hamilton, J. 2013. Historical Oil Shocks. In Parker, R. E..and R. Whaples, 2013. Handbook of Major Events in Economic History. Preprint available from http://econweb.ucsd.edu/~jhamilton/oil_history.pdf.

1 eliminate the effects of oil prices on product demand shifts. Thus, vehicle electrification

2 will increase macroeconomic stability for the United States and especially for Michigan.

3 Q. How does accelerating electric vehicle adoption in Michigan help to maintain

4 Michigan’s leadership in the automobile industry?

5 A. As the automobile industry shifts from internal combustion engines to electric motors,

6 Michigan’s dominance in the automobile industry may be challenged. Toyota initiated its

7 development and production of the Prius in Japan. California-based Tesla has already

8 established itself as a major contributor to the early development of the US electric

9 vehicle market. Although Michigan-based manufacturers are now establishing a

10 significant presence in the electric vehicle segment of the auto industry and investments

26 11 by both federal and Michigan governments established a Michigan-based battery25F25F25F and

12 electric drive train supply chain, relatively little of the electric vehicle charging supply

13 chain is reported to have developed in Michigan.

27 14 Michael Porter26F26F26F and many other researchers who have followed through on his seminal

15 works have shown that high levels of regional economic performance derive from the

16 formation and maintenance of industrial clusters with many reinforcing supportive

17 factors. Amongst the factors that are typically important to both formation and

18 maintenance of such clusters is the presence of a strong domestic market for the cluster’s

19 output. It follows that strong adoption of electric vehicles in the Michigan market will

20 contribute toward maintenance of Michigan’s dominance in the automotive industry as it

21 transitions to electric vehicles.

26 See ELPC’s 2015 report on Michigan’s clean energy supply chain companies, at http://elpc.org/2015-clean- energy-supply-chain-reports/#michigan 27 See particularly Porter, M. 1980 Competitive Strategy, Free Press, and Porter, M. 1990. The Competitive Advantage of Nations.

1 Q. How does accelerating electric vehicle adoption potentially benefit electric utility

2 customers?

3 A. Electric vehicle charging will increase electricity sales, which if well integrated into the

4 electric power system can dilute the fixed costs of transmission and distribution and

5 lower electricity rates for all utility customers. An electric vehicle “can be recharged

28 6 while its owner is sleeping, eating, working, or doing anything other than driving.”27F27F27F

7 Consequently, if electric vehicle charging is well-integrated into the near-future electric

8 power system, it can “fill valleys” in load without proportionally increasing overall

9 capacity requirements; this can reduce the average cost of power for all utility customers.

10 As variable renewable resources like wind and solar generation gain larger shares of

11 electric power generation, flexible electric vehicle charging can add value to the electric

12 power system by facilitating the integration of these resources and balancing electricity

13 generation with demand; this can stabilize power flows and reduce the average cost of

14 power.

15 Q. How much will vehicle electrification contribute to utility sales?

29 16 A. According to EPA fuel economy labels28F28F28F for electric vehicles, current model electric

17 vehicles use between 28 kWh and 54 kWh per 100 miles, with most models that have

18 significant sales using between 35 kWh and 42 kWh per 100 miles. I assume for this

19 illustrative calculation that future vehicles will average 40 kWh per 100 miles. According

30 20 to the Michigan Department of Transportation29F29F29F , vehicle miles traveled in Michigan in

28 A nice turn of phrase in SC-NRDC-ELPC-2: NRDC, 2016. Driving Out Pollution: How Utilities Can Accelerate the Market for Electric Vehicles. 29 These can be viewed at fueleconomy.gov. 30 See Table AVMT13, available from http://www.michigan.gov/documents/mdot/MDOT_TOTALS_LENAVMT_372811_7.pdf.

1 2013 totaled 95,136,461,000. If this amount of vehicle travel had been fully electrified,

2 then electric vehicles would have consumed about 38.055 TWh. This would have been a

3 36.8% increase in electricity sales. Of course, this amount will scale with electric vehicle

4 adoption and will therefore develop only gradually.

5 Q. How much would vehicle electrification dilute fixed costs of transmission and

6 distribution?

7 A. Many details are important to such a calculation. However, for a rough approximation I

8 perused the annual reports of major Michigan utilities and determined that approximately

9 60% of electric utility revenue is to recover generation costs and about 40% is for

10 transmission, distribution, customer service, and administration. If non-generation costs

11 could remain unchanged and generation costs per kWh were unchanged as a result of

12 adding load to fully electrify vehicle travel in Michigan, then average rates would be

31 13 reduced by about 11%30F30F30F . In the alternative, rates could be held constant if generation

14 costs per kWh were unchanged and the costs of transmission and distribution increased

15 by as much as 37%. It is likely that some additions to distribution system costs, in

16 particular, will be required if electric vehicles are ubiquitous but nonetheless likely that

17 the net effect will be significant dilution of fixed costs of transmission and distribution

18 over enlarged electricity sales.

19 Q. Does vehicle electrification cancel the benefits of Energy Optimization?

32 20 A. No. The 2013 Energy Efficiency Potential Study31F31F31F performed by GDS Associates for the

21 Commission, DTE Energy and Consumers Energy estimated economic potential

31 This is calculated by multiplying the generation share of costs by the percentage increase in load, adding unchanged transmission and distribution costs, and dividing the result by the increased load. 32 Available at http://www.michigan.gov/documents/mpsc/mi_ee_potential_study_rep_v29_439270_7.pdf

1 electricity efficiency as 36.1% of projected 2018 sales, absent energy efficiency

2 programs. This is roughly the same amount of electricity as would be required to fully

3 electrify vehicle travel in Michigan.

4 While vehicle electrification likely will counterbalance electricity sales reductions

5 through increased efficiency, this does not diminish the value to customers of greater

6 energy productivity in their businesses and residences. Vehicle electrification, while

7 adding back the sales reduced by Energy Optimization, creates new value for customers

8 and society as described earlier in my testimony.

9 Q. How much can “valley-filling” by electric vehicle charging reduce the average cost

10 of power?

33 11 A. Pacific Northwest National Laboratory32F32F32F found that nationally there is sufficient

12 generation capacity to charge almost all passenger vehicles through “valley-filling”.

13 Michigan currently has total generation capacity of about 28 GW, providing

14 approximately 103 TWh per year for a load factor of about 42%. If vehicle electrification

15 added 37 TWh generation per year and this load was accommodated by “valley-filling”,

16 then this load factor would rise to 57%. A 57% load factor is somewhat high for most

17 utilities but not unreasonable with the load-scheduling flexibility of electric vehicles.

18 Assuming consistent with the current generation portfolio that generation capacity

19 represents an average of 25% of total utility costs and that fuel and other variable costs

34 20 represent an average of about 35% of total utility costs, then a revision3F33F33F of the

33 Kintner-Meyer, M., K. Schneider, and R. Pratt, Impacts Assessment of Plug-in Hybrid Vehicles on Electric Utilities and Regional U.S. Power Grids, Pacific Northwest National Laboratory, November 2007, energyenvironment.pnnl.gov/ei/pdf/PHEV_Feasibility_Analysis_Part1.pdf. 34 In this case, multiplying only the variable costs of generation by the increased load, adding the unchanged costs of distribution, transmission, and generation capacity, then dividing the result by the increased load.

1 calculation I made above concerning the dilution of fixed costs suggests that vehicle

2 charging would increase utility sales by 37% but only increase utility costs by about 13%

3 so that rates would be reduced by 17.6%. In the alternative, rates could be held constant if

4 the incremental costs of transmission, distribution, and generation capacity to support

5 electric vehicle charging were less than 50% of the current costs of transmission,

6 distribution, and generation capacity.

7 In Exhibit SC-2, the NRDC authors present the following graph illustrating a similar but

8 more detailed analysis for San Diego Gas and Electric, consistent with my results.

10 Q. To what extent can electric vehicle charging buffer the variability of wind and solar

11 generation?

12 A. Two strategies for integrating electric vehicle charging with generation from renewables

13 have been the subject of recent studies. One strategy focuses on integration at a utility

1 customer site, usually combining solar generation with building loads and electric vehicle

2 charging. The other strategy, more relevant here, focuses on integration at utility scale.

35 3 Exhibit SC-434F34F34F is a good summary of some of that work which concludes that “[t]he

4 existing literature is fairly unanimous and conclusive in its assessment that EVs can

5 increase the amount of renewable energy that can be brought online while reducing the

6 negative consequences for the grid.” This conclusion is based in part on a number of

7 studies that look at regional and national scale balancing and show that smart electric

8 vehicle charging allows significantly greater increases in renewable generation than the

9 amount of vehicle charging load. With 50% of US electricity generation from wind, the

10 required regulation services can be provided by electrification of just 3.2% of the vehicle

11 fleet and operating reserves can be provided by electrification of 38% of the vehicle

36 12 fleet.35F35F35F In short, vehicle electrification is a key enabler of very high penetration of

13 renewable generation and is nearly sufficient for that purpose.

14 Michigan is far from a level of renewables penetration where electric vehicle charging or

15 other new storage options are necessary for renewable resource integration to the grid.

16 However, given the current power sector market trends and reinforcing policies that are

17 shifting the nation’s generation mix towards greater renewables penetration, it is prudent

18 to prepare for the strategic integration of these resources and explore other valuable grid

19 services that electric vehicles can provide. Thus, the Commission should be mindful of

20 this long-run benefit but remain focused on the rate reduction that electric vehicles offer

21 through dilution of fixed costs and load “valley-filling”.

35 Richardson, D. 2013. Electric vehicles and the electric grid: A review of modeling approaches, impacts, and renewable energy integration. 36 Kempton, W and J Tomic. 2005. Vehicle-to-grid power implementation: from stabilizing the grid to supporting large-scale renewable energy. Journal of Power Sources 144: pp 280-294.

1 Q. What is the market coordination problem between electric vehicle adoption and

2 electric vehicle infrastructure development?

3 A. A driver is reluctant to purchase an electric vehicle unless vehicle charging infrastructure

4 is generally available, since the absence of charging infrastructure limits the uses of an

5 electric vehicle and hence reduces its value to the driver. On the other hand, businesses

6 cannot see a business case for providing electric vehicle charging infrastructure if there

7 not enough electric vehicles in use to provide sufficient use and revenue to repay the

8 investment. This problem is common in network industries and has been studied in

9 contexts including but not limited to information technology hardware, software,

10 telecommunications, broadcasting, markets for information, banks and ATMs, and

37 11 airlines.36F36F36F The universal effect of these coordination problems is that such a market

12 grows or changes more slowly than the market optimum, sometimes to the point that it

13 never develops. The particular form of this coordination problem present in the case of

14 electric vehicle charging is called “indirect network effects”. Indirect network effects

15 arise because a decision by one driver to buy an electric vehicle increases the demand for

16 vehicle charging infrastructure, supply of which attracts electric vehicle purchase(s) by

17 other driver(s); thus one purchase indirectly increases other purchase(s). In the case of

18 electric vehicle charging, there are indirect network effects on both sides of the market.

37 See Shy, Oz. 2001. The Economics of Network Industries. Cambridge University Press.

1 Q. Why is this market coordination problem best addressed through utility

2 engagement in accelerated development of charging infrastructure?

38 3 A. Exhibit SC-4 is a recent paper37F37F37F that specifically estimates the quantitative elements of

4 this coordination problem. The authors estimate that a 10% increase in the number of

5 non-residential charging stations will increase EV sales by 8% and that a 10% increase in

6 the number of EVs will increase non-residential charging station deployment by 6%.

7 Thus any non-market “shock” to the supply of either electric vehicles or charging stations

8 will produce a “virtuous circle” of feedback between the two markets that will

9 significantly accelerate electric vehicle adoption. They further show based on their

10 parameter estimates that a given financial subsidy to electric vehicle infrastructure will

11 increase electric vehicle sales by more than twice the amount of increase if the financial

12 subsidy is offered for electric vehicle purchase.

13 They also apply their model to each Standard Metropolitan Statistical Area defined by the

14 US Census to determine whether, given current numbers of electric vehicles and non-

15 residential charging stations, investment in charging stations or electric vehicles should

16 lead the development of the local market. I have inserted their map below for ready

17 reference.

38 Li, S. et al. 2016. The Market for Electric Vehicles: Indirect Network Effects and Policy Design. SSRN 2515037.

2 Their analysis does not include rural areas, but for that part of Consumers Energy’s

3 service territory that falls within SMSAs, the map shows that accelerated investment in

4 charging infrastructure is clearly the better leading investment. My opinion, based on

5 careful reading of this work, is that such is also true of the rural areas in Consumers

6 Energy’s service territory.

7 Exhibit SC-3 is a 2015 report of The National Research Council Committee on

8 Overcoming Barriers to Electric Vehicle Deployment. After examining the case for

9 various entities to provide electric vehicle charging infrastructure in various settings, the

10 committee concluded with respect to electric utilities:

11 “The electric utility companies could emerge as a willing source of 12 capital for public charging stations. That conclusion reflects the

1 prospect that a network of public charging stations would induce 2 more utility customers to purchase PEVs, which would lead not 3 only to electricity consumption at the public chargers, but also to 4 much greater consumption of electricity at residences served by the 5 utilities. If public charging infrastructure drives greater eVMT and 6 greater deployment of vehicles, capital and variable costs for 7 public infrastructure might be covered by the incremental revenue 8 from additional electricity that PEV drivers consume at home, 9 where roughly 80 percent of PEV charging takes place (Francfort 39 10 2011).”38F38F38F

11 No entity other than the electric utility is able to benefit from the indirect network effects

12 of providing non-residential charging stations, especially in settings where additional

13 market failures prevail (which I discuss below). It is therefore uniquely possible for a

14 utility to strategically scale and equitably locate charging infrastructure during early

15 development of the electric vehicle market. Thus it is logical that, if the Commission is

16 moved by the benefits described above to accelerate the adoption of electric vehicles,

17 then the logical strategy is to support utility investment in electric vehicle charging

18 infrastructure.

19 Further, because the utility already has established connections to its customer base it is

20 also well positioned to provide education and outreach to both potential electric vehicle

21 drivers and charging site hosts. The benefit of increased electricity sales from electric

22 vehicle load should also incentivize the utility to leverage its existing customer

23 relationships to meaningfully engage potential electric vehicle drivers and site hosts on

24 the aforementioned benefits of vehicle electrification.

25 IV. UTILITY EV CHARGING PROGRAM STRUCTURE

26 Q. How should utility programs be structured in order to accelerate electric vehicle

39 Ibid, p 92.

1 adoption?

2 A. There are two essential features such programs must have. First, they must

3 comprehensively meet the growing vehicle charging needs of electric vehicle drivers.

4 Second, they must equitably enable electric vehicle adoption.

5 Q. What is necessary to comprehensively meet the vehicle charging needs of electric

6 vehicle drivers?

7 A. This is shaped by the technical possibilities for vehicle charging and depends on the type

8 of electric vehicle and driving pattern of the driver.

9 Chapter 2 of Exhibit SC-5 is a detailed discussion of charging technologies. I summarize

10 the most salient points here. The industry has developed standards and equipment for

11 three types of charging.

12 AC Level 1 Charging standard is for charging equipment that plugs into a 120 V wall

13 outlet and delivers up to 12 amps to a SAE J1772 plug that connects into a socket in the

14 car. AC Level 1 equipment is typically carried in the car and enables charging wherever

15 there is access to a “wall outlet”. At 12 amps, an AC Level 1 charger transfers energy at a

16 rate of 1.4 kW. Each hour of AC Level 1 charging adds range of 4 to 5 miles, depending

17 on vehicle efficiency and driving conditions.

18 AC Level 2 Charging standard is for charging equipment that uses 240V, split-phase

19 alternating current circuit and connects to the car through a SAE J1772 plug. AC Level 2

20 charging allows up to 80 amps of current, which would transfer up to 19 kW power but

21 many vehicles cannot accept that throughput. Most residential circuits and many small

22 commercial circuits cannot support that much current, so common installations are 40

23 amps or less. Each hour of charging at maximum current for AC Level 2 could add

1 approximately 60 miles to vehicle range but vehicle and circuit limits make 20 to 30

2 miles per hour of charging more representative.

3 DC Fast Charging has multiple, competing, incompatible “standards”. All fast chargers

4 installed in the United States are CHAdeMO or Tesla superchargers. Tesla superchargers

5 only work with Tesla vehicles. Faster charging is accomplished by connecting a high-

6 amperage direct current directly to the vehicle battery, unlike the AC chargers which go

7 through an AC-DC conversion on-board the vehicle. Only certain vehicle models,

8 discussed below, are currently able to accept fast charging. Fast charging capability was

9 only recently added to the SAE J1772 standard, as a distinct “combo plug” and only a

10 few recently announced vehicles will be able to accept fast charging through the SAE

11 J1772 combo connection. CHAdeMO fast chargers typically are able to transfer energy at

12 the rate of 44 kW, which can add range to a typical compatible vehicle at a rate of more

13 than 100 miles per hour of charging. Tesla charging stations are currently undergoing an

14 upgrade from 90 kW to 120 kW energy transfer rate, which adds range at a rate of 200

15 miles or more per hour of charging.

16 It should be apparent that AC Level 1 and AC Level 2 charging is suitable for either quite

17 limited driving range or long-dwell vehicle parking. Fast charging is intended to support

18 longer distance (highway) travel but still requires a stop of sufficient duration that most

19 customers will require comfort and alternative activity while waiting for charging to

20 complete.

21 Two forms of wireless charging are emerging as commercial offerings. One form is

22 functionally similar to AC Level 1 or 2 charging, but cordless. The other is dynamic

1 wireless charging that occurs along the roadway while the vehicle is in motion, but this is

2 currently only being implemented for fixed, short route vehicles.

3 A significant number of plug-in electric vehicle models are produced or have been

4 announced, with a variety of specifications. A number of them are intended for only local

5 use and are purely electric with modest battery capacity and AC charging (Limited-range

6 BEV). Two approaches have been taken for vehicles that are used for greater distances.

7 Plug-in hybrid vehicles (PHEV) can be powered electrically but also have on-board

8 engines such that in short-range usage they function as electric vehicles but for extended-

9 range usage they function more like a typical gasoline hybrid vehicle. Long-range battery

10 electric vehicles (Long-range BEV) rely exclusively on electricity but use large batteries

11 and fast charging to support extended-range travel. Most recently announced models are

40 12 battery electric vehicles with range of at least 80 miles.39F39F39F

13 Given these technologies, the evolving paradigm for charging infrastructure to

14 comprehensively meet the needs of electric vehicle drivers is to supply AC Level 1 or AC

15 Level 2 charging in places where people naturally park for extended periods and DC Fast

16 Charging along travel corridors. The various charging and vehicle technology

17 combinations and the related effects of infrastructure are well summarized in Table 5-1 of

18 Exhibit SC-3, reproduced here for ready reference.

40 https://www.fueleconomy.gov/feg/pdfs/guides/FEG2016.pdf, which does not yet list the Chevrolet Bolt that is reported to have a range of about 200 miles.

2 The typical electric vehicle is driven 4% of the time, is parked at home 50% of the time,

41 3 and is parked elsewhere 46% of the time.40F40F40F In most cases, the majority of time parked

4 elsewhere is at the workplace.

5 Q. What is necessary to equitably enable electric vehicle adoption?

6 A. In order to equitably enable electric vehicle adoption, each infrastructure category needs

7 to be equivalently available to all potential electric vehicle drivers. In particular, AC

8 charging at home is a “virtual necessity” and must potentially be available before a

9 potential electric vehicle driver will make an electric vehicle purchase. Employers with

10 employees who commute any significant distance will need workplace charging. For

41 Exhibit SC-2, Figure 3.

1 extended range travel using battery electric vehicles, fast charging must be available

2 along enough routes to effectively connect most trip origin-destination combinations.

3 Q. What is your evaluation of Consumers Energy’s proposal by these criteria?

4 A. Consumers Energy’s proposal is incomplete and is not well-focused.

5 The foundational vehicle charging infrastructure category is home charging. According to

6 witness Morales, Consumers Energy proposes to address this need by offering $1,000

7 reimbursement incentives to up to 2500 customers in the next three years, toward the

8 costs of installing AC Level 2 Chargers. Consumers Energy apparently has neglected to

9 identify cost recovery for these reimbursement incentives or to make tariff provisions,

10 and it is unlikely that they will provide these incentives without cost recovery. These

11 reimbursement incentives will likely accrue primarily to single-family homeowners with

12 access to privately-owned on-site parking. The reimbursement approach will likely

13 exclude residents of multifamily residences who face fundamental market failures that

14 would likely prevent them from acquiring a charging station on their own.

15 Witness Morales makes little mention of programmatic consideration of multifamily

16 residences. Most multifamily housing has a shared parking area, typically without

17 assigned parking. Someone who lives in a multifamily setting and is contemplating an

18 electric vehicle purchase faces a number of challenges not faced by an owner-occupant of

19 a single-family dwelling. Parking is a common area not under exclusive control of the

20 erstwhile electric vehicle owner, so some kind of permission will be required. Exclusive

21 control of a parking place equipped for charging may be difficult, and shared

22 infrastructure may be appropriate to the setting. Costs of charging infrastructure at

23 remove from the building, such as in a parking lot, will likely be higher than installation

1 in a single-family house garage. In the case of a renter, investment in charging

2 infrastructure may not be recoverable within their expected tenure. In response to

42 3 discovery41F41F41F , Consumers Energy’s response concerning whether the proposed placement

4 of 750 240V chargers included an effort to locate them at apartments or condominiums

5 was that “The current infrastructure program does not include a segment dedicated to

6 multi-family applications. The current program is structured to expand PEV

7 infrastructure to the broad base of customers, whereas multifamily residence applications

8 could minimize the accessibility of those stations to a select few. In most cases,

9 customers residing in multifamily residences will be eligible for the residential

10 reimbursement component of the program.” This response indicates that Consumers

11 Energy has not thought through the challenges of home charging for residents of

12 multifamily properties, nor valued their opportunity to use an electric vehicle. Inclusion

13 of a well-formulated program for multifamily residential parking is both virtually

14 essential to electric vehicle adoption by the residents of such housing and to equitable

15 treatment of these residents vis-à-vis Consumers Energy’s proposed offer of

16 reimbursement incentives in a fashion that is clearly better targeted to single-family

17 dwellings.

18 Witness Morales also makes no mention of charging infrastructure for fleets. School

19 buses, local delivery fleets, local transit fleets, garbage trucks, and similar short-range

20 fleets are typically parked overnight in a way that is analogous to residential charging.

21 The second-most important charging location is the workplace. While witness Morales’

22 description of Consumers Energy’s 750-station AC Level 2 charging proposal includes

42 Exhibit SC-6

1 “large workplaces”, his characterization of the hosts to be targeted seems more focused

2 on serving the patrons of the host businesses than the employees. This interpretation is

3 reinforced by his responses to discovery as shown in Exhibit SC-6. Thus, Consumers

4 Energy’s proposal to deploy 750 240V AC charging stations falls more into the Intracity

5 Level 1 and 2 categories shown in the preceding table. On-site workplace charging

6 potentially provides a focused benefit to employees and thereby provides value to the

7 employer; employees may be able to negotiate the provision of vehicle charging

8 infrastructure in on-site employee parking. Workers in downtown areas where parking is

9 primarily in shared public or private parking systems are unlikely to be able to negotiate

10 provision of electric vehicle charging in the same way that they might for on-site parking

11 at their workplace, for reasons similar to those that impeded at-home charging for

12 residents of multifamily dwellings. Thus, there is arguably a greater need for Consumers

13 Energy to engage in the provision of charging infrastructure in shared “public”

14 workplace-oriented parking than in exclusive workplace-oriented parking.

15 Intracity AC Level 1 and 2 charging like that proposed by Consumers Energy can add

16 value for an electric vehicle owner, so it should not be neglected. However, Consumers

17 Energy’s plan does not seem well targeted to provide that value. Dwell time of customers

18 varies considerably amongst the types of businesses listed. The host selection process

19 does not appear to be designed to identify and give preference to locations where trip

20 lengths suggest particular value for range extension. AC Level 1 or 2 charging at airports

21 is likely to result in long occupancy periods when the vehicle is not actually charging.

22 Consumers proposes that the charging stations will be virtually free to the hosts, so there

23 will be no market pressure to guide host selection. Finally, Consumers indicates that they

1 will place stations to serve “all income and socioeconomic levels” but does not explain

2 how this will be accomplished.

3 Consumers Energy’s proposed locations for Intercity Fast Charging stations are not

4 unreasonable, but its analytical support is thin for such relatively substantial

43 5 investments.42F42F42F

6 Finally, Consumers Energy has not provided a plan nor committed to systematic

7 evaluation of results from the program so that experience and learning will guide

8 subsequent charger placement.

9 V. INTEGRATING EV CHARGING WITH THE ELECTRIC POWER SYSTEM

10 Q. How should the Commission ensure that vehicle charging will be well integrated

11 with the electric power system?

12 A. There are four issues for the Commission to consider.

13 First, as I outlined earlier, it is possible to support substantial vehicle charging load

14 without significant additional generation capacity, through load “valley-filling”. The

15 most effective way to do this is through clear price signals that are passed through as

16 actual costs to the person making charging decisions. Absent such price signals, the

17 driver of an electric vehicle will have no reason – and likely no awareness of the need- to

18 avoid charging at high load times. I therefore recommend that the Commission require

19 that all vehicle charging be done through time-of-use tariffs with critical peak pricing.

20 Second, pricing will be more effective and demand response in the charging process

21 becomes possible if the vehicle and charging equipment are enabled to automate the

43 Exhibit SC-7

1 charging process. The relevant open standard for this is the Smart Energy Profile 2.0, as

44 2 embedded in SAE J2836/143F43F43F . The Commission should require Consumers Energy adopt

3 and use that standard in its AC Level 2 charging infrastructure, to the extent practicable.

4 In addition, the Commission should expect that with this capability in place, Consumers

5 Energy will engage in efforts to capture and appropriately share with customers the value

6 that electric vehicle charging can provide as grid services, which can include frequency

7 response, traditional and advanced demand response – all of which are very valuable in

8 the grid now but will become increasingly valuable with increasing renewables

9 penetration. The following table from Exhibit SC-2 summarizes the value of such

10 services in several recent studies:

12 Third, the Commission should be concerned that increasing load from electric vehicle

13 charging not add to the carbon emissions of the grid. Moreover, most purchasers of

45 14 electric vehicles have environmental motivations4F44F44F . I therefore recommend that

15 electricity sales through public charging stations provided by Consumers Energy,

44 SAE J2836/1 can be obtained from http://standards.sae.org/j2847/1_201311/ 45 Exhibit SC-5.

1 including fast charging and AC Level 1 and 2, be 100% obtained through Consumers’

46 2 green power programs.45F45F45F

3 Fourth, incremental load due to charging may drive some distribution circuits to capacity

4 or affect (positively or negatively) voltage, power factor, power quality or other

5 operational characteristics. The extent of these effects depends on charging technology,

6 usage, and grid conditions so it is important for Consumers Energy and the Commission

7 to learn from experience. The Commission should require Consumers Energy to perform

8 appropriate analyses to minimize such costs when siting fast charging stations and to

9 annually develop and submit to the Commission a report of the effects of the AC Level 2

10 stations on the distribution system.

11 VI. PROMOTING DEVELOPMENT OF A COMPETITIVE EV CHARGING

12 MARKET

13 Q. Why should the Commission promote development of a competitive electric vehicle

14 charging market?

15 A. First, it is a well-established conclusion of economics that in the long-run effective

16 competition produces better prices and greater supply of services. Secondly, this is a

17 period of rapid innovation in the electric vehicle and vehicle charging markets and the

18 Commission should avoid locking-in a particular business model or set of technologies

19 for vehicle charging infrastructure.

46 Station hosts who acquire equipment themselves or through a third-party service must be enabled to choose to participate in the green power program as well.

1 Q. How should the Commission promote development of a competitive electric vehicle

2 charging market, while supporting utility engagement?

3 A. It is important to understand in some detail the structure of costs and scope of potential

4 competition for vehicle charging. The following diagram represents the approach Pacific

5 Gas & Electric (PG&E) has taken to vehicle charging infrastructure and is a useful

6 reference for examining this question.

8 The EV Service Connection to the left in this diagram is on the utility side of the meter

9 and is within the scope of the traditional “natural monopoly” of a distribution utility.

10 The EV Supply Infrastructure that connects from the meter to the end-use equipment

11 seemingly falls within the traditional scope of the customer premise and not the utility. In

12 a residence or business with charging infrastructure for only one or two vehicles, it is

13 likely that the natural physical implementation will be to use the current premise meter

1 and wire the charging equipment connection to the existing or an enlarged electric service

47 2 panel.46F46F46F However, as electric vehicles become ubiquitous and parking lots come to have

3 large numbers of vehicle chargers, this will not make sense. Instead, we will likely have a

4 separate service connection to the parking lot and it is likely that metering and payment

5 will be centered in the charger. In this likely future, the meter demarcation of the utility

6 boundary would naturally be at the charger.

7 In the California proceedings concerning utility engagement in electric vehicle charging,

8 the combination of EV Service Connection and EV Supply Infrastructure has been

9 termed as site “make-ready”. For the two utility cases that have come before the

10 California Public Utilities Commission for decision, these costs have been authorized to

48 11 be included in utility base rates47F47F47F . When electric vehicles become ubiquitous this is likely

12 the natural utility regulatory structure and I recommend that the Michigan Public Service

13 Commission adopt this approach now in anticipation of that future.

14 The EV Charger Equipment to the right in this diagram is analogous to other end-use

15 equipment that is normally supplied by competitive markets; there are currently a number

16 of competitors in the marketplace for manufacturing, installing, and servicing such

17 equipment. This is also the locus of innovation activity in vehicle charging technology

18 and business models and should therefore be the focus of any effort by the Commission

47 While the use of an existing meter is the simplest physical configuration for 1-2 chargers, the ideal configuration may be separate or sub-metering of EV load, depending on the applicable electricity rate structure. 48In addition to authorizing “make-ready” costs, the Commission approved additional recovery for the EV charging equipment. In San Diego Gas & Electric’s program, the capital cost of the charging stations will be included in the base rates, and the utility will own the stations. In Southern California Edison’s program, the hosts of the EV charging equipment will receive a rebate from the utility for a portion of the equipment unit cost. The rebates vary by market segment (50% in multi-unit dwellings, 25% for fleets, workplaces and destination centers, and 100% in disadvantaged communities), and will be treated as expenses, to be recovered from ratepayers in the year the expense is incurred. California Public Utilities Commission, Decisions 16-01-045 (Decision Regarding Underlying Vehicle Grid Integration Application and Motion to Adopt Settlement Agreement) and 16-01-023 (Decision Regarding Southern California Edison Company’s Application For Charge Ready and Market Education Programs), available at http://docs.cpuc.ca.gov/DecisionsSearchForm.aspx.

1 to promote development of a competitive market for vehicle charging. There are three

2 ways that competition can occur in this realm: utility procurement, utility-facilitated

3 market access, and open market competition. In the utility procurement approach, the

4 utility would specify and purchase equipment; this approach risks stifling market

5 competition and innovation if the utility’s specifications and price competition dominate

6 the process. Utility-facilitated market access would provide a list of qualified vendors

7 who will work with the utility but could potentially allow competing business models and

8 technologies to compete for selection by the prospective site host. Open market

9 competition would provide all of the benefits of competition but risks shifting onto the

10 site host considerable transaction costs in the form of learning about electric vehicle

11 charging, vendor selection, coordination of utility and vendor installation, and resolution

12 of any operating problems. I recommend that the Commission require Consumers Energy

13 to facilitate open market competition in addition to either utility procurement or utility-

14 facilitated market access.

15 Q. What about payment services as shown in this diagram?

16 A. Consumers Energy’s proposal with respect to payment is problematic. Firstly, they

17 propose that site hosts pay for the energy delivered through the AC Level 2 chargers.

18 This completely obliterates any price signal to electric vehicle drivers and will lead to

19 higher utility costs as charging occurs on-peak. It also means that these charging stations

20 will likely only be located where the host can capture sufficient economic value to make

21 the expense worthwhile, an approach that has not been demonstrated successful for

22 vehicle charging and is highly to result in placement with hosts where higher-income

23 customers enable higher sales margins. Further, if Consumers Energy effectively makes

1 vehicle charging free to vehicle drivers and transfers all costs of vehicle charging

2 equipment into base rates, then there will be no scope for open market competition. And,

3 failure to charge appropriately for use of chargers in the kinds of public locations

4 Consumers Energy proposes to target will inevitably result in inefficient utilization of the

5 chargers.

6 Second, Consumers Energy has not explained how it proposes to charge for use of fast

49 7 chargers.48F48F48F

8 I therefore recommend that the Commission require Consumers to recover at least the

9 costs of delivered power, and perhaps some of the costs of vehicle charging equipment,

10 from vehicle drivers. This will necessitate use of a payment network. Payment networks

11 are ubiquitous in our society, predominantly in the form of bank credit and debit card

12 networks. In some economic sectors, including vehicle fuels, some businesses have

13 developed value-added payment networks (usually based on bank card networks).

14 Consumers Energy should be required to accept bank card payment but should allow

15 other forms of payment including both cash payment to the host and value-added

16 services.

17 Q. If make-ready costs are paid by Consumers Energy and rate-based, and electric

18 vehicle drivers pay for delivered power when charging, how do you recommend that

19 costs of electric vehicle charging equipment be recovered?

20 A. There are two basic approaches available, both of which are compatible with both

21 development of a competitive market and with utility engagement in this market. The

22 first alternative is to charge the electric vehicle driver in addition to the delivered energy

49 Current tariffs are not generally designed for the kind of load pattern that will occur at fast charging stations and are unlikely to accurately capture cost causation or reflect the attributes of charging that are valuable to customers.

1 costs. This could be in the form of a markup on energy delivered, but in long-dwell

2 parking a car that is fully charged can continue to occupy a space and thereby block

3 access to charging stations by other drivers. It might therefore be more appropriate to

4 charge for parking space occupancy rather than marking up energy, while keeping the

5 basic energy charge. However, during market development when vehicle charging

6 infrastructure is leading vehicle sales, this approach may not be able to recover sufficient

7 revenue at reasonable prices. At the same time, during market development most

8 charging stations will be local monopolies in which unregulated pricing could be

9 excessive, risking electricity prices that eliminate fuel cost savings and may likely exceed

10 gasoline prices, so the Commission should ensure that pricing is appropriate for use of

11 charging stations in which Consumers Energy invests. This issue is particularly acute for

12 EV drivers at multi-family dwellings, where the on-site charging stations are the only

13 reasonable option for their primary overnight charging needs.

14 The second alternative is to allow a station host to pay for the equipment, either upfront

15 or in “rental” rate via monthly charges that include maintenance and operations as well as

16 recovery of and on capital. This approach is particularly attractive during this period of

17 market development when infrastructure will be leading electric vehicle purchases as it

18 provides a way for a site host to subsidize the station as a benefit to employees or

19 customers. However, it is unlikely that this approach will work for fast-charging stations

20 as they are more expensive than the AC Level 2 charging stations and with short dwell

21 time may not offer as much value to a station host; in the long run, fast charging station

22 costs likely need to be charged directly to the electric vehicle driver even if the market

23 develops such that hosts pay some or all of the costs of AC Level 2 charging stations.

1 During the market development period when charging infrastructure leads electric vehicle

2 ownership, utilization rates of equipment may not be high enough to support full cost

3 recovery of the costs of electric vehicle charging equipment. AC Level 2 host “skin in the

4 game” partly addresses this need but host contributions likely will not be sufficient for

5 fast charging stations and for AC Level 2 charging stations in situations with structural

6 market failures like multifamily dwellings and “public” parking that benefits multiple

7 employers or businesses. If the Commission treats initial development of vehicle

8 charging infrastructure as a development project, there is room for some Company or

9 ratepayer subsidy of charging equipment during early market development. If the

10 Commission pursues this path, however, it should ensure that any ratepayer subsidy is

11 equally available for host-owned and third-party owned charging equipment as well as

12 for Company-owned charging equipment. Another attractive approach to this market

13 development period could be some temporary risk-sharing on the development of sales

14 volume through application of debt-rate capital charges rather than blended-debt-and-

15 equity rate capital charges on a regulatory liability associated with such equipment.

16 VII. SPECIFIC VEHICLE CHARGING RECOMMENDATIONS

17 Q. Please summarize your preceding recommendations.

18 A. I want to clearly recommend that the Commission act to accelerate adoption of electric

19 vehicles through utility engagement in electric vehicle charging infrastructure. However,

20 in doing so the Commission needs to attend to a number of deficiencies of Consumers

21 Energy’s proposal.

22 Provisions for residential charging must explicitly include provisions for multifamily

23 residential settings and address the specific market failures that occur in those settings.

1 Provision should also be made for inclusion of fleet charging. Workplace charging should

2 be a higher priority than Intracity AC Level 2 charging and should extend to small

3 employers and shared “public” parking. Intracity AC Level 2 charging should be better

4 targeted to locations with longer dwell times and where patron trip length indicates

5 significant value-added, and host sites should have some “skin in the game” by paying

6 for equipment (with some potential exceptions in order to provide equitable opportunities

7 for electric vehicle use). Fast charger locations should be based on a more analytical

8 approach to siting, based in traveler origin-destination patterns and route choices.

9 Electricity used in vehicle charging should ordinarily be paid by the electric vehicle

10 driver based on time-of-use rates with critical peak pricing, to incent “valley filling” load

11 profiles. In support of smart charging practices, Consumers Energy should support the

12 use of the Smart Energy Profile 2.0 to shape load, delivered through equipment that

13 complies with SAE J2836/1. Power used in vehicle charging should be provided through

14 Consumers Energy’s green power programs. The Commission should also require

15 Consumers Energy to provide annual reports on its experience with distribution system

16 effects and costs related to electric vehicle charging and to incorporate lessons learned in

17 subsequent siting and development activities.

18 In providing for cost recovery of an electric vehicle charging program, the Commission

19 should allow charging station “make-ready” costs for both Company and third-party

20 charging stations to be recovered through general rates, reflecting both an expectation

21 that charging will become ubiquitous and that rate dilution due to extra energy sales will

22 allow these costs to be recovered within the rate trajectory absent vehicle electrification.

23 Costs of delivered power used in vehicle charging should ordinarily be paid by the

1 electric vehicle driver, although the Commission should not foreclose host decisions to

2 pay such charges or different approaches at third-party charging stations, assuming limits

3 for reasonable electricity pricing. In the long run, costs of AC Level 2 charging

4 equipment for multifamily residential parking, employee parking, or intracity parking can

5 be recovered through either charges to the host site or parking fees. In the long run, costs

6 of fast charging should be paid by electric vehicle drivers. In the short run, there will be

7 insufficient sales volume for full cost recovery, due to the “chicken or egg” problem; the

8 Commission should consider sharing subsidy and volume risk responsibility with

9 Consumers Energy in order to enable utility engagement in accelerating the adoption of

10 electric vehicles.

11 Q. Do you have any other recommendations with respect to electric vehicle charging?

12 A. Yes. Leading the market requires learning by doing. The Commission should actively

13 engage in such learning both to ensure that Consumers is actively learning but also for

14 the benefit of the Commission and other stakeholders. To that end, I recommend that the

15 Commission constitute an ongoing stakeholder collaborative led by the Commission and

16 use that venue for continuous monitoring and review of Consumers electric vehicle

17 charging demonstration project.

18 As both accountability for the effectiveness of Consumers Energy’s efforts and as

19 material for learning within the collaborative, the Commission should require Consumers

20 to periodically report certain information to the Commission. This should include but not

21 be limited to stations proposed, planned, and implemented by market segment and

22 geography; make-ready and charging equipment costs; station usage and load patterns;

23 distribution system impacts; host and customer satisfaction and issues; electric vehicle

1 sales and vehicle miles traveled in Michigan. In addition, the Commission should require

2 Consumers Energy to contract for and deliver to the Commission program evaluation

3 research that examines the effects of Consumers Energy’s charging infrastructure on

4 electric vehicle sales and usage; the effectiveness of Consumers Energy’s grid integration

5 of vehicle charging; implications of ubiquitous vehicle charging on Consumers Energy’s

6 future distribution system architecture; costs of vehicle charging infrastructure and

7 whether it has the expected dilution effect on Consumers Energy rates; and the effects of

8 Consumers Energy’s programs on development of a competitive market for vehicle

9 charging equipment and services.

10 VIII. LINE LOSS REDUCTION AND DISTRIBUTION SYSTEM OPTIMIZATION

11 Q. What is Consumers Energy proposing in this case with respect to its distribution

12 system?

13 A. Consumers Energy witness Andrew J. Bordine presented the Company’s plans for

14 distribution system investments and changes, which are also summarized in Exhibits A-

15 15, A-16, A-17, A-18, and A-19. According to Mr. Bordine, the Company’s primary

16 emphasis is to improve distribution system reliability through vegetation management,

17 service restoration readiness, corrective maintenance, and asset management programs

18 for the distribution system. In addition, Consumers Energy is proposing Grid

19 Modernization expenditures to “support system conditioning to reduce losses and support

20 automation, Grid Analytics, and the deployment of a D[istribution] M[anagement]

21 S[ystem]”.49F49F49F50 Mr. Bordine explains that “Grid Analytics will leverage

22 information generated by Advanced Meter Infrastructure (“AMI”) and Gird

50 Direct testimony of Anderew J. Bordine, p 40, line 7-8.

1 Modernization systems to help identify system losses, overloaded equipment, and

2 automatic phase identification.”. He also describes a proposed System Conditioning

3 activity which he describes as “a systematic and targeted program to address system

4 limitations in providing energy efficiency and adequate customer voltage.”

5 Mr. Bordine describes general benefits of these activities as including “Reduction of

6 Outage Frequency”, “Reduction of Outage Duration”, “Energy Efficiency & System

7 Optimization”, as well as other operational benefits.

8 Mr. Bordine also describes other distribution system expenditures about which I will not

9 be testifying.

10 Q. Does Consumers Energy project reductions in losses as a result of these programs?

11 A. Yes. In discovery response 17990-MEC-CE-49, which I am sponsoring as Exhibit SC-9,

12 Mr. Bordine indicates that the Company projects a 1% reduction in system losses from

13 these programs.

14 Q. Do you think that these expenditures are appropriate?

15 A. I think that these expenditures are directionally appropriate. Further, by the completion of

16 this case, Consumers Energy will be approaching complete deployment of AMI, which

17 significantly enables the direction they propose. Thus, it is timely that the Company

18 propose and the Commission approve the Company’s efforts in this direction in this case.

19 However, Consumers Energy has not provided sufficient forecast of or commitment to

20 results for me to fully evaluate the actual amounts proposed. I am therefore

21 recommending that the Commission approve the revenue recovery requested in

22 conjunction with Consumers Energy’s Grid Modernization and System Conditioning

1 activities but require Consumers Energy to report to the Commission on both the work it

2 accomplishes and the benefits achieved through its distribution system investments.

3 Q. Has the Commission recently addressed the need to evaluate opportunities to reduce

4 energy losses and optimize distribution system operations in planning to replace

5 aging distribution infrastructure?

6 A. Yes. In its Final Order dated November 19, 2015 in the last Consumers Energy rate case,

7 Case No. U-17735, the Commission stated at page 93:

8 Notwithstanding, the Commission recognizes the importance of 9 understanding potential opportunities to reduce energy waste 10 through the mitigation of line losses in the event such opportunities 11 are cost-effective relative to other investments. Given that the 12 functioning of the grid and replacement of aging distribution 13 infrastructure will likely be of ever increasing importance in the 14 coming years with the advent of emerging technologies, the 15 Commission finds it is important to examine distribution planning 16 in a holistic manner and base investment decisions on strong 17 analytical support of the costs and benefits. The Commission 18 therefore directs the Staff to engage with stakeholders on the 19 process going forward, to educate and enhance understanding of 20 this complex issue. 21 In its Final Order dated May 20, 2016 in Consumers Energy’s 2014 Power Supply Cost

22 Recovery Plan Case, the Commission stated at page 31:

23 In addition, the Commission expects that as part of its efforts to 24 maximize the functionality and energy savings benefits of AMI, 25 Consumers should evaluate the implementation of voltage control 26 facilitated by AMI. 27 Both of these decisions by the Commission were in primarily in response to testimony

28 that I presented in those cases.

29 Q. Please explain how distribution system investments can reduce line losses.

30 A. The laws of physics require that there will be system losses in Consumers Energy’s

31 distribution system. However, losses can be reduced by actions available to the

1 Company. Some of those actions are relatively inexpensive but all have costs. The most

2 cost-effective way to address this issue is to do so holistically, as part of a large-scale

3 distribution capital spending program if one is in process. Therefore, when asking the

4 Commission for approval to rate base a large amount of distribution capital spending,

5 Consumers Energy should be expected to demonstrate that it will exercise appropriate

6 diligence in ensuring that the combined costs of system losses and available mitigation

7 measures have been or are being minimized. It is highly likely that such measures will be

8 less expensive in context of other work on the distribution system and therefore that more

9 such measures will be cost-effective.

10 There are a number of examples of practices that will mitigate system losses that should

11 be examined:

12 (1) Using metering at intermediate points in the distribution system, such as feeder line

13 interfaces to substations, in combination with data from smart meters to identify and

14 resolve identifiable problems with excessive loss;

15 (2) As discussed further below, improving balance between phases in three-phase

16 circuits;

17 (3) As discussed further below, right-sizing transformers, based on site-specific AMI

18 data, and using more efficient transformers and other line equipment as normal

19 replacement is undertaken;

20 (4) As discussed further below, using dynamic volt-VAR control and circuit

21 configuration to optimize power flows, which may not be justifiable for system loss

22 reduction alone but is highly likely to be justified by the accumulation of multiple

23 benefits;

1 (5) As discussed further below, deploying conservation voltage reduction to reduce load

2 at high-load times, which will have the secondary benefit of disproportionately

3 reducing system losses;

4 (6) Remote monitoring and as-needed maintenance of line equipment, particularly

5 transformers, capacitors and voltage regulators, can both reduce system losses and

6 enhance system reliability;

7 (7) Selectively changing distribution system topology, upgrading line voltage or phasing,

8 or re-conductoring to address grid locations with high losses;

9 (8) Deploying efficiency programs, demand response programs, distributed generation or

10 storage targeted to reduce distribution load at high-load times or in locations with

11 high system loss rates.

12 Several of these practices are significantly enabled by Advanced Metering Infrastructure,

13 suggesting that the Commission should specifically require these measures now both

14 because they may be more economical in conjunction with other work on the distribution

15 system and because they are important steps for customers to benefit in their investment

16 in AMI.

17 Q. What is your understanding of the status of Consumers Energy’s deployment of

18 advanced metering infrastructure?

19 A. In Direct Testimony pre-filed in this case, Company witness Lincoln Warriner represents

20 that Consumers Energy will complete deployment of advanced meters during 2017.

1 Q. What are the most important ways to leverage AMI to improve the distribution

2 system through AMI?

3 A. AMI data can be used for phase balancing, transformer balancing, and control of voltage

4 and power factor. Each of these enhances distribution system reliability, reduces system

5 maintenance costs, and reduces generation requirements.

6 Q. Please explain the best practices and expected benefits of distribution system phase

7 balancing.

8 A. Primary distribution circuits are three-phase while most domestic and small commercial

9 loads are single phase. Some loads, especially thermal end uses, of primary customers are

10 single phase. As a result, the load on each phase of a three-phase distribution line can

11 become unbalanced. Waveform changes due to inductance and capacitance of end-uses

12 can also cause alternating current misalignments between the phases so that when

13 superposed there is a loss of real power and efficiency.

14 When phases are unbalanced, one or two phases carry excessive current and one or two

15 phases carry insufficient current. Those carrying excessive current experience higher line

16 and transformer losses while those carrying insufficient current experience lower line and

17 likely lower transformer losses. Because losses are proportional to the square of current,

18 the imbalance causes greater overall losses than if the phases were balanced. The phases

19 with higher load also experience greater voltage drop than the phases with lower load,

20 affecting delivered voltage; phase balancing can thereby enable greater voltage control,

21 which I discuss later. Three phase equipment also generally operates less efficiently and

22 with greater wear when the phases are unbalanced. It is therefore beneficial to balance

23 load between phases.

1 Basic load balancing can be accomplished by optimizing the assignment of load to fixed

2 wiring but additional performance is possible through distribution automation using

3 sectionalizing reclosers like those discussed for reliability sectionalization as discussed

4 by Company witness Bordine.

5 It has historically been difficult to identify phase imbalance and to determine best ways to

6 rebalance loads between phases. However, with AMI it becomes relatively easy. The

7 phase to which a single-phase load is attached can be determined by regressing voltage at

51 8 each meter to reference voltage on each phase to find the best fit.50F50F50F Various algorithms

9 are then available to optimally combine load profiles from advanced metering into

10 phases.

11 Direct benefits of phase balancing are generally fairly small and mostly are reductions in

12 losses. However, because the costs of phase balancing are also quite low, the cost-

13 effectiveness of phase balancing is generally high. For instance, EPRI’s Green Circuits

52 14 project51F51F51F included detailed analysis of 66 distribution circuits and included an evaluation

15 of phase rebalancing at the substation (additional benefits can be obtained by phase

16 balancing using additional equipment further along the circuit). The following graph

17 shows the results of that analysis

51 T. A. Short, Advanced Metering for Phase Identification, Transformer Identification, and Secondary Modeling. IEEE Transactions on Smart Grid 4:651-658, 2013. 52 EPRI. Green Circuit Distribution Efficiency Case Studies 2011 Technical Report, available from http://www.epri.com/abstracts/Pages/ProductAbstract.aspx?ProductId=000000000001023518&Mode=download.

2 This graph shows that for some circuits, rebalancing caused negative line loss reductions,

3 which can be attributed to the EPRI analysis using phase balancing at peak load rather

4 than a more advanced algorithm that takes account of the entire load profile. This

5 limitation of method was likely detrimental to the positive results for other circuits.

6 Nonetheless, 12 circuits showed definite beneficial results.

7 The EPRI project looked in more detail at 6 circuits, intended to represent the variety of

8 relevant conditions in the larger sample. The following table shows the benefit-cost ratio

9 of each of the strategies they analyzed for each of the circuits.

2 Phase balancing shows consistently good benefit-to-cost ratios across these example

3 circuits. The analysts also observed that phase balancing was most cost effective when

4 applied to circuits with high load density (load per length of circuit), which could be used

5 as a basis for prioritizing phase balancing analysis of Consumers Energy’s distribution

6 system. EPRI did not produce a table showing detailed phase balancing results across

7 these 6 circuits, but did show detailed results for two circuits. These were not chosen

8 based on the phase balancing results so may or may not be representative. They do

9 indicate that examination of phase balancing by Consumers Energy is warranted. I

10 prepared the following table of the best phase balancing solution for each circuit based on

11 their presentation, where each entry is a percentage of that circuit’s base load or demand.

Annual Annual Peak Peak Annual Total Levelized Circuit Consumption Losses Demand Losses Energy Peak Cost Saved Saved Saved Saved Saved Reduction $/kWh Saved A 3.1% 2.3% 0.1% 0.1% 3.1% 0.1% $0.009 F 2.4% 2.5% 0.1% 0.2% 2.4% 0.1% $0.004 12

1 Q. Please explain the best practices and expected benefits of transformer load

2 balancing.

3 A. Distribution transformers account for 26% of typical utility transmission and distribution

53 4 system losses and 41% of distribution and subtransmission losses.52F52F52F Transformers lose

5 energy through two primary mechanisms. Resistive losses in the windings of the

6 transformer are proportional to the square of current and electromagnetic field losses

7 related to energizing the transformer core are more or less proportional to voltage. As a

8 result, power losses in a transformer as a percent of load are dependent upon load as is

54 9 shown in the following graph53F53F53F where the various lines represent various load profiles

10 used in European standards:

53 T. A. Short, Electric Power Distribution Equipment and Systems. Boca Raton, FL, USA: CRC/Taylor & Francis, 2006. 54 From The Scope for Energy Savings in the EU Through the Use of Energy-Efficient Electricity Distribution Transformers by Hans De Keulenaer et al, available at http://www.cired.net/publications/cired2001/4_27.pdf.

1 This illustrates why common engineering practice is to size transformers for average load

2 around 50% of nominal capacity. In practice, many above-ground transformers are

3 overloaded for some hours of the year since they do not fail when rated capacity is

4 exceeded; rather they suffer greater losses (following the square of load) as heat which

5 accelerates transformer aging but with a reasonable economic tradeoff. It should also be

6 noted that when represented as actual loss rather than relative loss, losses at high loads

7 are much higher than losses at low loads. Thus transformer load balancing has potential

8 to reduce overall energy loss factors through reducing both overload and underload

9 conditions, but particularly to reduce energy and capacity requirements associated with

10 peak loads.

11 When transformers are consistently underloaded, this causes excessive losses as a

12 percentage of load. When transformers are consistenly overloaded, this causes excessive

13 losses as a percentage of load (and in absolute amounts) and accelerated aging of the

14 transformer. Aged transformers are most likely to fail during a period of high load and in

15 such circumstances are likely to spill cooling oil, requiring environmental remediation.

16 With the deployment of advanced meters, it is now possible to empirically examine

17 transformer loading. The basic strategy is to add up the hourly (or more granular) loads

18 served by a particular transformer and compare this to the transformer’s rating.

19 Transformer loading problems can then be addressed by reassigning load to a different

20 transformer, moving transformers around wihtin the distribution system to better match

21 transformer ratings to loads, or by taking steps to reduce high loads through targeted

22 efficiency, demand response, distributed generation, or storage. Such analyses can be

1 done using database queries and spreadsheets or by using commercially-available

2 software developed for this purpose.

3 As an illustration of the benefits of transformer load balancing that might be achieved by

4 Consumers Energy, I draw the Commission’s attention to the recently-filed testimony of

5 Karen R Lefkowitz on behalf of Potomac Electric Power Company (PEPCO) before the

6 Maryland Public Service Commission in Case No. 9418. PEPCO’s submittal letter,

7 application cover page, the first page of Lefkowitz’s testimony and the relevant pages 25

8 and 40 through 44 of her testimony are included in Exhibit SC-8. For context, PEPCo,

9 with approximately 550,000 customers, is approximately one-third the size of Consumers

10 Energy. In this testimony Lefkowitz describes that during 2014 and 2015, PEPCO

11 analyzed over 9,500 transformers serving over 74,000 customers located on its 2014 and

12 2015 Priority Feeders. This study identified 420 overloaded transformers that could

13 impact nearly 4,000 customers. The overloading determination was based on occurrence

14 of overloads of more than five hours in length that equaled or exceeded 160% of

15 transformer nameplate capacity for overhead transformers. For underground

16 transformers, the overloading determination was made based on overloads of more than

17 five hours in length and a loading magnitude of 100% of transformer nameplate capacity.

18 (The delineation in criteria between overhead and underground transformers is

19 attributable to the ability of overhead transformers to dissipate heat into the outdoor air.)

20 By methods outlined in pages 40 through 44 of her testimony, Lefkowitz estimates net

21 present value net benefits of about $13,480,000 from instituting this practice, dominated

22 by reliability benefits. This estimate did not include benefits from power quality

23 improvements or line loss reductions, as these were not quantified.

1 There is no a priori reason that Consumers Energy’s transformers are better or worse

2 matched to load than PEPCO’s. A calculation of potential energy and capacity savings

3 through transformer load balancing is necessarily speculative absent specific studies by

4 Consumers Energy, but can be approximated based on PEPCO’s experience. PEPCO

5 found that roughly 5% of its transformers were severely overloaded. A similar number

6 are likely significantly underloaded. Assuming that these transformers serve primary and

7 secondary distribution customers, about 80% of Consumers Energy’s load is potentially

8 affected. Based on a reasonable estimate that transformer load balancing reduces energy

9 losses by 0.3% of the load served by those transformers, we can approximate loss

10 reduction as 10%*80%*0.5% = 0.04% or about 15,000 MWh per year. Given that losses

11 increase with the square of load and that Consumers Energy’s peak load is approximately

12 double its average load, the effect on generation capacity requirements is likely about

13 0.16%, which with planning reserve margin added would be approximately 0.175% or

14 about 15 MW. The net present value of such savings would be about $27 million

15 depending upon assumptions about future costs of capacity and energy.

16 While the potential benefits described above are speculative, they justify an effort by

17 Consumers Energy to evaluate and as appropriate implement the practice of using AMI

18 data to balance load with transformer capacity.

19 Q. Please explain the best practices and benefits of dynamic voltage and power factor

20 control?

21 A. Total power delivery consists of real power, measured as volt-amps or watts, and reactive

22 power, measured as volt-amps reactive or VARs. Greatly oversimplifying, these

23 respectively measure the flow of power when voltage and current are in phase and can do

1 useful work and when they are out of phase. Power factor is the ratio of real power to

2 total power, so power factor and VARs are just different ways to measure the same

3 phenomena. The Commission’s regulations typically deal with power factor but the

4 technology for jointly managing real and reactive power is called dynamic volt-VAR

5 control and I will use that label.

6 Dynamic volt-VAR control uses sensors in the distribution system to monitor voltage and

7 VARs, processes those data through special-purpose software, then adjusts equipment in

8 the distribution system to bring voltage and VARs closer to their optimum levels in near

9 real time. This contrasts with static voltage and VAR control that relied on fixed settings

10 or settings controlled by clock. Dynamic volt-VAR control is enabled by Smart Grid

11 communications infrastructure. Importantly for the present case, advanced meters can

12 serve as the sensors for dynamic volt-VAR control, making the implementation of

13 dynamic volt-VAR control cheaper and easier than if implemented without advanced

14 metering.

15 It is common that implementation of dynamic volt-VAR control flattens voltage along a

16 distribution circuit. Voltage is typically controlled to some target at a substation, but as

17 power is used along a distribution circuit, voltage drops. Because much end-use

18 equipment uses more energy when fed at higher voltage, this causes customers close to

19 the substation to use more power than customers further along the circuit when making

20 use of similar equipment for similar end-use services of electric power. Thus, flattening

21 the voltage profile of a distribution circuit benefits customers near the substation with

22 lower bills (and sometimes less equipment wear) and customers near the end of the

1 distribution circuit with better service and less equipment wear through better power

2 quality.

3 When voltage is dynamically controlled, it is less variable, which enables greater use of a

4 practice called conservation voltage reduction (CVR). End-use and grid-attached

5 equipment is designed to work well within a range of voltages around the nominal

6 voltage of the distribution system interconnection. American National Standards Institute

7 recommends and this commission requires that voltage be regulated within 5% above or

8 below nominal voltage. Generally speaking, equipment that is exposed to recurrent low

9 or high voltage wears out more quickly.

10 CVR is the practice of reducing target voltage below nominal while ensuring that it does

11 not fall below 5% less than nominal voltage. The advantage of CVR is that most end-use

12 equipment uses less power when voltage is reduced but without a noticeable deterioration

13 in service quality. Incandescent lights will be slightly dimmed, water heating will take

14 slightly longer, etc. Some equipment compensates by drawing greater current so that the

15 flow of power is unchanged. The average ratio of reduced power usage to reduced

16 voltage is referred to as the Conservation Voltage Reduction Factor (CVRf). The value of

17 this factor varies by circuit and from time-to-time but average around 0.8, meaning that a

18 1% reduction in voltage will produce a 0.8% reduction in demand. Some uses, such as

19 water heating will still ultimately use just as much energy but at a slower rate, so it is

20 common for the energy savings from CVR to be around 0.65 of the percentage voltage

21 reduction.

1 A similar phenomenon occurs with reactive power, but with a stronger effect. A

2 percentage reduction in voltage will generally produce a 3% reduction in VARs; and

3 hence an improvement in power factor.

4 CVR is used by a number of utilities and is required in some jurisdictions. There are

5 generally two modes of application. Some utilities use CVR to reduce peak loads,

6 effectively treating as demand response. Progress Energy, before it was acquired by Duke

7 Energy, labeled its CVR program as Distribution System Demand Response. Others

8 apply CVR at all times, thereby accomplishing a measure of energy efficiency.

9 Because the tools of dynamic volt-VAR control and conservation voltage reduction are

10 nearly the same, it is common to implement them together under the label of distribution

11 system voltage optimization. As a long-standing practice, there have been many studies

12 of conservation voltage control and a significant number that incorporate dynamic

13 controls and look at voltage optimization. The EPRI Green Circuits study I referenced

14 above is one such study and produced representative results.

15 The following Figure from the Green Circuits report shows the annual energy savings

16 from voltage optimization on each of the 66 circuits included in the study.

2 As is shown in the figure above, it is common for voltage optimization to reduce annual

3 distribution load by between 2% and 2.5%. Although not shown in this figure, the

1 reduction in peak load is typically somewhat larger than the energy savings because of

2 the load-flattening effect of equipment like water heaters that run longer to accomplish

3 the same work when voltage is reduced. It is therefore reasonable to expect that a

4 comprehensive voltage optimization program at Consumers Energy would reduce

5 distribution system customer energy requirements by at least 2% and capacity

6 requirements by 2.5%. Since primary and secondary distribution customers take about

7 80% of Consumers Energy’s load and slightly more of its peak demand, this practice

8 should reduce Consumers Energy’s annual energy requirement by about 1.6% and its

9 peak load by about 2%.

10 The same benefit-to-cost ratio table for six circuits that EPRI studied more closely, that I

11 presented above (and here for easy reference), shows that both base voltage feedback and

12 VAR optimization perform well in many circuits.

14 I therefore conclude that there are large potential benefits if Consumers Energy

15 implements voltage optimization including both dynamic volt-VAR control and

16 conservation voltage reduction throughout its distribution system.

1 Q. What is your recommendation regarding the operational benefits of AMI?

2 A. I recommend that the Commission order, as a condition of its approval of rate relief

3 requested in this case, that Consumers Energy submit a comprehensive plan and report

4 covering these practices to the Commission by date certain and before or concurrent with

5 its next filing of a general rate case or a request for Certificate of Necessity for new

6 generation. I suggest that the date certain deadline for such report be 180 days after the

7 Commission enters its order in this case. I further note that once implemented these

8 practices should measurably reduce Consumers Energy’s peak demand forecasts, energy

9 sales forecast, and line loss factor and the Commission should expect future filings to

10 reflect those effects.

11 Q. Do you have a recommendation regarding the loss factor used for the Company’s

12 PSCR plans?

13 A. Yes. As I mentioned early in my testimony, Consumers Energy states that it projects a

14 reduction in losses as a result of the programs it is proposing. The programs I have

15 discussed, if adopted, would likely result in further reductions. However, as described in

16 discovery response 17990-MEC-CE-49, the Company does not project to reduce the loss

17 factor applied to its projected PSCR plan as a result of these anticipated loss reductions.

18 The Company’s position is that PSCR loss factors are set based on historic information,

19 not projections. The problem with that position, however, is that the PSCR loss factor is

20 only reviewed in general rate cases, whenever they happen to occur. Therefore, I

21 recommend that the Commission direct that if the loss factor in Consumers Energy’s

22 proposed PSCR tariff (Exhibit A-11, Schedule F-5, page 25) is accepted, that this number

1 reviewed and subject to re-opening in each PSCR plan case, based upon the Company’s

2 report and its tracking of results from these programs as I have described.

3 Q. Does that complete your testimony?

4 A. Yes.

STATE OF MICHIGAN

BEFORE THE MICHIGAN PUBLIC SERVICE COMMISSION

EXHIBITS

DOUGLAS JESTER

ON BEHALF OF

THE MICHIGAN ENVIRONMENTAL COUNCIL, THE NATURAL RESOURCES DEFENSE COUNCIL, THE SIERRA CLUB AND ENVIRONMENTAL LAW AND POLICY CENTER

July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-1; Source: D. Jester Resume Page 1 of 5

Douglas B. Jester

Personal Contact Information: Information 115 W Allegan Street, Suite 710 Lansing, MI 48933 517-337-7527 [email protected]

Professional January 2011 – present 5 Lakes Energy experience Principal Member

Co-owner of a consulting firm working to advance the clean energy economy in Michigan and beyond. Consulting engagements with foundations, startups, and large mature businesses have included work on public policy, business strategy, market development, technology collaboration, project finance, and export development concerning energy efficiency, smart grid, renewable generation, electric vehicle infrastructure, and utility regulation and rate design. Policy director for renewable energy ballot initiative. Supported startup of the Energy Innovation Business Council, a trade association of clean energy businesses.

February 2010 - December 2010 Michigan Department of Energy, Labor and Economic Growth Senior Energy Policy Advisor

Advisor to the Chief Energy Officer of the State of Michigan with primary focus on institutionalizing energy efficiency and renewable energy strategies and policies and developing clean energy businesses in Michigan. Provided several policy analyses concerning utility regulation, grid-integrated storage, performance contracting, feed-in tariffs, and low- income energy efficiency and assistance. Participated in Pluggable Electric Vehicle Task Force, Smart Grid Collaborative, Michigan Prosperity Initiative, and Green Partnership Team. Managed development of social-media-based community for energy practitioners. Organized conference on Biomass Waste to Energy.

August 2008 - February 2010 Rose International Business Development Consultant - Smart Grid . Employed by Verizon Business’ exclusive external staffing agency for the purpose of providing business and solution development consultation services to Verizon Business in the areas of Smart Grid services and transportation management services.

Douglas B Jester Page 1 of 5 6/29/2016 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-1; Source: D. Jester Resume Page 2 of 5

December 2007 - March 2010 Efficient Printers Inc President/Co-Owner . Co-founder and co-owner with Keith Carlson of a corporation formed for the purpose of acquiring J A Thomas Company, a sole proprietorship owned by Keith Carlson. Recognized as Sacramento County (California) 2008 Supplier of the Year and Washoe County (Nevada) Association for Retarded Citizens 2008 Employer of the Year. Business operations discontinued by asset sale to focus on associated printing software services of IT Services Corporation.

August 2007 - present IT Services Corporation President/Owner . Founder, co-owner, and President of a startup business intended to provide advanced IT consulting services and to acquire or develop managed services in selected niches, currently focused on developing e-commerce solutions for commercial printing with software-as-a- service currently in first customer use.

2004 – August 2007 Automated License Systems Chief Technology Officer . Member of four-person executive team and member of board of directors of a privately-held corporation specializing in automated systems for the sale of hunting and fishing licenses and in automated background check systems. Executive responsible for project management, network and data center operations, software and product development. Brought company through mezzanine financing and sold it to Active Networks.

2000 - 2004 WorldCom/MCI Director, Government Application Solutions . Executive responsible in various combinations for line of business sales, state and local government product marketing, project management, network and data center operations, software and product development, and contact center operations for specialized government process outsourcing business. Principal lines of business were vehicle emissions testing, firearm background checks, automated hunting and fishing license systems, automated appointment scheduling, and managed application hosting services. Also responsible for managing order entry, tracking, and service support systems for numerous large federal telecommunications contracts such as the US Post Office, Federal Aviation Administration, and Navy-Marine Corps Intranet. . Increased annual line-of-business revenue from $64 million to $93 million, improved EBITDA from approximately 2% to 27%, and retained all customers, in context of corporate scandal and bankruptcy. . Repeatedly evaluated in top 10% of company executive management on annual performance evaluations.

Douglas B Jester Page 2 of 5 6/29/2016 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-1; Source: D. Jester Resume Page 3 of 5

1999-2000 Compuware Corporation Senior Project Manager . Senior project manager, on customer site with five project managers and team of approximately 80, to migrate a major dental insurer from a mainframe environment to internet-enabled client-server environment.

1995 - 1999 City of East Lansing, Michigan Mayor and Councilmember . Elected chief executive of the City of East Lansing, a sophisticated city of 52,000 residents with a council-manager government employing about 350 staff and with an annual budget of about $47 million. Major accomplishments included incorporation of public asset depreciation into budgets with consequent improvements in public facilities and services, complete rewrite and modernization of city charter, greatly intensified cooperation between the City of East Lansing and the East Lansing Public Schools, significant increases in recreational facilities and services, major revisions to housing code, initiation of revision of the City Master Plan, facilitation of the merger of the Capital Area Transportation Authority and Michigan State University bus systems, initiation of a major downtown redevelopment project, City government efficiency improvements, and numerous other policy initiatives. Member of Michigan Municipal League policy committee on Transportation and Environment and principal writer of league policy on these subjects (still substantially unchanged as of 2009).

1995-1999 Michigan Department of Natural Resources Chief Information Officer . Executive responsibility for end-user computing, data center operations, wide area network, local area network, telephony, public safety radio, videoconferencing, application development and support, Y2K readiness for Departments of Natural Resources and Environmental Quality. Directed staff of about 110. Member of MERIT Affiliates Board and of the Great Lakes Commission’s Great Lakes Information Network (GLIN) Board.

1990-1995 Michigan Department of Natural Resources Senior Fisheries Manager . Responsible for coordinating management of Michigan’s Great Lakes fisheries worth about $4 billion per year including fish stocking and sport and commercial fishing regulation decisions, fishery monitoring and research programs, information systems development, market and economic analyses, litigation, legislative analysis and negotiation. University relations. Extensive involvement in regulation of steam electric and hydroelectric power plants. . Considerable involvement with Great Lakes Fishery Commission, including: o Co-chair of Strategic Great Lakes Fishery Management Plan working group o Member of Lake Erie and Lake St. Clair Committees o Chair, Council of Lake Committees o Member, Sea Lamprey Control Advisory Committee o St Clair and Detroit River Areas of Concern Planning Committees

Douglas B Jester Page 3 of 5 6/29/2016 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-1; Source: D. Jester Resume Page 4 of 5

1989-1990 American Fisheries Society Editor, North American Journal of Fisheries Management . Full responsibility for publication of one of the premier academic journals in natural resource management.

1984 - 1989 Michigan Department of Natural Resources Fisheries Administrator . Assistant to Chief of Fisheries, responsible for strategic planning, budgets, personnel management, public relations, market and economic analysis, and information systems. Department of Natural Resources representative to Governor’s Cabinet Council on Economic Development.

1983-present Michigan State University Adjunct Instructor . Irregular lecturer in various undergraduate and graduate fisheries and wildlife courses and informal graduate student research advisor in fisheries and wildlife and in parks and recreation marketing.

1977 – 1984 Michigan Department of Natural Resources Fisheries Research Biologist . Simulation modeling & policy analysis of Great Lakes ecosystems. Development of problem-oriented management records system and “epidemiological” approaches to managing inland fisheries.

Education 1991-1995 Michigan State University PhD Candidate, Environmental Economics Coursework completed, dissertation not pursued due to decision to pursue different career direction.

1980-1981 University of British Columbia Non-degree Program, Institute of Animal Resource Ecology

1974-1977 Virginia Polytechnic Institute & State University MS Fisheries and Wildlife Sciences MS Statistics and Operations Research

1971-1974 New Mexico State University BIS Mathematics, Biology, and Fine Arts

Douglas B Jester Page 4 of 5 6/29/2016 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-1; Source: D. Jester Resume Page 5 of 5

Citizenship and Youth Soccer Coach, East Lansing Soccer League, 1987-89 Community Co-organizer, East Lansing Community Unity, 1992-1993 Involvement Bailey Community Association Board, 1993-1995

East Lansing Commission on the Environment, 1993-1995

Councilmember, City of East Lansing, 1995-1999

Mayor, City of East Lansing, 1995-1997

East Lansing Downtown Development Authority Board Member, 1995- 1999

East Lansing Transportation Commission, 1999-2004

East Lansing Non-Profit Housing and Neighborhood Services Corporation Board Member, 2001-2004

Lansing Smart Zone Board of Directors, 2007-present

Council on Labor and Economic Growth, State of Michigan, by appointment of the Governor, May 2009 – May 2012

East Lansing Downtown Development Authority Board Member and Vice-Chair, 2010 – present.

East Lansing Brownfield Authority Board Member and Vice-Chair, 2010 – present.

East Lansing Downtown Management Board and Chair, 2010 – present.

Douglas B Jester Page 5 of 5 6/29/2016 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution NRDC Page 1 of 21 .JUNE2016 R: I6-05B

1GOUTPOLLUTION: How Utilities Can Aceelera te the Market fbr Electric Vehiôles..

MaxBaumhefner RolandHwang PierreBull

1- A July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 2 of 21

Acknowledgments Report written by: Max Bamnhefner Roland 1-Twang Pierre Bull

Reviews by: Sheryl Carter (NRDC) JJ McCoy (Northwest Energy Coalition) Pat Remick (NRDC) Luke Tonachel (NRDC) Seth Zuckerman (Climate Solutions)

Other substantive contributions by: Mary Heglar (NRDC) Allison Kotch (NRDC)

About NRDC The Natural Resources Defense Council is an international nonprofit environmental organization with more than 2.4 million members and online activists. Since 1970, our lawyers, scientists, and other environmental specialists have worked to protect the world’s natural resources, public health, and the environment. M{DC has offices in New York City; Washington, D.C., Los Angeles, San Francisco, Chicago, Montana, and Beijing. Visit us at urdc.org.

ARDC ChiefCommunications Officer: Lisa Benenson NRDCDepuIy Directors of Communications: Michelle Egan and Lisa Goffreth NRDCPo1icy Editor: Mary AnnaIse Heglar Design andProduction. www.suerossi.com

©Natural Resources Defense Council 2016 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 3 of 21 Table of Contents

Executive Summary 4

A. Introduction 6

B. The Need for Utffity Electric Vehicle Programs 7

1. Electric Vehicles and Clean Power Are Needed to Solve Global Warming and Air Pollution 7 2. Utility Investment in Charging Infrastructure Is Needed to Expand the Electric Vehicle Market 7 3. Widespread and Well-Managed Electric Vehicle Charging Can Benefit AU Utility Customers 9 4. Electric Vehicles Can Provide Valuable Grid Services 10

C. The Three Phases of Utifity Policy to Accelerate the Electric Vehicle Market 14

1. Removing Barriers to Adoption, Ensuring Grid Reliability, and Maximizing Fuel Cost Savings 14 2. Closing the Charging Infrastructure Gap and Promoting Equity 16 3. Capturing the Value of Grid Services and Integrating Renewable Energy 17

U. Conclusion 18

Endnotes 19

LIST OF FIGURES AND TABLES

Figure 1: SDG&E Cost of Service Before and After Widespread Electric Vehicle Adoption 9 Figure 2: Present Value of EV Adoption in California Through 2030, by Rate Scenario 9 Figure 3: Estimated Percentage of Time EVs Spend by Location 10 Figure 4: Modeled Storage Needs of Four Exemplary Subregions of the U.S. Western Grid with 27 Percent Renewables in 2022 12 Figure 5: Most Important Reason to Acquire an Electric Vehicle 15 Figure 6: Residential EV Charging in Dallas/Fort Worth Region by Tune of Day 15 Figure 7: Residential EV Charging in San Diego Region by Time of Day 15

Table ES-i: The Three Phases of Utility Electric Vehicle Market-Acceleration Policy 5 Table 1: Grid Services That Electric Vehicles Could Potentially Provide, by Grid Segment II Table 2: Estimated Value of Select Electric Vehicle Pilot Programs 13 Table 3: The Three Phases of Utility Policy to Accelerate the Electric Vehicle Market 14

PageS DRIVING OUT POLLUTION: HOW UTILITIES CAN ACCELERATETHE MARKET FOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 4 of 21 Executive Summary

Widespread adoption of electric vehicles (EVs) is an essential strategy for driving carbon pollution out of the transportation sector.1 Large-scale deployment of EVs can also help replace dirty power plants with clean energy like wind and solar power. And EVs powered by those renewable resources are virtually emissions-free.

Realizing this potential requires a robust network of As noted in a recent National Academies of Science study, charging stations where consumers live, work, and play. only utilities can capture the “incremental revenue from Such a network will pave the way for a larger and more additional electricity that EV drivers consume at home, diverse EV market. Electric utilities are uniquely positioned where roughly 80 percent of the charging takes place,” and to facilitate the creation of this network because they can use it to increase access to electricity as a &ansportation make use of spare grid capacity to charge EVs, generating fuel.7 This means serving the “garageless” who cannot significant new revenues. In turn, the growing customer buy a plug-in electric vehicle because they are not able to investment in EVs with large, advanced batteries can be plug it in at home, and growing the market in low-income leveraged to bring more renewable energy into the system. communities that are historically exposed to dangerous air With the right policies, the power and scale of the electric pollution and also the most vulnerable to volatile gas prices. industry could be unleashed to transform the way America It also means deploying charging stations at workplaces and travels while saving us money and protecting our health, other visible locations to drive new sales, alleviate “range environment, and economy from dangerous climate change. anxiety” (fear of running out of juice while driving), and make greater use of solar energy that generally peaks when Putting the transportation sector, which accounts for a people are away from home. combined with residential third of U.S. carbon pollution, on track to meet the nation’s charging, which ensures EVs are plugged in overnight when climate goals requires greatly accelerating EV sales.2 EVs wind power is abundant, this maximizes the availability of will need to account for 40 percent or more of new vehicle EVs to support the grid. sales by 2030, up from the current less than 1 percent, in order to meet long-term carbon-reduction targets, Researchers at the Pacific Northwest National Laboratory according to numerous studies.3 found sufficient spare capacity in the nation’s grid to power nearly all of our passenger cars and trucks, if vehicle This is not impossible. In a period of about two weeks, charging is properly managed.8 charging EVs during ahnost 400,000 people put down $1,000 deposits for the hours when the grid is underutilized increases utility next-generation, moderately priced Tesla.4 However, revenues without commensurate increases in costs, putting Tesla may be forced to return many of those deposits, if downward pressure on electricity rates. This effect is the the charging infrastructure network does not catch up to opposite of the utility “death spiral,” whereby increasing consumer demand. A major barrier to the growth of the EV costs borne by a decreasing pool of customers causes rate market is the lack of charging stations outside of single- increases that drive away more customers, leaving those family detached homes, where more than 80 percent of who cannot afford distributed (onsite) generation or home current EV owners live.5 energy storage to pay for an aging grid. In fact, a recent A substantial investment is needed. By way of example, study estimates large-scale commercialization of EVs in to meet california’s EV deployment goals, it is estimated california would generate net revenues of $2 billion to $8 125,000 to 220,000 publicly accessible charging ports are billion for Southern california Edison (SCE), San Diego Gas needed by 2020, a dramatic increase from the estimated & Electric (SDG&E), Pacific Gas & Electric (PG&E) and the 10,000 the state has today.° And hundreds of thousands of Sacramento Municipal Utility District (SIVEUD),enough to additional charging stations at apartment complexes and allow those utilities to both invest in charging infrastructure other multi-unit dwellings will be needed over the same time and reduce consumer bills.9 period to unleash the pent up demand from consumers who Electric utilities can also leverage the growing number do not live in single-family detached homes. of EV batteries already on the road to absorb increasing The electric industry is uniquely positioned to accelerate amounts of wind and solar electricity that may otherwise the EV market and help meet air quality standards and be dumped if it is not generated at times when there is climate goals by deploying more charging stations and sufficient demand. The charging of EVs can be managed to educating customers on the benefits of driving on electricity. avoid hours when the grid is strained and to coincide with

PaqeS DRIVINGOUT POLLUTION: HOW UTILITIES CANACCELERATETHE MARKET FOR ELECTRICVEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 5 of 21

TABLEES-I:THETHREEPHASESOFUTILITYELECTRICVEHICLE MARKET-ACCELERATIONPOLICY 1. Removing Barriers To Adoption, Ensuring Grid Reliability, And Maximizing fuel Cost Savings Clarify that electric vehicle charging companies will not be regulated as utilities Inform distribution system planning Provide consistent and fair treatment of electric vehicle load Adopt appropriate rates to maximize fuel savings and manage charging Target customer education and outreach programs

2. closing the Charging Infrastructure Gap and Promoting Equity I

Utility-facilitated deployment of charging infrastructure ‘

Increase access to electricity as transportation fuel in disadvantaged communities Promote broader awareness through mass-market education and outreach - 3. capturing the Value of Grid Services and Integrating Renewable Energy Implement traditional demand response programs for electric vehicle customers Implement advanced demand response programs for electric vehicle customers

Integrate V2G and battery second life programs into wholesale and retail markets

hours when renewable energy is plentiful, avoiding the need supports the grid and returns the value of doing so to EV to either spill valuable clean energy or invest in stand-alone drivers. Utility policies to accelerate the EV market can energy storage. be broken down into three phases, as shown in Table ES- In the future, EV batteries could even put electricity 1. Phase 1 removes barriers to EV purchases, facilitates back onto the grid when it is most needed. This can be a competitive market for third-party charging services, accomplished both via “vehicle-to-grid” or “V2G” (storing prepares utilities to integrate EV load, and encourages energy in EVs and putting it back onto the grid later) and via drivers to charge in a manner that avoids adverse grid “Battery Second Life” (storing energy in used EV batteries impacts and maximizes their fuel cost savings relative redeployed as stationary energy storage and putting it to gasoline. Phase 2 focuses on infrastructure, equity, back onto the grid later). American drivers have already and education programs to accelerate the EV market and purchased, in the form of EV batteries, more than enough increase access to electricity as a transportation fuel. Phase energy storage to power all the homes in the District of 3 develops managed charging programs so that EVs can Columbia on an average day.O That sunk investment grows facilitate the integration of renewable energy and provide with every EV purchase. Researchers at the National other grid services, and returns the value of such services to EV to Renewable Energy Laboratory (NREL) estimate massive drivers further accelerate the market. amounts of energy storage will likely be needed to balance Phase 1 presents the most pressing issues, but the a U.S. electric grid that is 80 percent renewable by the year foundations for Phase 2 and Phase 3 must be laid today in 2050.” That need could theoretically be met entirely with order to realize the long-term benefits of widespread EV batteries from as few as 10 percent of the EVs on the road adoption. Now is the time for utilities and utility regulators in that year.’2 Stand-alone energy storage on that scale to act. Short-term delays could result in a near-impossible could require an investment somewhere between $120 task in the future, as it takes decades to turn over the billion and $180 billion.’3 Directing even some portion of nation’s vehicle fleet. It is estimated that traffic pollution that investment away from capital-intensive, utility-scale causes more than 50,000 premature deaths annually projects and toward EV drivers to provide energy storage in the lower 48 states, which is more than 1.5 times the with the batteries they have already purchased could reduce deaths from traffic accidents on an annual basis.’4 The the cost of transitioning to a cleaner grid and accelerate the electric industry should move quickly to bring forward the electrification of the transportation sector, environmental and economic benefits of moving America off for To realize this potential, we need utility policies to unleash oil—once and good. greater investments in charging infrastructure and other programs that expand EV adoption in a manner that

Pag&. DRIVINGOUT POLLUTION: HOW UTILITIES CAN ACCELERATE THE MARKETFOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 6 of 21 A. Introduction

Achieving long-term greenhouse gas (GHG) reduction and compensating EV drivers for valuable grid services, targets in the United States and internationally requires deploying charging infrastructure that can act as a grid large-scale deployment of electric vehicles (EVs), including resource and reduce range anxiety, and conducting both pure battery electric vehicles and hybrids that can customer education and outreach. be plugged into the grid, increasingly fueled by renewable The electric industry is one of the few that have the power electricity. Electric utilities are singularly positioned to to challenge the market dominance of the oil industry. provide ubiquitous access to charging—while supporting Utility-scale investment is also needed to facilitate the grid and facilitating its transition to variable resources the expansion of the nascent competitive EV charging like wind and solar energy, benefiting all customers, and service industry. Since they provide the fuel, utilities and value to EV drivers. returning independent EV charging service companies play a critical Studies show that putting the transportation sector on role in determining the success of the EV market. With track to meet long-term GHG reduction gods requires the right programs to manage charging, widespread EV greatly accelerating the sale of EVs—currently less than 1 adoption could benefit the entire utility system and its percent of new sales—in the near term.’5 Fortunately, the customers. initial U.S. market launch of EVs has been highly successful. Electric vehicles are not the only form of fuel-switching that Over the first five years of sales, from 2011 to 2015, about can reduce overall emissions, but substituting electricity 388,000 EVs were sold; that is comparable to the number for petroleum fuels has the greatest potential to reduce of conventional hybrid electric vehicles (which cannot be emissions of any electrification opportunity. Likewise, plugged into the grid) sold during their first five years on the no other single customer-side resource combines the U.S. market.’° vehicles have also enjoyed Electric extremely attributes of a typical EV that could be utilized to support high levels of customer satisfaction: 98 percent of Tesla the grid. A 2016 Nissan LEAF can store as much electricity Model S owners, 85 percent of Chevrolet Volt owners, as the average American home uses in a day, equal the and 77 percent of Nissan LEAF owners report they would instantaneous demand of several homes, and be recharged definitely purchase the same vehicle again.’7 Over the next while its owner is sleeping, eating, working, or doing couple of years, the number of models will almost double, anything other than driving. and these next-generation models promise improved range and performance. Battery costs have been falling more Utility programs that maximize the storage, power, and rapidly than previously predicted and will continue to drop. flexibility of EVs can benefit all utility customers. Charging In a period of about two weeks, almost 400,000 people put EVs when there is spare grid capacity avoids the need for down $1,000 deposits for the next-generation, moderately new capital investments and provides additional revenue priced Tesla Model 318 without a commensurate increase in costs, thereby putting downward pressure on electricity rates. By providing However, without major new market-transformation valuable grid services that facilitate the integration policies, the EV market could stall before it reaches a of variable renewable resources like wind and solar, critical tipping To achieve mass commercialization, point. widespread EV adoption can also lower the costs of de EVs must overcome three key barriers: higher initial carbonizing the electricity sector. Achieving this promise purchase prices, concerns over lack of charging stations requires the rapid adoption of major new utility programs and range, and low consumer awareness. Utilities can and policies that can drive out pollution while benefiting the accelerate the mass commercialization of EVs by reducing power grid and its customers. the cost of ownership through appropriate rate design

Th,geG DRIVINGDUT POLLUTION: HOW UTILITIES CANACCELERATETHE MARKET FOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 7 of 21 B. The Need for Utility Electric Vehicle Programs

I.ELECTRICVEHICLESANDCLEANPOWERARENEEDED long-term emissions reduction goals. Electric utilities are TOSOLVEGLOBALWARMINGANDAIRPOLLUTION singularly positioned to close the charging infrastructure Numerous independent studies have come to the same gap. Utilities can work with third-party charging service conclusion: reducing global warming pollution to 80 providers to leverage existing customer relationships and percent below 1990 levels by 2050 will require a dramatic their knowledge of the electric grid to capture the value of shift to electric vehicles powered by renewable and other grid services provided by EVs and increased revenues from zero-carbon energy sources.’9 Because just 15 million to well-managed residential charging, which meets the vast 17 million passenger vehicles are sold each year in the majority of fueling needs.27 United States, it will take decades to transform the existing Recent studies have concluded that expanding charging U.S. stock of 250 million vehicles. To meet long-term GHG infrastructure is critical to increasing EV adoption. As emissions reduction targets, studies have estimated EVs will explained by the “network effect” of market diffusion, need to account for 40 percent or more of new vehicle sales consumer valuation of EVs increases with the number of by 2030.20 charging stations, but investors are less willing to build Electric vehicles are also increasingly needed to meet clean stations when the EV market is small (this is also known air standards in the most polluted areas of the country. It as the classic “chicken or the egg” problem). Researchers is estimated that traffic pollution causes more than 50,000 from Cornell University analyzed network effects associated premature deaths annually in the lower 48 states, which with quarterly EV sales in 353 metro areas. They found that is more than 1.5 times the deaths from traffic accidents on “the increased availability of public charging stations has an annual basis.2’ In California, regulators have concluded a statistically and economically significant impact on EV that broad deployment of zero- and near-zero-emissions adoption decisions.”2’ Another recent study of global EV technologies in the South Coast and San Joaquin Valley air markets concluded that of all the factors examined, charging basins will be needed between 2023 and 2032 to comply infrastructure was the best predictor of a country’s EV with current federal health-based air quality standards. The market share.29 regulators also project that by 2040, all passenger vehicles Recent investments by automakers further illustrate the sold in California will need to be zero-emissions vehicles.22 importance of infrastructure in driving increased sales. Major metro areas outside of California with dangerous BMW, Volkswagen, and Nissan have pledged to help finance levels of air pollution, such as Houston and Dallas, are more than 1,000 publicly available stations in key U.S. increasingly looking to EVs to comply with federal air markets. In Japan, Nissan, Honda, Toyota, and Mitsubishi quality standards.22 In light of the pressing need to combat have agreed to fund one-third of the cost of installing 12,000 dangerous air pollution, 13 North American and European public charging stations (with the balance provided by governments, including those of Germany, the United the government). According to Nissan’s market research, Kingdom, California, Connecticut, Maryland, and New York, sufficient charging infrastructure would double the number signed a pact to ensure that all new vehicles sold be zero- of Leading, Environmentally-friendly, Affordable, Family emissions vehicles by 2050.24 (LEAF) car owners who would repurchase an EV.3°Nissan also saw a marked increase in LEAF sales in 2013 when the company deployed a large number of direct current (DC) 2.UTILITYINVESTMENTINCHARGINGINFRASTRUCTURE fast charging stations at dealerships across North America IS NEEDEDTOEXPANDTHEELECTRICVEHICLEMARKET and Europe.2’ Similarly, Tesla officials report their DC fast It is clear that a model is becoming increasingly new charging network has been critical to growing sales of the needed to deliver the robust charging network necessary Model S sedan.32 to accelerate EV adoption. Market research shows that consumers’ top four reasons for rejecting EVs were all However, deploying charging infrastructure is not the core related to lack of infrastructure or range.2’ Survey analysis business of automakers. AFter all, automakers did not enter by the National Renewable Energy Laboratory (NREL) the gas station business to sell gasoline-powered vehicles. shows that the lack of infrastructure for alternative-fuel Likewise, while state and federal programs have supported vehicles is as much of a barrier to adoption as incremental much of the existing charging network, public funding vehicle price.2’ To date, the limited charging infrastructure alone will likely not be sufficient to meet the scale of the that exists beyond single-family homes has generally challenge. been deployed by government, automakers, and start Even California, which has been a strong supporter of up charging service companies. This model is unlilcely infrastructure deployment and has almost 30 percent of the to deliver the comprehensive network needed to meet nation’s publicly available charging ports, still falls far short

Paqe DRIVING OUT POLLUTION: HOW UTILITIES CANACCELERATETHE MARKETFOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 8 of 21 of what is needed to scale up the EV market.33 According Workplaces and Other Long-Dwell- to a study by NREL, to support 1 million EVs, California Time Locations would need 125,000 to 220,000 publicly accessible charging Adding charging stations to workplaces can both extend ports (including those deployed at workplaces), a dramatic range for drivers and increase EV visibility, which can increase from the estimated 10,000 the state has today.34 spur additional vehicle sales. Nissan credits a workplace To meet emissions reduction goals, we need to rapidly charging initiative with a fivefold increase in monthiy increase our investment in infrastructure. Unfortunately, EV purchases by employees at Cisco Systems, Coca private financing of the installation and operation of Cola, Google, Microsoft, and Oracle.4’ Likewise, the U.S. charging stations alone does not appear to be sufficient. Department of Energy (DOE) recently concluded that A recent study commissioned for the state of Washington employees of companies participating in its Workplace found that “charging station business models that rely solely Charging Challenge were 20 times more likely to drive an on direct revenue from EV charging services currently are EV than the average worker.42 Workplace charging can also not financially feasible” and that viable business models increase electric miles driven, especially for drivers of plug- must “capture other types of business value in addition to in hybrid vehicles with shorter all-electric ranges, reducing selling electricity.”35 The challenge is especially acute for their reliance on petroleum. Utility-facilitated deployments DC fast charging stations, which have high capital costs of grid-integrated charging infrastructure at workplaces and can be subject to demand charges meant to recover and other long-dwell-time locations such as park-and- investments needed to meet peak electricity demand. ride commuter lots also ensure EVs are available in the Utilities are uniquely situated to capture the system-wide afternoon to serve as a form of energy storage to absorb benefits of a comprehensive charging network. As noted in peak production from solar energy (see Section 4). a recent National Academies of Science study, utilities can capture the “incremental revenue from additional electricity Public Fast Charging that EV drivers consume at home, where roughly 80 percent Nine days out often that a car is driven, it is driven less of the charging takes place” and use that revenue to both than 70 miles, which is well within the range of today’s deploy charging stations and reduce rates and bills for all pure battery electric vehicles, but the lack of fast charging customers.3° infrastructure needed to make that one-in-ten trip remains Increasing access to electricity as a transportation fuel is a significant obstacle to the purchase of pure battery EVs.’13 a natural fit for the electric industry. A multi-state survey Drivers’ purchase decisions are often disproportionately conducted by researchers at the University of California, influenced by rare use cases; for example, the off-road Davis reveals EV drivers believe utilities should lead the capability of SINs remains a driving force behind their deployment of charging infrastructure.37 Building upon market dominance, even though that capability is almost their history of helping to transform the market for energy never used. Consumer research shows the lack of “robust efficiency and renewable energy, utilities are also well DC fast charging infrastructure is seriously inhibiting the situated to deploy infrastructure, especially in segments value, utility, and sales potential” of typical pure-battery where the need is greatest, such as: electric vehicles.44 Unfortunately, without extremely high utilization rates, it is difficult for private firms to recoup Apartment Complexes and Other installation costs and cover operating expenses, including Multiunit Dwellings utility demand charges that are meant to recover grid investments needed to serve customers with high power Drivers are unlikely to purchase plug-in vehicles if they requirements.’15 As the keepers of the electric grid, utilities cannot plug in at home, where cars are parked for 12 hours are singularly situated to facilitate the deployment of fast out of every day and the vast majority of driving needs can charging stations that incorporate strategies to minimize be met with overnight charging.38 Unfortunately, less than the need for additional grid investments, including managing haff of U.S. vehicles have reliable access to a dedicated off- the demand of both charging stations and other loads, street parking space at an owned residence where charging as well as on- and off-site energy storage. Utility-funded infrastructure could be installed.39 More than 80 percent researchers are also in the process of developing more of EV drivers live in single-family detached homes.4° It is efficient utility fast charging stations that require less power essential for the EV market to move beyond the suburbs to deliver the same amount of electricity.46 Utilities are to meet long-term climate and air quality goals. Installing also uniquely able to fund the deployment of fast charging charging stations at apartment buildings and other multiunit stations needed for widespread EV adoption with additional dwellings could unlock the potential for a broader, younger, revenues derived from the residential charging that will and more diverse market for the next generation of EVs. occur as a result of greater adoption and use of EVs. Utilities can leverage existing customer relationships, knowledge of the electric grid, and economies of scale to deploy charging stations in this critical but underserved market.

DRIVING OUT POLLUTION: HOW UTILITIES CAN ACCELERATE THE MARKET FOR ELECTRIC VEHICLES NROC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 9 of 21 WIDESPREADANDWELL-MANAGED 3. FIGUREI:SDG&ECOSTOFSERVICEDEFOREAND ELECTRICVEHICLECHARGINGCANBENEFIT AFTERWIDESPREADELECTRIC VEHICLEADOPTION ALLUTILITYCUSTOMERS Charging electric vehicles predominantly during off-peak $200 electricity hours (when the electric grid is underutilized and there is plenty of spare capacity in the generation, = transmission, and distribution system) allows utilities to . 150 avoid new capital investments while capturing additional S Generation revenues, lowering the average electricity cost for all their 00 — • Transmission customers. This effect is the opposite of the utility “death $60.58 S Distribution spiral,” whereby increasing costs borne by a decreasing pool of customers causes rate increases that drive away more II50 : $33.16 customers, leaving those who cannot afford distributed generation or home energy storage to pay for an aging grid. BeforeWidespread AfterWidespread Researchers at the Pacific Northwest National Laboratory EVAdoption EVAdoption found sufficient spare capacity in the nation’s electric grid (Adapted from Kintner-Myer et al, 2OO7)’ to power virtually the entire light-duty passenger vehicle fleet, if vehicle charging load is integrated during off-peak in the cost of service) is significant. All customers would hours and at lower power levels.47 As the grid becomes benefit from this shift in the form of lower electricity bills. more dominated by renewable energy generation that Using standard regulatory cost benefit tests, recent analysis varies depending upon the weather, time of day, and season, conducted by Energy and Environmental Economics the amount of spare capacity may grow even larger. The (E3) demonstrates that the body of utility customers is likely to same researchers also modeled impacts on the marginal benefit from the additional revenue provided by properly cost of utility service associated with transformative managed EV charging. Figure 2 presents results from the transportation electrification for two utilities, Cincinnati Ratepayer Impact Measure test, a restrictive test that fails Gas & Electric and San Diego Gas & Electric (SDG&E). The to account for systemic benefits, for a typical California results of a 60 percent plug-in hybrid penetration scenario utility under typical rate structures. It reveals that, by 2030, in SDG&E territory are illustrated in Figure 1. EVs will contribute $2 billion to $8 billion more in revenue These results do not reflect all the complexities of SDG&E’s to SCE, SDG&E, PG&E and SMUD than they cost to serve, systems, but the directional shift (-‘20 percent reduction putting downward pressure on rates for all customers.

FIGURE2:PRESENTVALUEOFEVADOPTIONINCALIFORNIATHROUGH2030,BYRATESCENARIO

$16 II

14 Notes: Based on California utility system, Cd, D NetRevenue Cz 12 assuming charging occurs predominantly -J-I when the system is underutilized, Net I0 RateBase revenues are positive under “Tiered,” z “Flat,” and “TOU” (time-of-use) rate ‘I, 8 RRPS Cost structures and a “Mixed” TOU/Tiered U scenario. Under TOU rates, EV owners are rewarded for charging during hours -J 6 • CarbonCost of the day when the cost of energy is at its lowest, resulting in smaller, but still 4 a CapacityCost U significant, net revenues. 0U 2 • EnergyCost Utility $0 a Bills U I- U I.-. CLI I- U I Z CCd, Z CCd) CU) Z CU) U C.) U C U C.) CLI C.) U> U>

TIERED FLAT MIXED TOU

(Environmental and Energy Economics, California Transportation Etectrification Assessment—Phase 2: GridImpacts)

Paqet) DRIVING OUT POLLUTION: HOW UTILITIES CAN ACCELERATE THE MARKETFOR ELECTRIC VEHICLES NROC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 10 of 21 To capture the potential of widespread EV adoption to benefit all customers, utilities should implement rate FIGURE3:ESTIMATEDPERCENTAGEOFTIMEEVsSPENDBYLOCATION designs and programs to ensure EV charging occurs predominantly when there is excess capacity in the grid. Driving SDG&E has already demonstrated that the combination of 4% time-of-use rates and education and outreach can push 80 Parked percent of EV charging to the hours between midnight and elsewhere 5 a.m., the “super off-peak” period on the utility’s EV tariff (see Figure 7)50 Such time-of-use rates are likely sufficient to integrate EV load in the early market. However, analysis conducted by SMUD shows that more sophisticated forms of load management, such as the use of dynamic price signals or advanced demand response, will likely be needed to minimize costs and allow for net benefits as the EV market scales up.5’

4. ELECTRICVEHICLESCANPROVIDEVALUABLE GRIDSERVICES Already highly valued by grid operators and utilities, flexible resources that keep the grid stable by ensuring (Adapted from Langtoo a criaotomo, Vehicle-Grid Integration, California Public electricity demand and supply remain perfectly in sync Utilities Commission)” will become increasingly valuable as variable resources like wind and solar replace fossil and nuclear generation. With the right policy framework, utilities can leverage charging that can easily be accomplished during off-peak for the electricity grid. Using “level 2” charging the growing customer sunk investment in EV batteries to hours equipment, which plugs into a 240-volt outlet those capture this value and use it to drive additional EV sales. (like used by clothes dryers), 35 miles of electricity can be Rewarding EV customers for facilitating the transition to delivered in two hours. This provides renewable energy could prove a sustainable replacement an immense amount of for federal and state purchase incentives that are likely to flexibility, considering the typical EV is parked for 23 hours a day. That flexibility means an EV battery could satisfy be phased out as the EV market moves beyond the early- adopter segment. typical driving needs while supporting the electric grid and providing EV drivers with significant value. Electric Vehicles Represent a Unique Opportunity to Support the Grid The Types of Grid Services Electric Vehicles Could Provide American drivers have already purchased approximately Imagine a vehicle that stops charging when demand for 11 gigawatt-hours (GWh) of advanced battery storage in the electricity peaks in the early evening and begins again form of EV batteries, more than enough to power all the late at night when most people are asleep and electricity homes in the District of Columbia on an average day.5’ This is cheap. Now picture EV being driven to work in sunk investment grows with every new EV purchase and that the morning, charging up on excess solar generation during the represents a unique opportunity to support the electric grid. being driven home, selling back the There is no other single customer-side “smart appliance” afternoon, electricity to grid that combines the potential for immense flexibility with when demand peaks in the evening, and then recharging again midnight significant capacity for both power and storage. at when there is an oversupply of cheap wind energy. Imagine further that after many years of service, Peak demand for electricity generally occurs during when the battery in that EV has lost enough capacity that the early-evening hours when people return home from it no longer provides the range its driver requires, it is work, turn on the lights, crank up air conditioners, redeployed as a form of stationary energy storage that watch television, and do all the other things that require could be charged and discharged whenever or wherever electricity—most of which can be done only when people most needed to support the grid. All of these functions are at home and awake. In contrast, EVs can be charged are already being proved in the real world. They can be whenever they are not being driven, which is 96 percent of categorized as follows: the average day, as shown in Figure 3, provided they have access to charging stations. 1. Traditional Demand Response: Turning charging off. 2. Advanced Demand Response: Turning charging on or The average American drives 35 miles per day.54Using a off and/or changing the rate of charging. standard 120-volt wall outlet and the “level 1” charging cords that are provided with every EV, 35 miles’ worth 3. Vehicle-to-Grid, or V2G: Putting electricity stored in of electricity can be delivered in nine hours of low-power EVs back onto the grid.

DRIVING OUT POLLUTION: HOW UTILITIES CAN ACCELERATE THE MARKET FOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 11 of 21 4. Battery Second Life: Putting electricity stored in used EV year; a single typical peaker natural gas plant costs about batteries redeployed in stationary applications back onto $120 million to build at $1,200 per kW of capacity.55 thegrid. In some regions, peaker plants are supplemented by These four functions can potentially provide the full range “Pumped Hydro Storage” facilities that pump water uphill of services required to keep the grid stable at all levels. and use gravity to power turbines at a later time. This is Supply of electricity must instantaneously and precisely by far the most widely deployed form of energy storage match demand to prevent blackouts. Yet both demand currently operating, but it is expensive and it has a large and supply of electricity change by the second, minute, geographic footprint.56 hour, day, and season. Grid operators must maintain this Distribution System Grid Services equilibrium, even as they integrate greater levels of variable Local utilities often rely on customer-side resources renewable resources, like wind and solar. to defer upgrades to equipment such as neighborhood This holds true across the entire electric grid, which transformers. To date, this has meant rewarding comprises both a transmission and a distribution system participating utility customers for occasionally turning operated by different entities. The transmission system things off (“Traditional Demand Response”) or firing moves electricity from power plants in bulk, often up diesel generators during peak-demand hours. In the across state lines. It is kept in balance by independent future, “Advanced Demand Response” programs could system operators (ISOs), such as the California ISO, and reward customers for allowing things to be automatically regional transmission organizations (RTOs), such as turned off, on, down, and up in a manner that still meets PJM Interconnection, which operate wholesale energy their needs. Customer-side energy storage resources markets. The distribution system delivers electricity from could also provide “Power Quality” (keeping local the transmission system to retail customers in homes voltage, frequency, and power stable to protect critical and businesses. Its reliability is ensured by local utilities. equipment), “Energy Arbitrage” (storing electricity Utilities, ISOs, and RTOs rely on the transmission and during hours when it is cheap, and either using it later distribution system grid services to keep the whole system or selling it back to the grid when it is expensive), in balance. and “Demand Charge Mitigation” (managing on-site consumption to minimize “demand charges” on utility Transmission System Grid Services bills, which recover investments needed to accommodate “Day-Ahead Resources,” typically bid into wholesale peak demand). markets one day before deployment, must be able to turn on within minutes, reach full power within 30 As shown in Table 1, the four categories of EV functions minutes, and be maintained for about four hours. These can provide the full spectrum of grid services at both the transmission and distribution system levels. resources are typically used only 5 to 20 days per year, often to meet peak demand from air conditioners during Utility-scale projects take years to finance, permit, and heat waves. “Ancillary Services” (including “Frequency construct and can be difficult to site where they are Regulation” and “Spinning Reserves”) meet the most needed. In contrast, EVs can be scaled precisely instantaneous needs of the grid, requiring an immediate and deployed strategically to provide the full spectrum response for up to 30 minutes. These are called upon of grid services required at the transmission level by several times per day for around 50 days a year. For more ISOs and RTOs operating wholesale markets, and at the than a century, procuring “Day-Ahead Resources” and distribution level by local utilities that interact directly “Ancillary Services” meant building expensive fossil-fuel with retail customers. The challenges associated with power plants, called “peakers,” that sat idle most of the relying on vehicles, the primary purpose of which is to

TABLEI:GRIDSERVICESTHAT ELECTRICVEHICLESCOULDPOTENTIALLYPROVIDE,BYGRIDSEGMENT Potential Grid Services, by Grid Segment Electric Vehicle Function Transmission Distribution Traditional Demand Response: Day-ahead resource, spinning reserve Grid upgrade deferral, demand charge Powering charging down or off mitigation Advanced Demand Response: Day-ahead resource, spinning reserve, Grid upgrade deferral, demand charge Powering charging down, off, on, or up frequency regulation, one-way energy mitigation, energy arbitrage storage Vehicle-to-Grid (“V2G”): Day-ahead resource, spinning reserve, Grid upgrade deferral, power quality, Discharging energy stored in EVs frequency regulation, two-way energy demand charge mitigation, energy arbitrage back to the grid storage.. :wv Battery Second Life: Day-ahead resource, spinning reserve, Grid upgrade deferral, power quality, Deployingused EVbatteries as stationary frequency regulation, two-way energy demand charge mitigation, energy arbitrage energy storage storage

ii1i DRIVINGOUT POLLUTION: HOW UTILITIES CANACCELERATETHE MARKETFOR ELECTRIC VEHICLES t NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 12 of 21 provide mobility, to also provide grid seivices can be largely FIGURE4:MODELEDSTORAGENEEDSOFFOUR EXEMPLARYSUBREGIONS mitigated with sufficient scale; with a large enough fleet, OFTHEU.S.WESTERNGRIDWITH27PERCENTRENEWABLESIN2022 the fact that some vehicles are not available at any given moment becomes increasingly irrelevant because enough other vehicles will be available at that time. SOLAR-DOMINATEDREGIONS CahfoiaSrmerWeaLead Electric Vehicles Can Provide a Variety of Grid Support and Storage Services to Meet Unique Regional System Needs a Integrating higher levels of renewable energy, chiefly wind and solar, will increase the demand for flexible grid resources that could be provided by EV batteries across ass the nation, but resources and needs will vary regionally. 111* As shown in Figure 4, in solar-dominant California and the desert Southwest, solar generation can create an oversupply of electricity during the afternoon but does little to help 5q1W* meet peak demand during evening hours. Both Traditional Ot*dSflsnWnk laid Demand Response and Advanced Demand Response can a help avoid exacerbating that evening peak, and the latter can also help absorb excess solar generation during the a afternoon by ramping up EV charging. V2G and Battery Second Life (see Table 1) provide the additional benefit of Inns selling excess solar energy stored during the afternoon back to the grid to supply peak demand during the evening a (Energy Arbitrage). This regular pattern of excess solar generation during the afternoon will be most common in the spring and fall, when the biggest source of demand for — ss_ electricity resources—heating or cooling of buildings—is WIND-DOMINATED REGIONS at its lowest. Solar could provide as much as one-quarter of the regional electricity supply at these times.57 Wind energy generation often peaks during early-morning a hours, when demand for electricity is typically at its lowest point because the vast majority of the population is asleep.58 EVs can be conveniently refueled while their drivers are still in bed and electricity is cheap. In 2009, BMW and the European utility Vattenfall demonstrated the potential for EVs to function as a form of Advanced Demand Response in which overnight charging was ramped up and down to — tib’I match variable wind generation, integrating renewable —sat generation while effectively lowering the emissions of the PadficWWS*iqWnki.aad vehicles.59 However, the wind does not always blow at night, nor does it blow every day. Such varying intervals typical a of onshore wind generation as well as seasonal variation in hydropower resources are shown in Figure 4 in the Rocky Mountain (top right) and Pacific Northwest (bottom right) in regions. A similar pattern is likely to emerge for the wind- rich Midwest, South, and Northeast (counting both onshore eas and offshore generation), which also contain a sizable share ass of hydropower resources. In such regions, energy must often be stored for longer periods of time, requiring EV grid l51t; ma!zaq 5195351 services such as Advanced Demand Response and V2G to (Energy and Environmental Economics, Inc., Investigating a ThgherRenewables be supplemented with stationary forms of energy storage, Portfolio Standard in California)” including Battery Second Life.

Note: Top lins is ths nst gsnsration for a 27 psrcsnt renswables scenario over a one- month period. The bottom line is the net load, The green-shaded area represents the oversupply of electricity that will be wasted if it cannot be stored and used later to power Evs or pot back onto the grid when demand for electricity peaks. That need for energy storage could be met with Advanced Demand aesponse, v2G, or Battery 5econd Life programs.

PnqeJ2I DRIVINGOUT POLLUTION: HOW UTILITIES CANACCELERATE THE MARKETFOR ELECTRICVEHICLES NROC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 13 of 21 The Potential Value of Electric energy, the project is able to meet the PJM minimum Vehicle Grid Services 100-kW capacity required to participate in the ancillary services market.63 A similar V2G at an Air The full range of EV grid support functions is already pilot Force base in being demonstrated in projects that are providing Southern California, with 32 EVs that provide up to 655 kW of power to California ISO’s considerable value, especially in highly remunerative but ancillary services market, is annual revenues vehicle.64 limited ancillary services markets. However, leadership by returning of around $2,500 per Table 2 shows of value regulators and the utilities under their jurisdiction is needed per-vehicle estimates for pilot projects currently underway. for EVs to capture the full value of facilitating the transition to a grid dominated by variable renewable resources. Questions remain as to the willingness of automakers to allow their vehicles’ batteries to be used for V2G. Likewise, PG&E and BMW have partnered in a novel pilot project that the scalability of V2G combines Traditional Demand Response and Battery Second remains to be seen. While today’s Frequency Regulation markets are highly remunerative, Life to provide a flexible grid resource with a capacity of they could be saturated with as few as 136,000 EVs in the at least 100 kW. They are offering a $1,000 incentive to PJM market or 45,000 EVs in the California encourage 100 EV drivers to participate in an 18-month ISO region if one assumes V2G-enabled vehicles pilot study. Participants could earn an additional $540 by are able to return to grid responding to day-ahead requests to curtail charging during electricity the at the same rate at which they charge using typical level 2 equipment plugged hours when the grid is pushed to its limits. The $540 figure into 240- volt outlets.69 The for Frequency Regulation was derived using the tool approved by the California Public market could double or triple as more variable solar Utilities Commission to determine the value of investments and wind energy is. integrated into mix. deferred as the result of demand response programs. If the generation Over the longer term, however, one-way the response rate is not high enough to reduce demand by storage is likely to emerge as the greatest opportunity for drivers to earn value since it allows the full 100 kW, BMW will return energy to the grid from utilities to call upon EVs as a “dispatchable” resource used EV batteries redeployed in a stationary second-life to absorb low-cost wind and application to make up the balance.6’ In addition, the pilot solar, balance the grid, and improve the utilization of the system.7° is meant to build a technical foundation and customer interface that could be used in future Advanced Demand Researchers at NREL have estimated that 100 to 152 Response and V2G programs. BMW is also using this real- GW of energy storage will likely be needed to balance a world test to determine whether sufficient value can be U.S. electric grid that is based on 80 percent renewable derived from stationary storage to justify pre-engineering resources by the year 2050.’ That need could theoretically battery packs to be easily redeployed in Battery Second Life be met entirely with EV batteries from as few as 10 applications. percent of the EVs on the road in that year.72 Stand- , In the PJM Interconnection regional wholesale market, alone energy storage on that scale could require an investment somewhere between $120 $180 which serves approximately 50 million people in the mid- billion and billion.73 Directing Atlantic and Midwest states, the University of Delaware and even some portion of that investment away from capital-intensive, utility-scale projects and NRG Energy have a demonstration V2G project underway, toward EV drivers to provide energy in which a fleet of EVs is charged at optimal times and storage with the returns power to the grid to provide Frequency Regulation, batteries they have already purchased could reduce the cost of transitioning to a cleaner grid and accelerate the earning annual revenues of about $1,800 per vehicle.62 By electrification of the transportation managing the charging and discharging of the vehicles’ sector.

TABLE2:ESTIMATEDVALUEOFSELECTELECTRICVEHICLEPILOTPROGRAMS Project Electric Vehicle Function Market Grid Services Estimated $/ Vehicle

BMW/PG&EPilot Traditional Demand Response California Day Ahead Resource, Spinning Reserve $360 per year65 ISO Hypothetical at 40% Advanced Demand Response Retail One-way Storage (storing renewable $850 over vehicle Renewable Penetration energy and using to drive later) lifetime66 Univ. of Delaware & V2G PJM Frequency Regulation $1,800 per year°7 NRG Demonstration U.S.Dept. ofDefense V2G California Frequency Regulation $2,520 per year68 ISO

— Paqëi? DRIVINGOUT POLLUTION: HOW UTILITIES CANACCELERATETHE MARKET FOR ELECTRIC VEHICLES NRDC _____

July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 14 of 21 C. The Three Phases of Utility Policy to Accelerate the Electric Vehicle Market

Realizing the environmental, consumer, and grid benefits of terms of service, rates, and other policies adopted by state EV adoption requires regulators to move quickly to develop commissions. Sixteen states have adopted policies to make new programs and policies to accelerate the market. Based it clear that such companies are not subject to the full on experience in the major early EV markets, the necessary extent of utility regulatory authority. Policymakers should utility policies can be separated into three phases, make it clear that companies acting as customers of utilities introduced in Table ES-i, repeated below, and discussed in will not be regulated like public utilities, but they should detail in the subsections that follow. also avoid creating sweeping exemptions that could hinder future efforts to ensure the environmental performance and integrity of the electric grid. I.REMOVINGBARRIERSTOADOPTION,ENSURINGGRID RELIABILITY,ANDMAXIMIZINGFUELCOSTSAVINGS Inform Distribution System Planning Phase i removes barriers to consumer adoption, facilitates a competitive market for third-party charging services, A fundamental tool to minimize the costs of integrating prepares utilities to integrate EV load, and encourages vehicle charging is timely utility notification when a drivers to charge in a manner that avoids adverse grid customer buys an EV. In California, one of the world’s impacts and maximizes savings relative to gasoline. largest EV markets with more than 200,000 vehicles, Example policies include: costs associated with integrated EV load so far have been de minimis—only 0.i percent of EVs have required Clarify that Electric Vehicle Charging Companies a service line and/or distribution system upgrade.76 A Will Not Be Regulated as Utilities detailed analysis of California’s distribution systems also reveals that, with the right policies, a mass market for EVs Regulatory treatment of independent EV charging could be achieved without significant new investments.77 companies is a fundamental issue that must be decided at Flowever, the instantaneous demand of a single EV can be the state level. State codes often define the term electric comparable to that of an entire home, which could result in utility very broadly, potentially subjecting EV charging local distribution system impacts if not properly managed.78 service providers to the jurisdiction of state utility For example, the cost of replacing a transformer on an regulators. In most instances, such companies will simply emergency basis can be twice that of a planned upgrade.79 act as customers of utilities and will be subject to the Therefore, regulators and utilities need to know where EVs

TARLE3:THE THREEPHASESOFUTILITYPOLICYTO ACCELERATETHE ELECTRICVEHICLEMARKET

1. Removing Barriers to Adoption, Ensuring Grid Reliability, and Maximizing Fuel Cost Savings Clarify that electric vehicle charging companies will not be regulated as utilities Inform distribution system planning Provide consistent and fair treatment of electric vehicle load Adopt appropriate rates to maximize fuel savings and manage charging Target customer education and outreach programs 2. closing the Charging Infrastructure Gap and Promoting Equity Utility-facilitated deployment of charging infrastructure Increase access to electricity as transportation fuel in disadvantaged communities Promote broader awareness through mass-market education and outreach 3 Capturing th Value of Grid Services and Integrating Renewable Fnergy Implement traditional demand response programs for electric vehicle customers Implement advanced demand response programs for electric vehicle customers

Integrate V2G and battery second life programs into wholesale and retail markets

DRIVING POLLUTION: HOW UTILITIES Pafl j OUT CAN ACCELERATETHE MARKET FOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 15 of 21 are charging if they are to manage the load. Four states 5: IMPORTANT AN FIGUREMOST REASONTOACQUIREELECTRIC VEHICLEhave adopted some form of notification requirements.8° (FROMSURVEYOF16,000EVOWNERS) Notification is also essential to facilitate targeted customer outreach regarding EV rate options, policies, and programs. Existing utility rules generally require customers to provide ADesirefor VehiclePerformance5% notification Newest whenever they add significant new load, but Supportingthe customers Technology are often unaware of this requirement and Diffusionof EVTechnology5% contact their utility only if something goes wrong.8’ 5% Utilities L Other3% must proactively identify EV owners. Potential sources of Increased L actionable information include automakers, auto dealers, Energy charging equipment installers, local building permit offices, Independence smart meter data, and state departments of motor vehicles (DIVWs). 56% Legislative changes are sometimes necessary to allow access to DMV data, which is the most comprehensive Saving - source.8’ —0 Moneyon Carpool FuelCosts Provide Consistent and Fair Treatment Lane 37% of Electric Vehicle Load Utility regulators should resist calls to implement EV specific charges or fees. Existing utility rules are generally ReducingEnvironmentalImpacts22% sufficient to recover costs associated with integrating vehicle load. There is no reason to treat EV load less (center for Sustainable Energy, California Plug-in Electric Vehicle Owner Survey favorably than comparably demanding loads such as hot Dashboard)” tubs and air conditioners, which lack the corresponding environmental benefits.

FIGURES:RESIDENTIALEVCHARGINGINDALLAS/FORTWORTH Adopt Appropriate Rates to Maximize REGIONBYTIMEOFDAY Fuel Savings and Manage Charging

Line: Generally applicable utility rates may not be well suited for Wee kda Blue maximum 0.250 electricitydemand EV load. While time-of-use rates are not always the answer, across alldays E 0.200 they are generally a good fit for EVs. These rates do double Blackline= median duty: ensuring consumers can maximize their fuel cost 0.150 demand savings, and incentivizing minimal adverse grid impacts. 0.100 minimum Numerous surveys reveal fuel cost savings are the most - important motivator of EV purchase decisions, as shown in Figure 583 6:00 12:00 18:00 0:00 To illustrate the potential fuel savings from driving on Time of Day electricity to consumers accustomed to buying gasoline, the (The EV Project, 2013)” DOE has an online tool that translates the cost of driving an EV into “eGallons,” which are equivalent to the cost of driving a similar conventional vehicle on gasoline. On a FIGURE7:RESIDENTIALEVCHARGINGINSANDIEGO national average basis, electricity costs $L22/eGallon.8’ REGIONBYTIMEOFDAY

BlueLine maximum However, utility customers do not buy “average” electricity; Weekday electricitydemand prices can vary by utility territory, customer class, season, 1.600 across alldays 0 time of day, marginal consumption, and peak demand. a 1.200 Blackline= median ‘V demand Many of the nation’s largest EV markets have higher-than- OSOO average electricity prices and rates that increase with Red line minimum 0.600 demand marginal consumption. For example, residential customers in PG&E territory pay $3.31/eGallon on the default rate for 0.300 marginal consumption above a certain threshold, whereas customers charging off-peak on PG&E’s EV rate pay only 6:00 12:00 18:00 0:00 $O.97/eGallon.8° TimeofDay (The EV Project, 2013)”

— - Aii/aLc DRIVINGOUT POLLUTION: How UTILITIES CAN ACCELERATETHE MARKETFOR ELECTRIC VEHICLES NRDO July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 16 of 21 Time-of-use rates are also an important tool to encourage Utility leadership is especially needed in the 10 states that drivers to charge during off-peak hours when the electricity have adopted zero-emissions vehicle (ZEV) programs in grid has spare capacity. Without a price signal, drivers will order to meet federal air quality standards. Combined, they generally plug in and charge immediately upon arriving require about 3.3 million ZEVs by 2025, which will require a home after work, exacerbating system-wide evening peak comprehensive charging network where drivers live, work, demand, as shown in Figure 6 for the Dallas/Fort Worth and play. area. This stands in marked contrast to the charging pattern Utilities are beginning to move forward with infrastructure of EV drivers in San Diego (Figure 7), who meet 80 percent investments. A recent decision issued by the California of their refueling needs during the “super-off-peak” period Public Utilities Commission (CPUC) found agreement among from midnight to 5 a.m. on SDG&E’s time-of-use rate. Were charging companies, automakers, utilities, and nonprofit EVs in Dallas on similar rates, instead of exacerbating peak organizations that “utilities should have an expanded role demand and driving the need for additional investments, in EV infrastructure support and development in order to they would be charging while their drivers are asleep and realize the potential benefits of widespread EV adoption.”90 when Texas’ considerable wind generation is often at peak production.87 This decision allowed the CPUC to evaluate separate applications submitted by SDG&E, SCE, and PG&E to install Target Customer Education and more than 60,000 charging stations at public, workplace, Outreach Programs and multiunit residential locations. In January 2016, the CPUC approved a modified version of the SDG&E proposal At a minimum, utilities should prepare their customer and the first phase of the SCE proposal. In March 2016, a service agents and modify their websites to answer common widely supported settlement agreement was proposed in questions from new or prospective EV owners regarding the PG&E proceeding, building on the guidance provided home charger installation, availability of public charging, by the CPUC in its decisions approving the SCE and benefits of off-peak charging, fuel cost savings on applicable SDG&E programs.9’ Meanwhile, the Washington Utilities rates, and other issues. This can help ensure grid safety and and Transportation Commission approved a $3 million EV reliability, and promote a positive consumer experience in infrastructure deployment pilot proposed by Avista, which the early market. However, a more proactive approach is serves rural areas in the eastern Washington and northern needed to avoid adverse grid impacts and maximize fuel cost Idaho. Kansas City Power and Light is also investing $20 savings that motivate EV purchase decisions. million to install more than 1,000 public and workplace Even in California, where utilities have been very active charging stations, and Georgia Power and Light has a $12 with respect to vehicle electrification, the majority of million “Get Current. Drive Electric” charging program customers remain unaware of the potential savings of to install 60 public charging stations with both DC fast switching to time-of-use rates. Utilities should identify and chargers and Level 2 stations. Utilities in other states reach out to customers who would benefit financially by have taken notice; the majority of respondents to a recent switching to more appropriate rates. Rate options and other survey of utility professionals across the nation stated their programs for EV customers must be coupled with targeted utilities are pursuing EV infrastructure deployment as a customer education and outreach. new and emerging revenue stream.9’ Increase Access to Electricity as Transportation 2. CLOSING THE CHARGINGINFRASTRUCTURE Fuel in Disadvantaged Conunmilties GAPANDPROMOTINGEQUITY Phase 2 focuses on robust policies and programs that can To increase access to electricity as a transportation fuel accelerate the EV market and increase access to electricity in communities with the greatest need for cleaner air and as a transportation fuel. Example policies include: lower fuel bills, utilities can target charging infrastructure investments in low-income communities and communities of Utility-Facilitated Deployment of Charging color. Communities of color represent the fastest-growing Infrastructure consumer segment in America, and their buying power will be critical in using EVs to meet long-term air quality As noted in Section B, the lack of access to charging standards and GHG emission reduction targets.9’ As noted infrastructure remains a significant obstacle to widespread in the Greenlining Institute’s 2011 report “Electric Vehicles: EV adoption. Electric utilities are singularly positioned to Who’s Left Stranded?” communities of color are more close the charging infrastructure gap by capturing the value concerned about air pollution, making them a natural but of additional grid services and increased revenues from largely untapped market for ZEVs.94The SDG&E and SCE system-wide charging. Utility involvement is also necessary programs approved by the CPUC both include requirements to ensure that the charging network is expanded in a manner to deploy at least 10 percent of charging stations in that supports the grid to the benefit of all customers. “disadvantaged communities,” as identified by the California

Page Iti DRIVING OUTPOLLUTION: HOW UTILITIES CANACCELERATETHE MARKETFOR ELECTRIC VEHICLES NROC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 17 of 21 Environmental Protection Agency’s CalEnviroScreen 2.0, ice storage used for cooling needs), and it should not be which scores census tracts using 12 types of pollution and excluded from energy storage procurement mandates. environmental factors and seven population characteristics Rather, it should be valued for the services it provides. factors.95 and socioeconomic The settlement proposed SDG&E’s charging infrastructure deployment program, in the PG&E proceeding requires that at least 15 percent recently approved by the CPUC, includes a price-based form of the charging stations be deployed in disadvantaged of Advanced Demand Response that aggregates EV charging communities, with a stretch goal of 20 percent, and also load by deploying banks of grid-integrated charging stations sets aside $5 million for additional equity programs in those at multiunit dwellings and workplaces. It is partially meant communities. to test EV Advanced Demand Response as a form of energy storage. Participating customers charge on a real-time rate Promote Broader Awareness through that reflects hourly wholesale market prices and are billed Mass-Market Education and Outreach on their normal home energy bill. Customers can actively To expand the EV market, a general lack of consumer manage their charging by providing basic parameters (e.g., awareness must be overcome and common misperceptions, when they want the vehicle to be fully charged) or allow often fueled by misleading press coverage, must be the system to minimize costs by absorbing cheap electricity corrected.96 Consumers in the market for a new car need to (e.g., excess solar in the afternoon or excess wind at night) be educated about the benefits of vehicle electrification and and avoiding expensive electricity during evening peak applicable utility rates, incentives, and programs. Utilities hours. are better positioned to conduct this type of broad customer In the future, leveraging the “smarts” and communications education effort than individual automakers seeking to capabilities embedded in EVs themselves may prove a cost- promote specific vehicles, or charging service providers effective solution for Advanced Demand Response. The seeking to promote specific business models. Electric Power Research Institute, utilities, and automakers are developing an Open Vehicle Grid Integration Platform 3. CAPTURING THEVALUEOFGRIDSERVICES that uses non-proprietary communications protocols and advantage of in vehicles ANDINTEGRATINGRENEWABLEENERGY takes the connectivity to manage EV load in response to grid conditions.98 Phase 3 leverages the growing customer investment in EV .4) batteries to provide valuable grid services that can facilitate the integration of renewable energy, and return the value Integrate V2G and Battery Second Life of such services to EV drivers to further accelerate the Programs into Wholesale and Retail Markets market. Example policies include: As noted in Section B(4), leveraging EV drivers’ sunk investment in advanced batteries to provide energy Implement Traditional Demand Response storage that both absorbs excess renewable generation Programs for Electric Vehicle Customers and feeds electricity back to the grid during hours of peak Traditional Demand Response programs that either curtail demand could reduce the cost of integrating wind and or reduce the rate of charging can provide value to both the solar resources and accelerate the EV market. However, if utilities, and do grid and EV drivers without compromising transportation automakers, other relevant parties needs. Such programs can be deployed today with readily not act now to demonstrate V2G and Battery Second available technology. As noted in Section B(4), PG&E Life programs at scale, the opportunity could be lost. and BMW are demonstrating this functionality in the San Significant investments in large-scale energy storage projects are already being made.99 Grid Francisco Bay Area, but they are not alone. Pepco (a utility operators, utilities, serving Maryland and the District of Columbia), Eversource and regulators should prioritize V2G and Battery Second Life programs because they could provide value to utility (a utility serving the Northeast), and SCE have all launched pilot programs to test EV charging as a form of Traditional customers that would otherwise go to private interests and Demand Response.°7 prove more cost-effective by leveraging sunk investments. They also have a unique potential to simultaneously reduce Implement Advanced Demand Response emissions in the electricity and transportation sectors. Programs for Electric Vehicle Customers Advanced Demand Response programs not only curtail or reduce EV charging when necessary, but turn on and ramp up charging to absorb excess renewable generation. Although EV Advanced Demand Response does not feed power back to the grid, it is a form of energy storage that takes power off the grid for use at a later time (like

Papal? DRIVINGOUT POLLUTION: HOW UTILITIES CAN ACCELERATE THE MARKETFOR ELECTRIC VEHICLES £ . - NRDC { - July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 18 of 21 Conclusion

With more than 450,000 modern EVs in the United States consistently triple theft efficiency savings, reducing bills alone, policymakers and utilities need not rely upon for all customers.’°° But removing disincentives does not go conjecture to place transportation electrification on the far enough. Regulators should consider performance-based right path to meet air quality, climate, and equity goals while earnings opportunities to encourage utilities to accelerate supporting the grid and facilitating progress toward other transportation electrification in a maimer that supports clean energy goals. If regulators and the utilities under the grid and facilitates the integration of renewable their jurisdiction fail to take timely action, the expansion energy, in addition to the three phases of utility EV market of the EV market could stall and EV charging could strain acceleration policy outlined in Section C. the grid, necessitating otherwise avoidable costs. However, While they are generally aware of the potential benefits with the right policies and programs, utilities could provide associated with widespread transportation electrification, widespread benefits to all customers, reduce exposure to many utility executives remain focused on other, short-term dangerous and effects of climate air pollution the worst issues that go straight to today’s bottom line. Consequently, change, and provide consumers with a viable alternative to even in utilities with robust transportation programs, EVs the volatile world oil market. remain a secondary priority. Funding for transportation Regulatory directives will be crucial to this effort, but teams can be erratic and is often dependent on the simply commanding utilities to do the right thing is particular interests of company executives. Regulatory not always sufficient; the utility business model for incentives should be realigned to ensure that the most transportation electrification should be aligned with profitable option for utility shareholders minimizes adverse societal interests. This has been proved repeatedly with system impacts, facilitates the integration of renewable respect to energy efficiency; states in which regulators generation, and maximizes system benefits and consumer have decoupled the recovery of authorized expenditures savings relative to gasoline. This would provide a clear and from actual volume of electricity sales in order to remove durable signal to utility leadership to accelerate the pace of the disincentive for utility investments in energy efficiency transportation electrification.

Pag1cJlI DRIVING OUTPOLLUTION: HOW UTILITIES CAN ACCELERATETHE MARKET FOR ELECTRIC VEHICLES NRDC Pagem

July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 19 of 21 ENDNOTES

1 See Williams, J.H. et al., Pathways to Deep Decarbonization in the United States, Energy and Environmental Economics, Inc. (E3), November 2014; California Council on Science and Technology; CaliforniabEnergyFuiure: The View to 2050, May2011; Williams, J.H. et al., “The Technology Path to Deep Greenhouse Gas Emissions Cuts by 2050: The Pivotal Role of Electricity;” Science 335, no. 6064 (January 2012): 53-59; Cunningham, Joshua, “Achieving an 80% GHG Reduction by 2050 in California’s Passenger Vehicle Fleet, “SAElnternational Journal of Passenger Cars—Electronic and Electrical Systems 3, no. 2 (December 2010): 19-36; Wei, Max et al., “Deep Carbon Reductions in California Require Electrification and Integration across Economic Sectors,” EnvironmentalResearch Letters 8, no. 1 (2013); Melaina, M. and K. Webster, “Role of Fuel Carbon Intensity in Achieving 2050 Greenhouse Gas Reductions within the Light-Duty Vehicle Sector,” Environ. Sci. Technol 45, no. 9 (2011): 3865-3871; International Energy Agency, Transport; Energy; and C02:Moving Towards Sustainabilify, OECD/IEA, 2009; National Research Council, Transitions to Alternative Vehicles and Fuels (Washington, D.C.: The National Academies Press, 2013). 2 Ibid. 3 See California Mr Resources Board (CARE), Vision for Clean Air:A Framework forAir Quality and Climate Planning, public review draft, June 27,2012; and National Research Council, Transitions to Alternative Vehicles and Fuels. 4 Fehrenbacher, Katie, Tesla’s Model 3 Reservations Rise to Almost 400,000, Fortune, April15, 2016. 5 Center for Sustainable Energy; California Plug-in Electric Vehicle Owner Survey Dashboard. 6 NREL, California Statewide Plug-in Electric Vehicle Infrastructure Assessment, prepared for the California Energy Commission, CEC-600-2014-2013, May 2014; DOE, Alternative Fuel Data Center, accessed May, 2015. 7 Committee on Overcoming Barriers to Electric-Vehicle Deployment et al., OvercomingRarriers to Deployment ofPlug-inElectric Vehicles (Washington, D.C.: Na tional Academies Press, 2015), www.nap.edu/catalog/21725/overcoming-barriers-to-deployment-of-plug-in-electric-vehicles. 8 Kintner-Meyer, M., K. Schneider, and K. Pratt, ImpactsAssessmentofPlug-inHybrid Vehicles on Electric Utilities andRegional US. Power Grids, Pacific Northwest National Laboratory; November 2007, energyenvironment.pnnl.gov/ei/pdf/PHEV_Feasibffity_Analysis_Partl.pdf 9 Energy and Environmental Economics (E3), California Transportation ElectriScation Assessment, Phase 2: Grid Impacts, California Electric Transportation Coalition, October 2014, www.caletc.com.’%vp-content/uploads/2014/10/CaIETC_TEA_Phase_2_Final_10-23-14.pdf. 10 Assuming a sales-weighted average EV battery size of 24.6 kWh, based on sales by model provided by the Alternative Fuels Data Center, and cumulative sales of 453,069 (California Plug-in Electric Vehicle Collaborative, April 2016). U.S. Department of Energy (DOE), Energy Information Administration (ElM, “District of Columbia Electricity Profile 2012,” accessed April 10, 2015, www.eia.gov/electricity/state/DistrictofColumbia/. 11 NREL, Renewable Electricity Futures Study, Table 12-7, 2013, www.nrel.gov/docs/5’l2osti/52409-2.pdt 12 Ten percent estimate derived from taking the avenge within the range of 15 to 23 million vehicles at level 2 charging capacity (6.6 kW) that would be needed to Mfili the 100 to 152 GW of energy storage capacity modeled by NREL, divided by 175 million EVa (including both battery electric vehicles (8EVs) and plug-in hybrid electric vehicles [PI{EVs]) modeled out to 2050 in National Research Council, Transitions to Alternative Vehicles andFuels, Appendix H, Figure H. 13, “Vehicle Stock by Vehicle Technology Assuming Optintistic PEV Technology Estimates,” 2013, invw.nap.edu/downloa&php?record_id’18264#. 13 Using the current cost estimate of $1,200/kW for a peaker plant as a proxy ceiling price for utility-scale energy storage capacity cost. See California Energy Commission, Estimated Cost ofjirew Renewable and Fossil Generation in California, CEC-200-2014-003-SF, March 2013, p. 140, Table 59, “Mid Case” Capital Cost for Combustion Turbine 100 MW 14 See Calazzo, Fabio et ali, “Mr Poliution and Early Deaths in the United States,” Atmospheric Environment 79 (November 2013): 198-208; National Highway Traffic Safety Administration, Fatality Analysis Reporting System (FARS) Encyclopedia. 15 See Williams, J.H. et al., Pathways to Deep Decarbonization in the United States, Energy and Environmental Economics, Inc. (E3), November 2014; California Council on Science and Technology, Californiak Energy Future: The View to 2050, May2011; Williams, J.H. et al., “The Technology Path to Deep Greenhouse Gas Emissions Cuts by 2050: The Pivotal Role of Electricity,” Science 335, no. 6064 (January2012): 53-59; Cunningham, Joshua, “Achieving an 80% GHG Reduction by 2050 in California’s Passenger Vehicle Fleet, “SAElnternational Journal ofPassenger Cars—Electronic and Electrical Systems 3, no. 2 (December 2010): 19-36; Wei, Max et al., “Deep Carbon Reductions in California Require Electrification and Integration across Economic Sectors;’ EnvironmentalResearch Letters 8, no.1(2013); Melalna, M. and K. Webster, “Role of Fuel Carbon Intensity in Achieving 2050 Greenhouse Gas Reductions within the Light-Duty Vehicle Sector,” Environ. Sci. TechnoL 45, no. 9 (2011): 3865-3871; International Energy Agency, Transport, Energy; and C02:Moving Towards Sustainability, OECD/IEA, 2009; National Research Council, Transitions to Alternative Vehicles and Fuels (Washington, D.C.: The National Academies Press, 2013). 16 See DOE, Alternative Fuels Data Center, www.afdc.energy.gov/. 17 See Kane, Mark, “Consumer Reports: Tesla Model S Rated #1 in Consumer Satisfaction,” Inside EVs, 2015, insideevs.com/consumer-reports-tesla-model-s-rated-1-in- customer-satisfaction/. 18 Fehrenbacher, Katie, Tesla’s Model 3 Reservations Rise to Almost 400,000, Fortune, April iS, 2016. 19 See Williams, J.H. et al, Pathways to Deep Decarbonization in the United States, Energy and Environmental Economics, Inc. (E3), November 2014; California Council on Science and Technology; CalithrnialcEnergy Future: The View to 2050, May2011; Williams, J.H. et al, “The Technology Path to Deep Greenhouse Gas Emissions Cuts by 2050: The Pivotal Role of Electricity;” Science 335, no. 6064 (January 2012): 53-59; Cunningham, Joshua, “Achieving an 80% GHG Reduction by 2050 in California’s Passenger Vehicle Fleet,” SAElnternational Journal ofPassenger Cars—Electronic andElectrical Systems 3, no.2 (December2010): 19-36; Wei, Max et al., “Deep Carbon Reductions in California Require Electrification and Integration across Economic Sectors,” Environmental Research Letters 8, no. 1 (2013); Melalna, M. and K. Webster, “Role of Fuel Carbon Intensity in Achieving 2050 Greenhouse Gas Reductions within the Light-Duty Vehicle Sector,” Environ. Sd. TechnoL 45, no. 9 (2011): 3865-3871; International Energy Agency, Transport; Energy; and C02: Moving Towards Sustainability, OECD/IEA, 2009; National Research Council, Transitions to Alternative Vehicles andFueis (Washington, D.C.: The National Academies Press, 2013). 20 See California Mr Resources Board (CARE), Vision for Clean Air:A Framework forAir Quality and Climate P1anning, public review draft, June 27, 2012; and National Research Council, Transitions to Alternative Vehicles andFu etc. 21 See Caiazzo, Fabio et al., “Mr Pollution and Early Deaths in the United States,” Atmospheric Environment 79 (November 2013): 198-208; National Highway Traffic Safety Administration, Fatality Analysis Reporting System (FARS) Encyclopedia. 22 CARB, Vision for CleanAir. 23 Nichols et al., “Mr Quality Impacts of Electric Vehicle Adoption in Texas,” Transportation ResearchPaflD: Transport and Environment 34 (January 2015): 208-218. 24 Office of Governor Edmund Brown, “Day 4: UN Climate Change Conference: Governor Brown Highlights Bold Subnational Action at COP 21 Events,” press release, December 8, 2015, w%4w.gov.ca.gov/news.phpi’id19235. 25 Peterson, David, “1700 Fast Chargers by 2016,” presentation to the California PEV Coilaborative, Nissan North America, March10, 2015, slide 6 citing survey by PG&E and RDA Group, 2014. 26 Malaina, M., Y. Sun, and A. Brooker, “Vehicle Attributes and Alternative Fuel Station Availability Metrics for Consumer Preference Modeling,” NREL Transportation Center, presented at Energy Commission Workshops, Sacramento, California, March19, 2015.

DRIVINGOUT POLLUTION: HOW UTILITIES CAN ACCELERATETHE MARKETFOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 20 of 21 27 Idaho National Laboratory, Rfll8EVProjectftecfric Veizicle Chaqing Infrastructure Summary Report January 2013 through December 2013. 2$ Li, S. et al,, “The Market for Electric Vehicles: Indirect Network Effects and Policy Impacts,” Cornell University June 2015. 29 Sierzchula, W. et aL, “The Influence of Financial Incentives and Other Socio-Economic Factors on Electric Vehicle Adoption,” Energy Policy 68 (May 2014): 183-194. 30 Peterson, David, “1700 Fast Chargers.” 31 Rovito, M., “Will Nissan’s No Charge to Charge Program Drive LEAF Sales?” Charged Electric Vehicles Magazine, July 3, 2014. 32 Lankton, Cal, Director of EV Infrastructure, Tesla Motor Company, “Plenary Panel: Technology Marches On—The Impact of New Vehicle and Infrastructure Technologies,” EPRI Plug-in 2014 conference, San Jose, California, July 2014. 33 DOE, Alternative Fuel Data Center, accessed March 26, 2015. 34 BREL, California Statewide Plug-in Electric Vehicle Infrastructure Assessment, prepared for the California Energy Commission, CEC-600-2014-2013, May 2014; DOE, Alternative Fuel Data Center, accessed May, 2015. 35 Nigro, N. and Frades, M. Business Models forFinancial(y SustainableEVChargin,q Networks, Center for Climate and Energy Solutions, March 2015. 36 Committee on Overcoming Barriers to Electric-Vehicle Deployment et al, OvercomingRarriers to Deployment ofPlug-inElecfric Vehicles. 37 Electric Power Research Institute (EPRI), “Plug-In Electric Vehicle Multi-State Market and Charging Survey,” Product Abstract, www.epri.com/abstracts/Pages/ ProductAbstract.aspx?ProductId00000000300200293l. 38 See Figure 3 of this report, “Estimated Percent of Time EVs Spend by Location”; see also Alexander, Marcus, Transportation Statistics Anaysis ibrElectric Transportation, EPRI, December 2011. 39 Traut, Elizabeth et al,, “US Residential Charging Potential for Electric Vehicles,” TransportationResearchPartD 25 (November 2013): 139-145. 40 Center for Sustainable fnergy California Plug-in Electric Vehicle Owner Survey Dashboard. 41 White, Brandon, Senior Manager of EV Sales Operations, Nissan North America, “Taking the ‘Work’ Out of Workplace Charging,” presentation at EPRI Plug-in 2014 conference, San Jose, California, July 2014, 42 DOE, Workplace Charging Chattenge--Progress Update 2014: Employers Take Charge, November 2014. 43 Alexander, Marcus, Transportation StatisticsAna(ysis forEtectric Transportation. 44 Hajjar, Norman, New SurveyData. BE VDrivers and theDesire forDCfast Charging, California Plug-in Electric Vehicle Collaborative, March 11,2014. 45 Beauregard, Garrett, Lessons Learned on the E VProject andDC fast Charging, EV Project presentation at Navigant Research webinar, FastDC Charging forElectric Vehicles, April 9, 2013. 46 See EPifi, Fast Charging Technology. 47 Kintner-Meyer, M., K. Schneider, and R. Pratt, Impacts Assessment ofPtug-in Hybrid Vehicles. 48 Ibid. 49 Environmental and Energy Economics, California Transportation Electrification Assessment. 50 Idaho National Laboratory, 2013 EVProjectElectric Vehicle Charging Infrastructure Summary Report, January 2013 through December 2013.

51 Berkheimer, 3, et al. ‘iS’tectric Grid Integration Costs for Plug-In Electric Vehicles,” SAEInt, J Alt, Power 3, no. 1 (2014), doi:10.4271/2014-01-0344. 52 Assuming a sales-weighted average EV battery size of 24.6 kWh, based on sales by model provided by the Alternative Fuels Data Center, and cumulative sales of 453,069 (CaliforiaPlug-in Electric Vehicle Collaborative, April 2016). U.S. Department of Energy (DOE), Energy Information Administration (ElM, “District of Columbia Electricity Profile 2012,” accessed April 10, 2015, www.eia.gov/electricity/state/DistrictofColumbia/. 53 Chart adapted from Adam Langton and Noel Crisotomo, Vehicle-Grid Integration, California Public Utifities Commission, October 2013,, www.cpuc.ca.gov/ WorkArea/DownloadAsset.aspx?id”7744. 54 from 1970 to 2008: U.S. Department of Transportation, Federal HighwayAdmiistration, Highway Statistics 2009,2011, Table Wvl-1and annual. From 2009 on, see Appendix A for Car/Light Truck Shares, (Additional resources: www.fhwa.dot.gov.) 55 California Energy Commission, Estimated Cost ofNew Renewable and Fossil Generation in California. 56 NREL, RenewableEtectricity Futures Study, 12-16, 2013. www.nrel.gov/docs/fyl2osti/52409-2.pdf. 57 Energy and Environmental Economics, Inc. (E3), Investsqating a H(qherRenewables Portfolio Standard in California, January 2014. 58 See Electric Reliability Council of Texas (ERCOT), Wind IntegrationReport, May 17, 2014. 59 Vattenfall AG, Klimaentlastung durch den Einsatz erneuerbarerEnergien im Zusammenwirken mit emissionsfreien Elekfrofahrzeugen (“Climate Change Mitigation Through Usage of Renewable Energy Sources in Combination with Emission Free Electric Vehicles”), March 2011. 60 Energy and Environmental Economics, Inc., (E3), Investigating a HzqherRenewables Portfolio Standard in California, 61 These batteries were taken out of a fleet ofpreproduction vehicles used for several years by real-world customers. 62 New York State Energy Research and Development Authority (NYSERDA), Compilation of Utility Commission Initiatives Related to Plug-in Electric Vehicles and Electric Vehicle Suppty Equipment, April 2013, Plug-acc.pdf. 63 Shepard, Scott and Gartner, John, Vehicle to Grid Technologies, Navigant Research, 2013. 64 Gorguinpour, Camron, DOD, Office of the Assistant Secretary of the Mr force, “Plug-In Electric Vehicle Program: The DOD V2G Pilot Project Overview,” 2013, electricvehicle.ieee.org/ffies/2013/03/DoD-Plug-In-Electric-Vehicle-Program.pdf. 65 Annual value is calculated based on the 18-month pilot estimated to return a total of $540 per vehicle over that period, and multiplying by 2/3 (12 months/18 months) to equal $360 per year. 66 Energy and Environmental Economics (E3), California TransportationEtecfrificationAssessmentPltase2. 67 It is important to note that the estimated average of $1,800 per year per vehicle was based on a small fleet of 15 vehicles that were part of a closely managed duty cycle and usage reservation system. Therefore we caution against maldng broader conclusions from this pilot about the V2G value stream for a more broadly applicable scenario involving more random driving cycles with different vehicle types and charging capacities. Wald, Matthew L., “Electric Vehicles Begin to Earn Money from the Grid,”New York Times, April 25, 2013,

I’s.q.FOI DRIVING OUT POLLUTION:HOW UTILITIES CAN ACCELERATE THEMARKET FOR ELECTRICVEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-2; Source: NRDC Report on Driving Down Pollution Page 21 of 21 68 It is important to note that the estimated average of $2,520 per year per vehicle was based on a vehicle fleet that includes EV-retrofitted mid-sized trucks that were used for daily-patterned duty cycles and on a managed reservation system. Thus we caution in making broader conclusions from this pilot about the V2G value stream for a more broadly applicable scenario involving more random driving cycles with different vehicle types charging capacities. Gorguinpour, Camron, DOD, Office of the Assistant Secretary of the Air Force, “Plug-In Electric Vehicle Program: The DOD V2G Pilot Project Overview,” 2013, electricvehicle.ieee.org/ffies/2013/03/DoD-Plug- In-Electric-Vehicle-Program.pdt

69 A general nile of thumb used to determine the market demand for frequency regulation in a given regional grid today is to take 1 percent of the average total capacity. Thus for a California ISO that has an average system capacity of around 30 GW, the demand for frequency regulation is on average around 300 MW; ilkewise, for the PJM market the average system capacity is around 90 GW, and therefore the average demand for frequency regulation is around 900 MW. The average level 2 charging capacity standard is 6.6 kW. See Monitoring Analytics, State of the MarketReport for PJM 2014. www.monitoringanalytics.com/reports/PJM_State_of_the_ Market/2014/2014-som-pjm-volume2-sec3.pdf. 70 According to modeled estimates from Denhoim, P. et aL, “The Impact of Wind and Solar on the Value of Energy Storage,” REEL, 2013, www.nrel.gov/docs/ r14osti/60568.pdf. 71 NEEL (2013) Renewable Electric icy Futures Study, Table 12-7. 72 Ten percent estimate derived from taldng the average within the range of 15 to 23 million vehicles at level 2 charging capacity (6.6 kW) that would be needed to fiilfiil the 100 to 152 GW of energy storage capacity modeled by REEL, divided by 175 million EVa (including both battery electric vehicles (BEVs) and plug-in hybrid electric vehicles [PHEW]) modeled out to 2050 in National Research Council, Transitions toAltensative Vehicles and Fuels, Appendix H, Figure H. 13, “Vehicle Stock by Vehicle Technology Assuming Optimistic PEV Technology Estimates,” 2013, www.nap.edu/download.php?recordjd=18264#. 73 Using the current cost estimate of Sl,200/kW for a peaker plant as a proxy ceifing price for utility-scale energy storage capacity cost. See California Energy Conmilssion, Estimated Cost of/Yew Renewable and Fossil Generation in California, CEC-200-2014-003-SF, March 2013, p. 140, Table 59, “Mid Case” Capital Cost for Combustion Turbine 100 MW. 74 Center for Climate and Energy Solutions (C2ES), “Who Can Own/Operate a Charging Station,” accessed March 24, 2015, www.c2es.org/initiatives/pev/maps/who- can-own-operate-a-charging-station. -- 75 California Pubilc Utilities Code Section 216(i) strikes an appropriate balance. 76 See California Auto Outlook 12, no.1 (February 2016); Pacific Gas & Electric, San Diego Gas & Electric, Southern California Edison, JointlOVElectric Vehicle Load Research Report 4th Report, ified December 24, 2015. http://docs.cpoc.ca.gov/PublishedDocs/Effie/G000/M158/K122/158122370.PDF. 77 Energy and Environmental Economics (E3), California Transportation Electrification Assessm ent Phase 2. 78 Today’s EVs are generally capable of charging at 1.2,3.3, or 6.6 kilowatts, a range that is comparable to the range in peak summer demand between typical California coastal and inland homes, according to Pacific Gas & Electric. 79 California Public Utilities Commission, Joint IOUAssessnwnt Report forE VNotification, Rulemaking 09-08-009 (2011), at 27. 80 C2ES, “PEV integration with Electrical Grid,” accessed March 24, 2015, www.c2es.org/initiatives/pev/maps/pev-integration-with-electrical-grid. 81 California Public Utilities Commission, Joint IOUAssessmentReport forEVNotification, Rulemaking 09-08-009 (2011), at 27. 82 See California Senate Bill 859 (Padiila, 2011). 83 Center for Sustainable Energy, California Plug-in Electric Vehicle Owner Survey Dashboard; Steele, David E., J.D. Power and Associates, “Predicting Progress: What We Are Learning About Why People Buy and Do Not Buy EVs,” Electric Drive Transportation Association 2013 Annual Meeting, Washington, D.C., June 11, 2013; Maritz Research, “Consumers’ Thoughts, Attitudes, and Potential Acceptance of Electric Vehicles,” National Research Council meeting, Washington, D.C., August13, 2013. 84 Center for Sustainable Energy, California Plug-in Electric Vehicle Owner Survey Dashboard. 85 Using Department of Energy eGalinn assumptions and methodology (electric vehicle efficiency of 0.35 kilowatt-hour/mile and conventional vehicle fuel economy of 28.2 miles per gallon). 86 This margin will decrease with recently adopted default residential rates. 87 See ERCOT, Wind Integration Report. 88 Idaho National Laboratory, 2OlSEVProjectElectric Vehicle Cha rging Infrastructure Summary Report, January 2013 through December 2013. 89 Ibid.

90 California Public Utliities Comsnlssion, “Phase 1Decision Establishing Policy to Expand the Utifities” Role in Development in Electric Vehicle Infrastructure,” Deci sion 14-12-079, December 18, 2014. 91 See Baumhefiier, Max, “Proposal to Charge Electric Cars in Southern California Gets the Green Light,” Natural Resources Defense Council (NTtDC), January14, 2016; Bamnhefiier, Max, “San Diego Gas & Electric to Use Electric Cars to Integrate Renewable Energy,” NItDC, January 28, 2016; Agreement Proposed to Deploy Charging Stations for Electric Cars in Northern and Central California, Medium, March 21, 2016. 92 Utliity Dive, 2016 State of the Electric Utility Survey, p. U. 93 See genera4y Treuhaft, S., J. Scoggins, and J. Tran, “The Equity Solution: Racial Inclusion Is Eey to Growing a Strong New Economy,” PolicyLinit, October, 2014. 94 Song, C.C., “Electric Vehicles: Who’s Left Stranded?” Greenlining Institute, August 2011, p. 4. 95 The U.S. Environmental Protection Agency has developed a similar national tool: www2.epa.gov/ejscreen. 96 See Krause, R.M. et al., “Perception and Reality: Public Knowledge of Plug-In Electric Vehicles in 21 U.S. Cities,” EnerRy Policy 63 (December 2013): 433-440. 97 Itron, “Pepco Deploys Itron and ClippecCreek Electric Vehicle Smart Charging Solution;’ press release, November 6, 2014, nnv.itron.com/na/newsAndEvents/ Pages/Pepco-Deploys-Itron-and-ClipperCreek-Electric-Vehicle-Smart-Charging-Solution.aspx; Rivera-Linares, Corina, “NSTAR Electric to Kick Off Vehicle Program,” Electric Light & Power/PowerGrid International, June 27, 2014, tnvw.elp.com/articles/2014/06/nstar-electric-to-kick-off-electric-vehicle-program.htnil; California Public Utilities Comnilssion, Advice Letter 2746-E, January 3, 2013, www.sce.conVNR/sc3/tm2/pdf/2746-E.pdf. 98 See EPRI, “Unified Plug-in Electric Vehicle (PEV) to Smart Grid Integration Approach Within Automotive and Utility Industries,” Product Abstract, December 2013. 99 Kelly-Detwiler, Peter, “Storage Industry Celebrates Big Win in California,” Forbes, November 5, 2014. 100 Kushier, Martin, “122 vs. EERS: There’s One Clear Winner Among State Energy Efficiency Policies,” American Council for an Energy-Efficient Economy (ACEEE), December16, 2014, aceee.org/blog/2014/12/irp-vs-eers-there%E2%80%99s-one-clear-wiuner-.

Paqe2J j DRIVINGOUT POLLUTION: HOW UTILITIES CAN ACCELERATE THEMARKET FOR ELECTRIC VEHICLES NRDC July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 1 of 153

This PDF is available from The National Academies Press at http://www.nap.edu/catalog.php?record_id=21725

Overcoming Barriers to Deployment of Plug-in Electric Vehicles

ISBN Committee on Overcoming Barriers to Electric-Vehicle Deployment; Board 978-0-309-37217-6 on Energy and Environmental Systems; Division on Engineering and Physical Sciences; Transportation Research Board; National Research 204 pages Council 8.5 x 11 PAPERBACK (2015)

Visit the National Academies Press online and register for...

Instant access to free PDF downloads of titles from the

NATIONAL ACADEMY OF SCIENCES

NATIONAL ACADEMY OF ENGINEERING

INSTITUTE OF MEDICINE

NATIONAL RESEARCH COUNCIL

10% off print titles

Custom notification of new releases in your field of interest

Special offers and discounts

Distribution, posting, or copying of this PDF is strictly prohibited without written permission of the National Academies Press. Unless otherwise indicated, all materials in this PDF are copyrighted by the National Academy of Sciences. Request reprint permission for this book

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 2 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

OVERCOMING BARRIERS TO DEPLOYMENT OF PLUG-IN ELECTRIC VEHICLES

Committee on Overcoming Barriers to Electric-Vehicle Deployment

Board on Energy and Environmental Systems

Division on Engineering and Physical Sciences

and

Transportation Research Board

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 3 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

THE NATIONAL ACADEMIES PRESS 500 Fifth Street, NW Washington, DC 20001

NOTICE: The project that is the subject of this report was approved by the Governing Board of the National Research Council, whose members are drawn from the councils of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine. The members of the committee responsible for the report were chosen for their special competences and with regard for appropriate balance.

This project was supported by Contract DE-EE0004436 between the National Academy of Sciences and the U.S. Department of Energy. Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the authors and do not necessarily reflect the view of the organizations or agencies that provided support for this project.

International Standard Book Number-13: 978-0-309-37217-6 International Standard Book Number-10: 0-309-37217-8

Library of Congress Control Number: 2015939639

Additional copies of this report are available for sale from the National Academies Press, 500 Fifth Street, NW, Keck 360, Washington, DC 20001; (800) 624-6242 or (202) 334-3313; http://www.nap.edu.

Printed in the United States of America.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 4 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished scholars engaged in scientific and engineering research, dedicated to the furtherance of science and technology and to their use for the general welfare. Upon the authority of the charter granted to it by the Congress in 1863, the Academy has a mandate that requires it to advise the federal government on scientific and technical matters. Dr. Ralph J. Cicerone is president of the National Academy of Sciences.

The National Academy of Engineering was established in 1964, under the charter of the National Academy of Sci- ences, as a parallel organization of outstanding engineers. It is autonomous in its administration and in the selection of its members, sharing with the National Academy of Sciences the responsibility for advising the federal government. The National Academy of Engineering also sponsors engineering programs aimed at meeting national needs, encourages education and research, and recognizes the superior achievements of engineers. Dr. C. D. Mote, Jr., is president of the National Academy of Engineering.

The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the services of eminent members of appropriate professions in the examination of policy matters pertaining to the health of the public. The Institute acts under the responsibility given to the National Academy of Sciences by its congressional charter to be an adviser to the federal government and, upon its own initiative, to identify issues of medical care, research, and education. Dr. Victor J. Dzau is president of the Institute of Medicine.

The National Research Council was organized by the National Academy of Sciences in 1916 to associate the broad community of science and technology with the Academy’s purposes of furthering knowledge and advising the federal government. Functioning in accordance with general policies determined by the Academy, the Council has become the principal operating agency of both the National Academy of Sciences and the National Academy of Engineering in providing services to the government, the public, and the scientific and engineering communities. The Council is administered jointly by both Academies and the Institute of Medicine. Dr. Ralph J. Cicerone and Dr. C. D. Mote, Jr., are chair and vice chair, respectively, of the National Research Council.

www.national-academies.org

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 5 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 6 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

COMMITTEE ON OVERCOMING BARRIERS TO ELECTRIC-VEHICLE DEPLOYMENT

JOHN G. KASSAKIAN, Chair, NAE,1 Massachusetts Institute of Technology DAVID BODDE, Clemson University, Clemson, South Carolina JEFF DOYLE, D’Artagnan Consulting, Olympia, Washington GERALD GABRIELSE, NAS,2 Harvard University, Cambridge, Massachusetts KELLY SIMS GALLAGHER, Tufts University, Medford, Massachusetts (until June 2014) ROLAND HWANG, Natural Resources Defense Council, San Francisco, California PETER ISARD, Consultant, Washington, D.C. LINOS JACOVIDES, NAE, Michigan State University, East Lansing ULRIC KWAN, IBM Global Business Services, Palo Alto, California REBECCA LINDLAND, King Abdullah Petroleum Studies and Research Center, Riyadh, Saudi Arabia RALPH MASIELLO, NAE, DNVGL, Inc., Chalfont, Pennsylvania JAKKI MOHR, University of Montana, Missoula MELISSA SCHILLING, New York University, Stern School of Business, New York RICHARD TABORS, Across the Charles, Cambridge, Massachusetts THOMAS TURRENTINE, University of California, Davis

Staff ELLEN K. MANTUS, Project Codirector K. JOHN HOLMES, Project Codirector JAMES ZUCCHETTO, Board Director JOSEPH MORRIS, Senior Program Officer LIZ FIKRE, Senior Editor MICHELLE SCHWALBE, Program Officer ELIZABETH ZEITLER, Associate Program Officer IVORY CLARKE, Senior Program Assistant LINDA CASOLA, Senior Program Assistant

1 NAE, National Academy of Engineering. 2 NAS, National Academy of Sciences. v

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 7 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

BOARD ON ENERGY AND ENVIRONMENTAL SYSTEMS

ANDREW BROWN, JR., Chair, NAE,1 Delphi Corporation, Troy, Michigan DAVID T. ALLEN, University of Texas, Austin W. TERRY BOSTON, NAE, PJM Interconnection, LLC, Audubon, Pennsylvania WILLAM BRINKMAN, NAS,2 Princeton University, Princeton, New Jersey EMILY CARTER, NAS, Princeton University, Princeton, New Jersey CHRISTINE EHLIG-ECONOMIDES, NAE, Texas A&M University, College Station NARAIN HINGORANI, NAE, Independent Consultant, San Mateo, California DEBBIE NIEMEIER, University of California, Davis MARGO OGE, McLean, Virginia MICHAEL OPPENHEIMER, Princeton University, Princeton, New Jersey JACKALYNE PFANNENSTIEL, Independent Consultant, Piedmont, California DAN REICHER, Stanford University, Stanford, California BERNARD ROBERTSON, NAE, Daimler-Chrysler (retired), Bloomfield Hills, Michigan DOROTHY ROBYN, Washington, D.C. GARY ROGERS, Roush Industries, Livonia, Michigan ALISON SILVERSTEIN, Consultant, Pflugerville, Texas MARK THIEMENS, NAS, University of California, San Diego ADRIAN ZACCARIA, NAE, Bechtel Group, Inc. (retired), Frederick, Maryland MARY LOU ZOBACK, NAS, Stanford University, Stanford, California

Staff JAMES ZUCCHETTO, Board Director DANA CAINES, Financial Associate ALAN CRANE, Senior Scientist K. JOHN HOLMES, Senior Program Officer/Associate Director MARTIN OFFUTT, Senior Program Officer ELIZABETH ZEITLER, Associate Program Officer LANITA JONES, Administrative Coordinator LINDA CASOLA, Senior Program Assistant ELIZABETH EULLER, Program Assistant JONATHAN YANGER, Research Associate

1 NAE, National Academy of Engineering. 2 NAS, National Academy of Sciences. vi

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 8 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

TRANSPORTATION RESEARCH BOARD 2014 EXECUTIVE COMMITTEE1

KIRK T. STEUDLE, Director, Michigan Department of Transportation, Lansing, Chair DANIEL SPERLING, Professor of Civil Engineering and Environmental Science and Policy; Director, Institute of Transportation Studies, University of California, Davis, Vice Chair ROBERT E. SKINNER, JR., Transportation Research Board, Executive Director VICTORIA A. ARROYO, Executive Director, Georgetown University Climate Center, Assistant Dean, Centers and Institutes, Professor from Practice, and Environmental Law Program Director, Georgetown Law, Washington, D.C. SCOTT E. BENNETT, Director, Arkansas State Highway and Transportation Department, Little Rock JAMES M. CRITES, Executive Vice President of Operations, Dallas-Fort Worth International Airport, Texas MALCOLM DOUGHERTY, Director, California Department of Transportation, Madera A. STEWART FOTHERINGHAM, Professor, University of St. Andrews, United Kingdom JOHN S. HALIKOWSKI, Director, Arizona Department of Transportation, Phoenix MICHAEL W. HANCOCK, Secretary, Kentucky Transportation Cabinet, Frankfort SUSAN HANSON, Distinguished University Professor Emerita, School of Geography, Clark University, Worcester, Massachusetts STEVE HEMINGER, Executive Director, Metropolitan Transportation Commission, Oakland, California CHRIS T. HENDRICKSON, Duquesne Light Professor of Engineering, Carnegie Mellon University, Pittsburgh, Pennsylvania JEFFREY D. HOLT, Managing Director, Bank of Montreal Capital Markets, and Chairman, Utah Transportation Commission, Huntsville, Utah GARY P. LAGRANGE, President and CEO, Port of New Orleans, Louisiana MICHAEL P. LEWIS, Director, Rhode Island Department of Transportation, Providence JOAN MCDONALD, Commissioner, New York State Department of Transportation, Albany ABBAS MOHADDES, President and Chief Executive Officer, ITERIS, Inc., Santa Ana, California DONALD A. OSTERBERG, Senior Vice President, Safety and Security, Schneider National, Inc., Green Bay, Wisconsin STEVE PALMER, Vice President of Transportation (retired), Lowe’s Companies, Inc., Mooresville, North Carolina HENRY G. (GERRY) SCHWARTZ, JR., Chairman (retired), Jacobs/Sverdrup Civil, Inc., St. Louis, Missouri KUMARES C. SINHA, Olson Distinguished Professor of Civil Engineering, Purdue University, West Lafayette, Indiana GARY C. THOMAS, President and Executive Director, Dallas Area Rapid Transit, Dallas, Texas PAUL TROMBINO, Director, Iowa Department of Transportation, Ames PHILLIP A. WASHINGTON, General Manager, Denver Regional Council of Governments, Denver, Colorado

1 Membership as of October 2014. vii

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 9 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 10 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Preface

The plug-in electric vehicle (PEV) holds much prom- Ron Adner, Dartmouth College, ise—from reducing dependence on imported petroleum to William F. Brinkman, NAS, Princeton University, decreasing greenhouse gas emissions to improving urban air Yet-Ming Chiang, NAE, Massachusetts Institute of Technology, quality. However, there are many barriers to its mainstream George Eads, Charles River Associates, adoption regardless of incentives and enticing promises to Gregory A. Franklin, University of Alabama at Birmingham, solve difficult problems. Such vehicles have some -limita John D. Graham, Indiana University, tions owing to current battery technology, such as restricted Christopher T. Hendrickson, NAE, Carnegie Mellon University, electric driving range and the long times required for battery Jeremy J. Michalek, Carnegie Mellon University, charging. Furthermore, they cost more than conventional John O’Dell, Edmunds.com, vehicles and require an infrastructure for charging the bat- Margo Tsirigotis Oge, U.S. Environmental Protection tery. Given those concerns, the U.S. Congress asked the De- Agency (retired), partment of Energy to commission a study by the National Karl Popham, Austin Energy, and Research Council (NRC) that would investigate the barriers Mike Tamor, Ford Motor Company. and recommend ways to overcome them. In this final comprehensive report, the Committee on Although the reviewers listed above have provided Overcoming Barriers to Electric-Vehicle Deployment first many constructive comments and suggestions, they were not discusses the current characteristics of PEVs and charging asked to endorse the conclusions or recommendations, nor technologies. It then briefly reviews the market-development did they see the final draft of the report before its release. The process, presents consumer demographics and attitudes to- review of the report was overseen by the review coordina- ward PEVs, and discusses the implications of that infor- tor, Maxine Savitz, NAE, Honeywell Inc. (retired), and the mation and other factors on PEV adoption and diffusion. review monitor, M. Granger Morgan, NAS, Carnegie Mellon The committee next explores how federal, state, and local University. Appointed by NRC, they were responsible for governments and their various administrative arms can be making certain that an independent examination of the report more supportive and implement policies to sustain benefi- was carried out in accordance with institutional procedures cial strategies for PEV deployment. It then provides an in- and that all review comments were carefully considered. Re- depth discussion of the PEV charging-infrastructure needs sponsibility for the final content of the report rests entirely and evaluates the implications of PEV deployment on the with the committee and the institution. The committee grate- electricity sector. Finally, the committee discusses incentives fully acknowledges the following for their presentations dur- for adopting PEVs. ing open sessions of the committee meetings: The current report has been reviewed in draft form by persons chosen for their diverse perspectives and technical Ali Ahmed, Cisco Systems, Inc., expertise in accordance with procedures approved by the Marcus Alexander, Electric Power Research Institute, NRC Report Review Committee. The purpose of the inde- Menahem Anderman, Advanced Automotive Batteries, pendent review is to provide candid and critical comments Greg Brown, Serra Chevrolet, that will assist the institution in making its published report Allison Carr, Houston-Galveston Area Clean Cities Coalition, as sound as possible and to ensure that the report meets insti- William P. Chernicoff, Toyota Motors North America, Inc., tutional standards of objectivity, evidence, and responsive- Mike Cully, Car2Go, ness to the study charge. The review comments and draft Tammy Darvish, DARCARS Automotive Group, manuscript remain confidential to protect the integrity of the Patrick B. Davis, U.S. Department of Energy, deliberative process. We thank the following people for their Katie Drye, Advanced Energy, review of this report: Rick Durst, Portland General Electric,

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 11 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

x Preface

Alexander Edwards, Strategic Vision, Austin Energy, Austin, Texas, James Francfort, Idaho National Laboratory, Berlin Agency for Electric Mobility (eMO), Berlin, Germany, Linda Gaines, Argonne National Laboratory, Charging Network Development Organization, Tokyo, Japan, Camron Gorguinpour, U.S. Department of Defense, Climate Change Policy Headquarters, City of Yokohama, David Greene, Oak Ridge National Laboratory, Federal Government Joint Unit for Electric Mobility Doug Greenhaus, National Automobile Dealers Association, (GGEMO), Berlin, Germany, Britta K. Gross, General Motors, German Institute for Transportation Research (DLR), Jonna Hamilton, Electrification Coalition, Berlin, Germany, Steve Hanson, Frito-Lay, Innovation Centre for Mobility and Societal Change, Jack Hidary, Hertz, Berlin, Germany, John H. Holmes, San Diego Gas and Electric, Japan Charge Network, Co., Kanagawa, Japan, Dana Jennings, Lynda.com, Inc., Kanagawa Prefectural Government, Kanagawa, Japan, Donald Karner, ECOtality North America, Kyoto Prefectural Government, Kyoto, Japan, Elise Keddie, California Air Resources Board, Ministry of Economy, Trade, and Industry, Tokyo, Japan, Ed Kim, AutoPacific, Ministry of Infrastructure and the Environment and Neil Kopit, Criswell Automotive, Netherlands School of Public Administration, The Hague, Michael Krauthamer, eVgo, The Netherlands, Richard Lowenthal, ChargePoint, MRA-Elektrisch, Amsterdam, The Netherlands, Brewster McCracken, Pecan Street Inc., Nissan Motor Co., Yokohama, Japan, John Miller, JNJ Miller plc, NRG eVgo, Houston, Texas, Russ Musgrove, FedEx Express, Okayama Vehicle Engineering Center, Okayama, Japan, Michael Nicholas, Institute of Transportation Studies, Osaka Prefectural Government, Osaka, Japan, University of California, Davis, Pecan Street Research Institute, Austin, Texas, Nick Nigro, Center for Climate and Energy Solutions, Technical University of Eindhoven and BrabantStad, Sarah Olexsak, U.S. Department of Energy, Eindhoven, The Netherlands, John Rhow, Kleiner Perkins, Tesla, The Netherlands, Paul Scott, Downtown Los Angeles Nissan, Tokyo Electric Power Company, Kanagawa, Japan, Chuck Shulock, Shulock Consulting, Urban Development Group, City of Rotterdam, Lee Slezak, U.S. Department of Energy, The Netherlands, and John Smart, Idaho National Laboratory, Vattenfall, Berlin, Germany. Suresh Sriramulu, TIAX LLC, Mark Sylvia, Massachusetts Department of Energy Resources, The committee is also grateful for the assistance of the Mike Tamor, Ford Motor Company, NRC staff in preparing this report. Staff members who con- Joseph Thompson, Nissan, tributed to the effort are Ellen Mantus and K. John Holmes, Chris Travell, Maritz Research, Project Codirectors; James Zucchetto, Director of the Board Jacob Ward, U.S. Department of Energy, on Energy and Environmental Systems; Joseph Morris, Se- Jason Wolf, Better Place, and nior Program Officer for the TRB; Liz Fikre, senior editor; Tracy Woodard, Nissan. Michelle Schwalbe, Program Officer; Elizabeth Zeitler, As- sociate Program Officer, and Ivory Clarke and Linda Casola, The committee also wishes to express its gratitude to Senior Program Assistants. Tomohisa Maruyama, Ministry of Economy, Trade, and In- I especially thank the members of the committee for dustry, Tokyo, Japan, and Sumiyo Hirano, Next Generation their efforts throughout the development of this report. Vehicle Promotion Center, Tokyo, Japan, for arranging an informative visit to Japan and accompanying the members John G. Kassakian, Chair as they traveled through Japan. The committee also wishes Committee on Overcoming Barriers to thank the following for providing valuable information to Electric-Vehicle Deployment and extending hospitality to the committee during its visits to Germany, Japan, The Netherlands, and Texas:

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 12 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Contents

SUMMARY…………………………………………………………………………………………………………………………….1

1 INTRODUCTION……………………………………………………………………………………………………………8 Historical and Policy Context, 8 The Plug-in Electric Vehicle and Current Sales, 9 Plug-in Electric Vehicles: Benefits and Trade-offs, 12 The Committee and Its Task, 13 The Committee’s Approach to Its Task, 16 Organization of This Report, 16 References, 17

2 PLUG-IN ELECTRIC VEHICLES AND CHARGING TECHNOLOGIES………………………………………………19 Types of Plug-in Electric Vehicles, 19 High-Energy Batteries, 22 Relative Costs of Plug-in Electric and ICE Vehicles, 25 Vehicle Charging and Charging Options, 29 References, 34

3 UNDERSTANDING THE CUSTOMER PURCHASE AND MARKET DEVELOPMENT PROCESS FOR PLUG-IN ELECTRIC VEHICLES…………………………………………………………………………37 Understanding and Predicting the Adoption of New Technologies, 37 Demographics and Implications for Adoption and Diffusion of Vehicles, 39 The Mainstream Consumer and Possible Barriers to Their Adoption of Plug-in Electric Vehicles, 46 Vehicle Dealerships: A Potential Source of Information? 51 Strategies to Overcome Barriers to Deployment of Plug-in Electric Vehicles, 53 Federal Government Efforts to Familiarize Consumers with Plug-in Electric Vehicles: Clean Cities Coalition, 57 Fleet Purchases, 58 References, 60

4 GOVERNMENT SUPPORT FOR DEPLOYMENT OF PLUG-IN ELECTRIC VEHICLES…………………………65 Federal Government Research Funding to Support Deployment of Plug-in Electric Vehicles, 65 Institutional Support for Promoting Plug-in Electric Vehicle Readiness, 66 Transportation Taxation and Financing Issues Related to Plug-in Electric Vehicles, 67 Streamlining Codes, Permits, and Regulations, 74 Ancillary Institutional Issues Related to Support for Plug-in Electric Vehicles, 75 References, 78

5 CHARGING INFRASTRUCTURE FOR PLUG-IN ELECTRIC VEHICLES…………………………………………82 Charging Infrastructure and Effects on Deployment of Plug-in Electric Vehicles and on Electric Vehicle Miles Traveled, 82 Models for Infrastructure Deployment, 90 References, 95

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 13 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

xii Contents 6 IMPLICATIONS OF PLUG-IN ELECTRIC VEHICLES FOR THE ELECTRICITY SECTOR……………………98 The Physical and Economic Structure of the Electricity Sector, 99 Generation and Transmission, 100 Physical Constraints in the Distribution Infrastructure, 100 Potential Economic Constraints or Impediments within the Delivery System, 102 Electricity Sector Regulatory Issues for Operating a Public Charging Station, 105 The Utility of the Future, 107 References, 107

7 INCENTIVES FOR THE DEPLOYMENT OF PLUG-IN ELECTRIC VEHICLES………………………………109 Vehicle Price and Cost of Ownership, 109 Price and Cost Competitiveness of Plug-in Electric Vehicles, 110 Possibilities for Declines in Production Costs for Plug-in Electric Vehicles, 112 Incentives, 113 Price of Conventional Transportation Fuels as an Incentive or a Disincentive for the Adoption of Plug-in Electric Vehicles, 119 Past Incentives on Other Alternative Vehicles and Fuels, 120 Recommendations, 123 References, 123

APPENDIXES

A BIOGRAPHICAL INFORMATION ON THE COMMITTEE ON OVERCOMING BARRIERS TO ELECTRIC-VEHICLE DEPLOYMENT……………………………………………………………129 B MEETINGS AND PRESENTATIONS…………………………………………………………………………………133 C INTERNATIONAL INCENTIVES………………………………………………………………………………………136

BOXES, FIGURES, AND TABLES

BOXES 1-1 Statement of Task, 14 3-1 Calculating Electricity or Fuel Costs for Plug-in Electric and Other Vehicles, 49 5-1 Some Hypothetical Economics for Providers of Public Charging, 95 7-1 Derivation of Petroleum Equivalent for a Battery Electric Vehicle, 115 7-2 Financial Incentives, 117

FIGURES 1-1 U.S. BEV monthly sales data from 2010 to 2014, 10 1-2 U.S. PHEV monthly sales data from 2010 to 2014, 11 1-3 World PEV sales in 2012, 2013, and 2014, 11 1-4 The rate of PEV market growth in its first 34 months superimposed on the rate of HEV market growth during its first 34 months, 12 1-5 Projected annual light-duty PEV sales as a percentage of total light-duty vehicle sales, 13 2-1 The volume energy density and the mass energy density for various battery types, 23 2-2 Representation of a lithium-ion battery that shows lithium ions traveling between the anode and the cathode and electrons traveling through the external circuit to produce an electric current, 24 2-3 Effect of ambient temperature on battery capacity on a 20 kWh battery in a PHEV, 26 2-4 Change in the sales price of NiMH, Li-ion, and NiCd battery cells from 1999 to 2012, 28 2-5 For AC level 1, a vehicle is plugged into a single-phase 120 V electric socket through a portable safety device called an electric vehicle supply equipment (EVSE), 30 2-6 The SAE J1772 plug that connects all PEVs to AC level 1 and level 2 is an agreed-on universal standard for 120 V and 240 V ac charging, 30 2-7 For AC level 2 charging, a vehicle is plugged into a split-phase 240 V electric circuit like those used by electric dryers, stoves, and large air conditioners through a wall- or post-mounted safety device called an electric vehicle supply equipment (EVSE), 31

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 14 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Contents xiii 2-8 Four plugs and control protocols are now being used for DC fast charging, 32 2-9 DC fast charging a Nissan Leaf, 32 2-10 As of February 2015, Tesla had installed 190 units in the United States, 33 3-1 Years needed for fastest growing consumer technologies to achieve penetration (0-50 percent or 51-80 percent), 38 3-2 Distribution of adopter categories, 40 3-3 Women’s rate of participation in the markets for all vehicles and for PEVs, 42 3-4 Projected 2014 light-duty PEV volume in the 100 largest MSAs, 43 3-5 Worldwide growth of car sharing in terms of vehicles and members, 45 3-6 Clean Cities coalitions funded for community-readiness and planning for PEVs and PEV charging infrastructure, 58 3-7 Fleet sales for passenger vehicles for 2012 by fleet purchase agency, 59 4-1 Corporate Average Fuel Economy requirements by year, 68 4-2 Sources of revenue for the federal Highway Trust Fund, FY 2010, 69 4-3 U.S. annual light-duty fuel consumption and VMT, 70 4-4 Annual transportation-related taxes paid by Washington state drivers, 71 4-5 Historic and forecast gasoline-tax revenue for Washington state, FY 1990 to FY 2040, 73 4-6 PEV-specific measures for transportation funding, 74 5-1 PEV charging infrastructure categories, ranked by their likely importance to PEV deployment, with the most important, home charging, on the bottom, and the least important, interstate DC fast charging, at the top, 83 5-2 Vehicle locations throughout the week on the basis of data from the 2001 National Household Travel Survey, 84 6-1 U.S. electricity demand growth, 1950-2040, 99 6-2 Schematic of U.S. electric power delivery system, 99 6-3 Hourly demand for electricity at a substation in a residential distribution system, 101 6-4 Residential charging behavior in NES and PG&E service territories, as measured in the EV Project, 104 6-5 States that have regulations regarding who can own or operate a PEV charging station, 106 7-1 Japan’s clean energy vehicles promotion program, 119 7-2 U.S. HEV and PEV sales overlaid with U.S. gasoline prices, 120

TABLES S-1 Four Classes of Plug-in Electric Vehicles, 2 S-2 Effects of Charging Infrastructure by PEV Class and Entities Motivated to Install Infrastructure Categories, 5 2-1 Definitions and Examples of the Four Types of Plug-in Electric Vehicles, 20 2-2 Properties of Lithium-Ion Batteries in Four Plug-in Electric Vehicles on the U.S. Market, 24 2-3 Estimates of Dollars per Kilowatt-hour for a 25 kWh Battery, 27 2-4 Summary of Estimated Costs of Total Energy from Various Sources (2013 U.S. $/kWh), 28 3-1 Categories and Descriptions of Adopters, 40 3-2 Comparison of New BEV Buyers, PHEV Buyers, and ICE-Vehicle Buyers, 41 3-3 Comparison of All New-Vehicle Buyers to Buyers of Specific Plug-in Electric Vehicles, 41 3-4 Factors That Affect Adoption and Diffusion of Innovation, 46 3-5 Consumer Questions Related to Plug-in Electric Vehicle (PEV) Ownership, 51 3-6 Ratings of Dealer Knowledge about Various Topics, 53 3-7 Websites with Information on Plug-in Electric Vehicles, 55 3-8 Information Resources for Fleet Managers, 59 4-1 Factors Determining PEV Readiness and Organizations Involved, 67 4-2 Comparison of Unrealized Revenue from Battery Electric Vehicles and Plug-in Hybrid Electric Vehicles, 72 4-3 Types of Equity and Examples in the Transportation Tax System, 72 4-4 Variation in Residential Electric Permit Fees by City or State, 75 5-1 Effect of Charging-Infrastructure Categories on Mainstream PEV Owners by PEV Class, 85 5-2 Charging Patterns for Nissan Leafs and Chevrolet Volts, 87 5-3 Entities That Might Have an Incentive to Install Each Charging Infrastructure Category, 91 5-4 Costs of Installing Public DC Fast-Charging Stations for the West Coast Electric Highway Project, 93 6-1 Definitions, Advantages, and Disadvantages of Various Types of Electric Rates, 103 7-1 MSRPs and 5-year Cumulative Cost of Ownership for Selected Plug-in Electric Vehicles and Comparative Vehicles (dollars), 111 7-2 Incentives for Plug-in Electric Vehicles (PEVs) by Country and State, 118

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 15 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 16 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Summary

The plug-in electric vehicle (PEV) has a long history. THE COMMITTEE’S TASK In 1900, 28 percent of the passenger cars sold in the United States were electric, and about one-third of the cars on the The committee’s analysis was to be provided in two road in New York City, Boston, and Chicago were electric. reports—a short interim report and a final comprehensive re- Then, however, mass production of an inexpensive gasoline- port. The committee’s interim report, released in May 2013, powered vehicle, invention of the electric starter for the provided an initial discussion of infrastructure needs for gasoline vehicle, a supply of affordable gasoline, and de- PEVs, barriers to deploying the infrastructure, and possible velopment of the national highway system, which allowed roles for the federal government in overcoming the barriers. long-distance travel, led to the demise of those first PEVs. It did not offer any recommendations because the commit- In the 1970s and 1990s, interest in PEVs resurfaced, but the tee was still in the early stages of gathering data. The cur- vehicles simply could not compete with gasoline-powered rent report is the committee’s final comprehensive report that ones. In the last few years, interest in PEVs has been reig- addresses its full statement of task, which can be found in nited because of advances in battery and other technologies, Chapter 1. new federal standards for carbon-dioxide emissions and fuel This report focuses on light-duty vehicles (passenger economy, state zero-emission-vehicle requirements, and the cars and light-duty trucks) in the United States and restricts current administration’s goal of putting millions of alterna- its discussion to PEVs, which include battery electric vehi- tive-fuel vehicles on the road. People are also beginning to cles (BEVs) and plug-in hybrid electric vehicles (PHEVs).2 recognize the advantages of PEVs over conventional vehi- The common feature of these vehicles is that they can charge cles, such as lower operating costs, smoother operation, and their batteries by plugging into the electric grid. The distinc- better acceleration; the ability to fuel up at home; and zero tion between them is that BEVs operate solely on electricity tailpipe emissions when the vehicle operates solely on its stored in the battery (there is no other energy source), and battery. There are, however, barriers to PEV deployment, in- PHEVs have an internal-combustion engine (ICE) that can cluding the vehicle cost, the short all-electric driving range, supplement the electric power train or charge the battery dur- the long battery-charging time, uncertainties about battery ing a trip. PHEVs can use engines powered by various fuels, life, the few choices of vehicle models, and the need for a but this report focuses on those powered by gasoline because charging infrastructure to support PEVs whether at home, at they are the ones currently available in the United States. work, or in a public space. Moreover, many people are still The premise of the committee’s task is that there is a not aware of or do not fully understand the new technology. benefit to the United States if a higher fraction of miles is Given those recognized barriers to PEV deployment, Con- fueled by electricity rather than by petroleum. Two reasons gress asked the Department of Energy (DOE) to commission for this benefit are commonly assumed. First, a higher frac- a study by the National Academies to address market barri- tion of miles fueled by electricity would reduce the U.S. de- ers that are slowing the purchase of PEVs and hindering the pendence on petroleum. Second, a higher fraction of miles deployment of supporting infrastructure.1 Accordingly, the fueled by electricity would reduce carbon dioxide and other National Research Council (NRC), an arm of the National air pollutants emitted into the atmosphere. The committee Academies, appointed the Committee on Overcoming Bar- riers to Electric-Vehicle Deployment, which prepared this 2 BEVs and PHEVs need to be distinguished from conventional report. hybrid electric vehicles (HEVs), such as the Toyota Prius that was introduced in the late 1990s. HEVs do not plug into the electric grid but power their batteries from regenerative braking and an internal- combustion engine. They are not included in the PEV category and 1 See Consolidated Appropriations Act, 2012, P.L. 112-74, H. are not considered further in this report unless to make a compari- Rept. 112-331 (H.Rept. 112-118). son on some issue. 1

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 17 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

2 Overcoming Barriers to Deployment of Plug-in Electric Vehicles was not asked to research or evaluate the premise, but it did PLUG-IN ELECTRIC VEHICLES consider whether the premise was valid now and into the AND CHARGING TECHNOLOGIES future and asked if any recent developments might call the premise into question. Today, there are several makes and models of PEVs on First, a PEV uses no petroleum when it runs on elec- the market, and PEV sales reached about 0.76 percent of the tricity. Furthermore, the electricity that fuels the vehicle is light-duty sales in the United States by the close of 2014. Be- generated using essentially no petroleum; in 2013, less than cause the obstacles to consumer adoption and the charging 0.7 percent of the U.S. grid electricity was produced from infrastructure requirements depend on PEV type, the com- petroleum. Thus, PEVs advance the long-term objective of mittee used the all-electric range (AER) of the vehicles to U.S. energy independence and security. Second, on average, distinguish four PEV classes (see Table S-1). Several impor- a PEV fueled by electricity is now responsible for less green- tant points regarding the PEV classes should be highlighted. house gases (GHGs) per mile than an ICE vehicle3 or a hy- First, the Tesla Model S clearly demonstrates the possibility brid electric vehicle (HEV). PEVs will make further reduc- of producing a long-range BEV that has been recognized tions in GHG emissions as the U.S. electric grid changes to as a high-performing vehicle. Second, limited-range BEVs lower carbon sources for its electricity. Therefore, the com- are the only type of PEV that have a substantial range limi- mittee concludes that the premise for the task—that there is tation. Although they are not practical for trips that would an advantage to the United States if a higher fraction of miles require more than one fast charge given the substantial re- driven here are fueled by electricity from the U.S. electric fueling time required, their ranges are more than sufficient grid—is valid now and becomes even more valid each year for the average daily travel needs of the majority of U.S. that the United States continues to reduce the GHGs that it drivers. Third, the range-extended PHEV has a total range produces in generating electricity. A more detailed discus- that is comparable to that of a conventional vehicle because sion of the committee’s analysis of the near-term and long- of the onboard ICE, and the typical AER is comparable to or term impacts of PEV deployment on petroleum consumption larger than the average U.S. daily travel distance. The frac- and GHG emissions is provided in Chapter 1 of this report. tion of miles traveled by electricity depends on how willing and able a driver is to recharge the battery during a trip lon- Recommendation: As the United States encourages the ger than the AER. Fourth, minimal PHEVs with AERs much adoption of PEVs, it should continue to pursue in parallel the shorter than the average daily driving distance in the United production of U.S. electricity from increasingly lower carbon States are essentially HEVs. sources. There are three options for charging the high-energy batteries in PEVs.4 First, AC level 1 uses a 120 V circuit and provides about 4-5 miles of electric range per hour of 3 For this report, ICE vehicle or conventional vehicle refers to a light-duty vehicle that obtains all of its propulsion from an internal- 4 A fourth option might be considered wireless charging, but this combustion engine. option is not widely used today.

TABLE S-1 Four Classes of Plug-in Electric Vehicles PEV Class Description Example (Rangea) Long-range BEV Can travel hundreds of miles on a single battery charge 2014 Tesla Model S (AER = 265 miles) and then be refueled in a time that is much shorter than the additional driving time that the refueling allows.

Limited-range BEV Is made more affordable than the long-range BEV by 2014 Nissan Leaf (AER = 84 miles) reducing the size of the high-energy battery. Its limited 2014 Ford Focus Electric (AER = 76 miles) range can more than suffice for many commuters, but it is impractical for long trips.

Range-extended PHEV Typically, operates as a zero-emission vehicle until its battery 2014 Chevrolet Volt (AER = 38 miles; is depleted, whereupon an ICE turns on to extend its range. total range = 380 miles)

Minimal PHEV Its small battery can be charged from the grid, but its AER 2014 Toyota Plug-in Prius (AER = 6-11 miles; is much less than the average daily U.S. driving distance. total range = 540 miles) a The AERs noted are average values estimated by the U.S. Environmental Protection Agency. Total ranges are provided for PHEVs; the AER is the total range for BEVs. NOTE: AER, all-electric range; BEV, battery electric vehicle; ICE, internal-combustion engine; PEV, plug-in electric vehicle; PHEV, plug-in hybrid electric vehicle.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 18 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Summary 3 charging. It is considered too slow to be the primary charg- One strategy for dealing with market complexity has been ing method for fully depleted batteries of PEVs that have to identify a narrow market segment for which the new tech- large batteries because charging times would be longer than nology offers a compelling reason to buy. Offering a compel- the time a vehicle is normally parked at home or the work- ling value proposition specifically targeted to meet the needs place. Second, AC level 2 uses a 240 V, split-phase ac circuit of a narrow market segment rather than the broad mass market like those used by electric dryers, electric stoves or ovens, gives the technology a greater chance to dominate in that key and large air conditioners; it provides about 10-20 miles of market segment. Then, the momentum gained in the initial electric range per hour of charging depending on how much market segment can be used more efficiently and effectively current the vehicle is allowed to draw. Third, DC fast charg- to drive sales in related, adjacent segments. That approach ap- ing is an option available only to BEVs today and uses high- pears reasonable for PEVs because the PEV market has been voltage circuits to charge the battery much more rapidly. characterized by strong regional patterns that reflect such attri- DC fast charging is generally not an option for residential butes as expensive gasoline; favorable demographics, values, charging given the high-power circuits required. In the Unit- and lifestyles; a regulatory environment favorable to PEVs; ed States, there is one standard plug for the AC level 1 and and an existing or at least readily deployable infrastructure. AC level 2 chargers, but there are at least three incompatible The purchase of a new vehicle is typically a lengthy pro- plugs and communication protocols being used for DC fast cess that often involves substantial research and is strongly charging. Plug and protocol incompatibility is a barrier to affected by consumer perceptions. In evaluating the pur- PEV adoption insofar as it prevents all PEVs from being able chase process for PEVs specifically, the committee identified to charge at any fast-charging station. several barriers—in addition to the cost differences between PEVs and ICE vehicles—that affect consumer perceptions Recommendation: The federal government and proactive and their decision process and ultimately (negatively) their states should use their incentives and regulatory powers to purchase decisions. The barriers include the limited variety (1) eliminate the proliferation of plugs and communication of PEVs available; misunderstandings concerning the range protocols for DC fast chargers and (2) ensure that all PEV of the various PEVs; difficulties in understanding electricity drivers can charge their vehicles and pay at all public charg- consumption, calculating fuel costs, and determining charging stations using a universally accepted payment method ing infrastructure needs; complexities of installing home just as any ICE vehicle can be fueled at any gasoline station. charging; difficulties in determining the greenness of the vehicle; lack of information on incentives; and lack of knowl- UNDERSTANDING THE MARKET DEVELOPMENT edge of unique PEV benefits. Collectively, the identified AND CUSTOMER PURCHASE PROCESS FOR barriers indicate that consumer awareness and knowledge PLUG-IN ELECTRIC VEHICLES of PEV offerings, incentives, and features are not as great as needed to make fully informed decisions about whether Developers of new technologies, such as PEVs, face to purchase a PEV. Furthermore, many factors contribute to challenges in developing a market and motivating consum- consumer uncertainty and doubt about the viability of PEVs ers to purchase or use their products. Incumbent technolo- and create a perceptual hurdle that negatively affects PEV gies—in this case, ICE vehicles—can be difficult to unseat; purchases. Together, the barriers emphasize the need for they have years of production and design experience, which better consumer information and education that can answer make their production costs lower than those of emerging all their questions. Consumers have traditionally relied on technologies and thus more affordable. The necessary infra- dealers to provide vehicle information; however, in spite of structure, including the ubiquitous presence of gasoline and education efforts by some manufacturers, dealer knowledge service stations across the United States, is well-developed. of PEVs has been uneven and often insufficient to address Consumers know the attributes and features to compare to consumer questions and concerns. The committee does ac- evaluate their ICE-vehicle choices, and they are accustomed knowledge, however, that even well-informed consumers to buying, driving, and fueling these vehicles. Indeed, one of might not buy a PEV because it does not meet some of their the main challenges to the success of the PEV market is that basic requirements for a vehicle (that is, consumer informa- people are so accustomed to ICE vehicles. tion and education cannot overcome the absence of features Accordingly, adoption and diffusion of PEVs is likely to desired by a consumer). be a long-term, complex process. Even modest market penetration could take many years. Furthermore, market penetra- Recommendation: To provide accurate consumer information rates will likely be a function not only of the product it- tion and awareness, the federal government should make use self but also of the entire industry ecosystem. Hence, product of its Ad Council program, particularly in key geographic technologies (such as low-cost batteries), downstream infra- markets, to provide accurate information about federal tax structure (such as dealers and repair facilities), and comple- credits and other incentives, the value proposition for PEV mentary infrastructure (such as charging stations) will need ownership, and who could usefully own a PEV. to be developed simultaneously.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 19 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

4 Overcoming Barriers to Deployment of Plug-in Electric Vehicles GOVERNMENT SUPPORT FOR DEPLOYMENT costs have been introduced into the installation process. Be- OF PLUG-IN ELECTRIC VEHICLES cause most charging will occur at home, PEV deployment could be seriously impeded if the buyers must bear high The federal government can play a substantive role in permit and installation costs and experience delay in the ac- encouraging PEV deployment by supporting research that tivation of their home chargers. Accordingly, clarity, predict- has the potential to remove barriers. Specifically, investment ability, and speed are needed in the permitting and approval in battery research is critical for producing lower cost, higher process for installation of home and public charging stations. performing batteries. Improved battery technology will lower vehicle cost, increase the all-electric range, or both, and those Recommendation: Local governments should streamline per- improvements will likely lead to increased PEV deployment. mitting and adopt building codes that require new construction Furthermore, research is needed to understand the relationship to be capable of supporting future charging installations. between charging infrastructure availability and PEV adoption and use. Specifically, research should be conducted to de- CHARGING INFRASTRUCTURE FOR termine how much public infrastructure is needed and where it PLUG-IN ELECTRIC VEHICLES should be sited to induce PEV adoption and to encourage PEV owners to optimize their vehicle use. That research is espe- PEV deployment and the fraction of vehicle miles fu- cially critical if the federal government is allocating resources eled by electricity (eVMT) critically depend on the charging to fund public infrastructure deployment. infrastructure. For its analysis, the committee categorized the infrastructure by location (home, workplace, intracity, Recommendation: The federal government should continue intercity, and interstate) and power (AC level 1, AC level 2, to sponsor fundamental and applied research to facilitate and and DC fast charging), evaluated it from the perspective of expedite the development of lower cost, higher performing ve- the PEV classes defined in Table S-1, and determined which hicle batteries. Stable funding is critical and should focus on entities might have a motivation to install which category of improving energy density and addressing durability and safety. charging infrastructure. The results of the committee’s analysis are summarized in Table S-2. The table reflects the rela- Recommendation: The federal government should fund re- tive importance of each infrastructure category as assessed search to understand the role of public charging infrastructure by the committee, with home listed first (most important) (as compared with home and workplace charging) in encour- and interstate listed last (least important). aging PEV adoption and use. Several points should be made for the various infrastructure categories. First, home charging is a virtual necessity for The successful deployment of PEVs will involve many all PEV classes given that the vehicle is typically parked at a entities, including federal, state, and local governments. One residence for the longest portion of the day. Accordingly, the potential barrier for PEV adoption that is solely within gov- home is (and will likely remain) the most important location ernment control is taxation of PEVs and, in particular, taxa- for charging infrastructure, and homeowners who own PEVs tion for the purpose of recovering the costs of maintaining, have a clear incentive to install home charging. Residences repairing, and improving roadways. In the United States, that do not have access to a dedicated parking spot or one fuel taxes have been used to finance transportation budgets. with access to electricity clearly have challenges to over- Because BEVs use no gasoline and PHEVs use much less come to make PEV ownership practical for them. gasoline than ICE vehicles, there is the belief that PEV own- Second, charging at workplaces offers an important op- ers pay nothing to support transportation infrastructure and portunity to encourage PEV adoption and increase eVMT. should be taxed or charged a special fee. However, PEV Specifically, it could double the daily travel distance that is owners pay taxes and fees other than fuel taxes that support fueled by electricity if combined with home charging and transportation budgets. Furthermore, the fiscal impact at the could in principle make possible the use of limited-range present time and likely over the next decade of not collecting BEVs when no home charging is available. Some businesses fuel taxes from PEV owners is negligible, especially com- appear to be motivated to provide workplace charging as a pared with the impact of high-mileage vehicles that are being means to attract and retain employees or to brand the compa- produced to meet fuel-economy standards. ny with a green image. However, one concern is that utilities could impose demand charges if the businesses exceed their Recommendation: Federal and state governments should maximum power-demand thresholds; such charges could be adopt a PEV innovation policy where PEVs remain free from substantial. Another concern is the IRS requirement for busi- special roadway or registration surcharges for a limited time to nesses to assess the value of the charging and report it as encourage their adoption. imputed income.

Some federal and state permitting processes have been Recommendation: Local governments should engage with ill-suited for the simple installation of some PEV charging and encourage workplaces to consider investments in charging infrastructure. As a result, unnecessary permit burdens and infrastructure and provide information about best practices.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 20 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Summary 5

TABLE S-2 Effects of Charging Infrastructure by PEV Class and Entities Motivated to Install Infrastructure Categoriesa Infrastructure Categoryb PEV Class Effect of Infrastructure on Mainstream PEV Owner Who Has an Incentive to Install? Home Long-range BEV Virtual necessity Vehicle Owner, Utility AC levels 1 and 2 Limited-range BEV Virtual necessity Range-extended PHEV Virtual necessity Minimal PHEV Virtual necessity

Workplace Long-range BEV Range extension, expands market Business Owner, Utility AC levels 1 and 2 Limited-range BEV Range extension, expands market Range-extended PHEV Increases eVMT and value proposition; expands market Minimal PHEV Increases eVMT and value proposition; expands market

Intracityc Long-range BEV Not necessary Utility, Retailer, Charging Provider, AC levels 1 and 2 Limited-range BEV Range extension, increases confidence Vehicle Manufacturer Range-extended PHEV Increases eVMT and value proposition Minimal PHEV Increases eVMT and value proposition

Intracityc Long-range BEV Not necessary Utility, Charging Provider, Vehicle DC fast charge Limited-range BEV Range extension, increases confidence Manufacturer, Government Range-extended PHEV NA – not equipped Minimal PHEV NA – not equipped

Intercityc Long-range BEV Range extension, expands market Vehicle Manufacturer, Government DC fast charge Limited-range BEV 2 × Range extension, increases confidence Range-extended PHEV NA – not equipped Minimal PHEV NA – not equipped

Interstate Long-range BEV Range extension, expands market Vehicle Manufacturer, Government DC fast charge Limited-range BEV Not practical for long trips Range extended PHEV NA – not equipped Minimal PHEV NA – not equipped a Assumptions for analysis are that electricity costs would be cheaper than gasoline costs, that away-from-home charging would generally cost as much as or more than home charging, that people would not plan to change their mobility needs to acquire a PEV, and that there would be no disruptive changes to current PEV performance and only incremental improvements in battery capacity over time. b The term intercity refers to travel over distances less than twice the range of limited-range BEVs, and the term interstate refers to travel over longer distances. c It is possible that these infrastructure categories could expand the market for the various types of PEVs as appropriate, but that link is more tenuous than the cases noted in the table for other infrastructure categories. NOTE: AC, alternating current; BEV, battery electric vehicle; DC, direct current; eVMT, electric vehicle miles traveled; NA, not applicable; PEV, plug-in electric vehicle; PHEV, plug-in hybrid electric vehicle.

Third, public charging infrastructure has the potential to Recommendation: The federal government should refrain provide range confidence and extend the range for limited- from additional direct investment in the installation of public range BEV drivers, allow long-distance travel for long-range charging infrastructure pending an evaluation of the relation- BEV drivers, and increase eVMT and the value proposition ship between the availability of public charging and PEV adop- for PHEV drivers. However, fundamental questions that tion or use. need to be answered are how much and what type of public charging infrastructure is needed and where should it be IMPLICATIONS OF PLUG-IN ELECTRIC VEHI- located? Furthermore, although the committee has identi- CLES FOR THE ELECTRICITY SECTOR fied several entities that might be motivated to install public charging infrastructure, it could identify only two entities— An important concern raised by the public and policy BEV manufacturers and utilities—that might have an attrac- makers pertains to the capability of electric utilities to pro- tive business case for absorbing the full capital costs of in- vide for PEV charging. At the current time, PEV charging vestments in public charging infrastructure. The government requirements account for about 0.02 percent of the energy might decide that providing public charging infrastructure produced and consumed in the continental United States. serves a public good when others do not have a business case Were the PEV fleet to reach as high as 20 percent of private or incentive to do so. vehicles, the estimated impact would still be only 5 percent

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 21 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

6 Overcoming Barriers to Deployment of Plug-in Electric Vehicles of today’s electric production. Accordingly, PEV deploy- popular PEVs before costs can be substantially reduced. ment is not constrained by the transmission system or the Thus, the deployment of PEVs might be at risk unless the generation capacity. Although some capital investment in (or federal government extends manufacturer or consumer in- upgrades to) the distribution infrastructure might be required centives, at least temporarily. in areas where there is high, concentrated PEV deployment, Regulatory requirements and incentives for manufactur- PEV charging is expected to have a negligible effect on the ers and consumers have been introduced over the past few distribution system at the anticipated rates of PEV adoption. years by states and the federal government to encourage Thus, the constraints on PEV adoption that could arise PEV production and deployment. Most manufacturer incen- from the electricity sector are more likely to be economic tives and mandates are contained in the federal Corporate rather than physical or technical. Potential impediments to Average Fuel Economy standards, the federal GHG emission PEV adoption include (1) high electricity costs that reduce standards, and state zero-emission-vehicle (ZEV) programs. the financial benefit of PEV ownership, (2) regional differ- Most consumer incentive programs have involved purchase ences in electricity costs that add confusion and prevent a incentives in the form of tax credits, tax rebates, or tax ex- uniform explanation of the economic benefits of PEV own- emptions. However, states have also used ownership incen- ership, (3) residential electric rate structures that provide no tives (such as exemptions from or reductions in registration incentive to charge the vehicle at the optimal time for the taxes or fees and vehicle inspections) and use incentives utility, and (4) high costs for commercial and industrial cus- (such as exemptions from motor fuel taxes, reduced road- tomers if demand charges are incurred as noted above. The way taxes or tolls, and discounted or free PEV charging or committee notes that state jurisdiction over retail electricity parking). Some states have also offered nonfinancial incen- rates constrains the federal role in directing the electricity tives that allow access to restricted lanes, such as bus-only, sector to foster PEV growth. high-occupancy-vehicle, and high-occupancy-toll lanes. In- centives have also been provided to install charging stations, Recommendation: To ensure that adopters of PEVs have in- the availability of which might also influence people’s will- centives to charge vehicles at times when the cost of supply- ingness to purchase PEVs. ing energy is low, the federal government should propose that On the basis of the committee’s analysis, several points state regulatory commissions offer PEV owners the option of should be highlighted. First, existing federal and state regu- purchasing electricity under time-of-use or real-time pricing. latory programs for fuel-economy and emissions have been effective at stimulating manufacturers to produce some INCENTIVES FOR THE DEPLOYMENT PEVs, and sale of credits from these programs between OF PLUG-IN ELECTRIC VEHICLES manufacturers has also provided an important incentive for PEV manufacturers to price PEVs more attractively. The One of the most important issues concerning PEV de- committee emphasizes that the state ZEV requirements have ployment is determining what, if any, incentives are needed been particularly effective at increasing PEV production and to encourage PEV adoption. Determining the need for in- adoption. Second, the effectiveness of the federal income centives is difficult because little is yet known about the ef- tax credit at motivating people to purchase PEVs would be fectiveness of PEV incentive programs. However, two fac- enhanced by converting it into a rebate at the point of sale. tors to consider are vehicle price and cost of ownership. To Third, state and local governments offer a variety of finan- examine those factors, the committee considered sales and cial and nonfinancial incentives, but there appears to bea consumer survey data and compared manufacturer suggest- lack of research to indicate which incentives might be the ed retail prices (MSRPs) on selected PEVs, HEVs, and ICE most effective at encouraging PEV adoption. Fourth, the vehicles. The committee found that although sales data and many state and local incentives that differ in monetary val- consumer survey data are difficult to interpret, they are con- ue, restrictions, and calculation methods make it challenging sistent with the view that price is a barrier to some buyers but to educate consumers on the incentives that are available to that others might be rejecting PEVs for other reasons. Com- them and emphasize the need for a clear, up-to-date source parisons of MSRPs and cumulative ownership costs that in- of information for consumers. Fifth, the overall international corporate current federal tax credits provide mixed evidence experience appears to suggest that substantial financial in- on whether price is an obstacle to PEV adoption. However, centives are effective in motivating consumers to buy PEVs. in the absence of tax credits or other subsidies, comparisons at today’s MSRPs would be unfavorable to PEVs. Recommendation: Federal financial incentives to purchase Another factor to consider is the possibility of declines PEVs should continue to be provided beyond the current in production costs for PEVs so that manufacturers can price production volume limit as manufacturers and consumers them attractively in comparison with conventional vehicles. experiment with and learn about the new technology. The The decline over time in PEV production costs, however, is federal government should re-evaluate the case for incen- likely to occur gradually, and existing quotas of federal tax tives after a suitable period, such as 5 years. Its re-evaluation credits could be exhausted for manufacturers of relatively should consider advancements in vehicle technology and

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 22 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Summary 7 progress in reducing production costs, total costs of owner- development to improve and lower the cost of batteries (and ship, and emissions of PEVs, HEVs, and ICE vehicles. electric-drive technologies) for PEVs represents a technology-push strategy that complements the market-pull strategy Recommendation: Given the research on effectiveness of represented by the federal financial purchase incentives that purchase incentives, the federal government should consider lower the barrier to market adoption. A significant body of converting the tax credit to a point-of-sale rebate. research, however, demonstrates that having the right technology (with a compelling value proposition) is still insuffi- Recommendation: Given the sparse research on incentives cient to achieve success in the market. That technology must other than financial purchase incentives, research should be be complemented with a planned strategy to create market conducted on the variety of consumer incentives that are (or awareness and to overcome customer fear, uncertainty, and have been) offered by states and local governments to deter- doubt about the technology. mine which, if any, have proven effective in promoting PEV Equally important to recognize is a recommendation deployment. that the committee does not make. The committee does not at this point recommend additional direct federal investment CONCLUDING REMARKS in the installation of public charging infrastructure until the relationship between infrastructure availability and PEV The committee provides a number of recommendations adoption and use is assessed. That statement does not mean throughout this report and highlights several of the most im- or should not be construed to mean that no federal invest- portant in the summary. However, two points should be fur- ment or additional public infrastructure is needed. Other en- ther emphasized. First, vehicle cost is a substantial barrier to tities—including vehicle manufacturers, utilities, and other PEV deployment. As noted above and discussed in detail in private companies—are actively deploying and planning to Chapter 7, without the federal financial purchase incentives, deploy public infrastructure and have concluded that addi- PEVs are not currently cost-competitive with ICE vehicles tional public infrastructure is needed. However, the commit- on the basis of either purchase price or cumulative cost of tee is recommending research to help determine the relation- ownership. Therefore, one of the most important commit- ship between charging infrastructure availability and PEV tee recommendations is continuing the federal financial adoption and use. Although some data have been collected purchase incentives and re-evaluating them after a suitable through various projects, the data-collection efforts were not period. Second, developing lower cost, better performing designed to understand that fundamental relationship, and batteries is essential for reducing vehicle cost because it is the committee cautions against extrapolating findings on the high-energy batteries that are primarily responsible for the first adopters to the mainstream market. Given the strain the cost differential between PEVs and ICE vehicles. It is on federal resources, the suggested research should help to therefore important that the federal government continue to ensure that limited federal funds are spent so that they will fund battery research at least at current levels. Technology have the greatest impact.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 23 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Introduction

Plug-in electric vehicles (PEVs) that derive all or some vehicle (the Model T), the invention of an electric starter for of their propulsion from an external electricity source have the gasoline vehicle (which eliminated the need for a hand- received critical attention in recent years. They are espe- crank), a supply of affordable gasoline, and the development cially attractive because they have the potential to reduce of the national highway system, which allowed long-distance greenhouse gas (GHG) emissions and to decrease petroleum travel (Schiffer et al. 1994). In the 1970s, interest in PEVs consumption substantially, given that light-duty vehicles ac- resurfaced with the Arab oil embargo and the emerging en- count for nearly half of the petroleum consumption in the vironmental and energy security concerns. Over the next few United States today and that electricity is not typically gener- decades, interest in PEVs waxed and waned as gasoline prices ated from petroleum (EIA 2014). Globally, the demand for remained roughly constant. In the 1990s, interest in PEVs was PEVs is growing, and some countries see them as an impor- revived by California’s zero-emission-vehicle (ZEV) policies tant element of their long-term strategy to meet environmen- but lagged again primarily because battery technology was tal, economic, and energy-security goals. Although they hold not as advanced as it is today. Recent advances in battery and great promise, there are also many barriers to their penetra- other technologies, new federal standards for carbon-dioxide tion into the mainstream market. Some are technical, such as (CO2) emissions and fuel economy, state requirements for ze- the capabilities of current battery technologies that restrict ro-emission vehicles, and the current administration’s goal of their electric driving range and increase their purchase price putting millions of alternative-fuel vehicles on the road have compared with conventional vehicles; others are related to reignited interest in PEVs. consumer behavior and attitudes; and still others are related Recent incentives to increase the number of PEVs on the to developing an infrastructure to support charging of the road began with the Emergency Economic Stabilization Act vehicles and addressing possible effects of the new charg- of 2008, which provided a $2,500 to $7,500 tax credit for the ing infrastructure on the electric grid. Given the growing purchase of PEVs (Public Law 110-343 §205). The American concerns surrounding the perceived barriers, Congress in its Recovery and Reinvestment Act of 2009 (Public Law 111-5 2012 appropriations for the Department of Energy (DOE) re- §1141) increased incentives for PEVs by expanding the list of quested that DOE commission a study by the National Acad- vehicles that are eligible for a tax credit. It also appropriated emies to identify market barriers that are slowing the pur- $2 billion in grants for development of electric-vehicle bat- chase of PEVs and hindering the deployment of supporting teries and related components (DOE 2009) and $2.4 billion infrastructure.1 Accordingly, the National Research Council in loans for electric-vehicle manufacturing facilities (DOE (NRC), which is a part of the National Academies, appointed 2011). Along with private investors, DOE has invested $400 the Committee on Overcoming Barriers to Electric-Vehicle million to support infrastructure development, including dem- Deployment, which prepared this final report. onstration projects involving 13,000 PEVs and 22,000 public and private charging points in 20 U.S. cities (DOE 2011). Fur- HISTORICAL AND POLICY CONTEXT thermore, the DOE Office of Energy Efficiency and Renew- able Energy (DOE 2013a) and several national laboratories, The PEV is not a new invention of the twenty-first cen- including Argonne National Laboratory (ANL 2011, 2012, tury. In 1900, 28 percent of the passenger vehicles sold in the 2013) and the National Renewable Energy Laboratory (NREL United States were electric, and about one-third of the vehi- 2013), are conducting substantial research and development cles on the road in New York City, Boston, and Chicago were on electric-drive technologies for PEVs (NRC 2013a). electric (Schiffer et al. 1994). The demise of PEVs resulted Various state-level efforts—such as consumer incen- from the mass production of an inexpensive gasoline-powered tives that include tax credits for vehicle purchase, access to carpool lanes, free public parking, and emission-inspection 1 See Consolidated Appropriations Act, 2012, P.L. 112-74, H. exemptions—are aimed at increasing the number of PEVs Rept. 112-331 (H.Rept. 112-118). 8

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 24 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Introduction 9 on the road (DOE 2013b). Other efforts, such as reimburse- THE PLUG-IN ELECTRIC VEHICLE ments and tax incentives for purchasing or leasing charg- AND CURRENT SALES ing equipment and low-cost loans for installation projects, are aimed at building the charging infrastructure (DOE This report focuses on light-duty vehicles (passenger 2013b). California's ZEV program is particularly important cars and light-duty trucks) in the United States and restricts because of the size of the California motor-vehicle market. its discussion to PEVs, which include battery electric vehi- Each motor-vehicle manufacturer in the state is required to cles (BEVs)2 and plug-in hybrid electric vehicles (PHEVs).3 sell at least a minimum percentage of ZEVs (vehicles that The common feature of these vehicles is that they can charge produce no exhaust emissions of any criteria pollutant) and their batteries by plugging into the electric grid. The distinc- transitional ZEVs (vehicles that can travel some minimum tion between them is that BEVs operate solely on electricity distance solely on a ZEV fuel, such as electricity) (13 CCR stored in the battery (there is no other power source), and §1962.1 [2013]). Nine states—Connecticut, Maine, Mary- PHEVs have an internal-combustion engine (ICE) that can land, Massachusetts, New Jersey, New York, Rhode Island, supplement the electric power train,4,5 PEVs are often de- Vermont, and Oregon—have also adopted the California fined by the number of electric miles that they can drive. ZEV program as part of their plans to meet federal ambient A BEV that can drive 100 miles on one battery charge is air quality standards. designated as a BEV100; likewise, a PHEV that can drive The policies that promote early PEV deployment are 40 miles on one battery charge is designated as a PHEV40. aimed at benefits beyond near-term reductions in petroleum A more detailed discussion of PEV technology is provided in consumption and pollutant emissions. The strategy is to Chapter 2 of this report. speed the long-term process of converting the motor-vehicle Although a few makes and models of PEVs were avail- fleet to alternative energy sources by exposing consumers able in the mid-1990s (for example, the General Motors EV1 now to PEVs, by encouraging governments and service pro- and the Honda EV+, released in 1997; see UCS 2014), many viders to plan for infrastructure, and by encouraging the mo- consider the December 2010 introduction of the Nissan Leaf tor-vehicle industry to experiment with product design and and Chevrolet Volt—the first mass-produced PEVs—to be marketing. Gaining a major market share for PEVs will likely the start of the viable commercial market for PEVs. Every require advances in technology to reduce cost and improve few months, new PEVs have been added to the U.S. market, performance, but the premise of the early deployment efforts including a long-range BEV (the Tesla Model S); limited- is that market development and technology development that range BEVs (such as the Daimler Smart EV and the BMW proceed in parallel will lead to earlier mass adoption than if i3); range-extended PHEVs (such as the Ford Fusion Energi technology advances are required before beginning market and the Ford C-Max Energi); and minimal PHEVs (such as development. The early deployment efforts also might speed the Toyota Plug-In Prius).6 Several manufacturers are also technology breakthroughs by maintaining visibility and in- selling limited-volume BEVs, including the Ford Focus EV, terest in PEVs. The risk entailed by this strategy is that if the Honda Fit EV, the Fiat 500e, and the Chevrolet Spark PEV promotion efforts are premature relative to the develop- to meet fuel-efficiency and ZEV regulatory requirements. ment of the technology, the costs of the promotion will have In addition, a number of PEVs are not yet available in the had little benefit in the form of market development. United States, notably the Mitsubishi Outlander PHEV and a The motivation for pursuing PEV-deployment policies number of Renault BEVs and Volkswagen PHEVs. beyond their near-term benefit can be understood from the Figures 1-1 and 1-2 show monthly sales for BEVs findings of another NRC report, Transitions to Alternative and PHEVs, respectively. PEV sales in the United States Vehicles and Fuels. The committee that prepared that report were about 56,000 units in 2012, 96,000 units in 2013, and was asked to assess a range of vehicle technology options and to suggest strategies for attaining petroleum consump- 2 The term all-electric vehicle (AEV) is sometimes used instead tion and GHG reduction targets of 50 to 80 percent by the of BEV. 3 2030-2050 timeframe (NRC 2013b). An important finding BEVs and PHEVs need to be distinguished from conventional of that report is that major policy initiatives—such as tax hybrid electric vehicles (HEVs), such as the Toyota Prius, which was introduced in the late 1990s. HEVs do not plug into the elec- incentives, subsidies, or regulations—are required to obtain tric grid but power their batteries from regenerative braking and an such large-scale reductions. That conclusion is relevant for internal-combustion engine. They are not included in the PEV cat- the current study because it provides context as to why fed- egory and are not considered further in this report except to make a eral policy (or an NRC study) might focus on barriers. If comparison on some issue. 4 policy makers decide that such major reductions in petro- Several design architectures are available for PHEVs, and, depending on the design, the engine may be used to drive the vehicle leum consumption or GHG emissions are required to meet directly or act as a generator to recharge the battery or both. environmental and other goals, an understanding of the bar- 5 PHEVs can use engines powered by various fuels. This report, riers and the strategies that are needed to overcome them will however, focuses on PHEV engines that are powered by gasoline be required. because they are the ones currently available in the U.S. market. 6 PEV designations are discussed in detail in Chapter 2 of this report.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 25 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

10 Overcoming Barriers to Deployment of Plug-in Electric Vehicles 120,000 units in 2014 (Inside EVs 2015). Total U.S. vehi- Washington, HEVs comprise 10 percent of new passenger cle sales in 2014 were nearly 16.5 million, a record year in vehicle sales (see Chapter 3 for a discussion of factors that which people were replacing their vehicles after not buying affect vehicle preferences). Figure 1-4 compares HEV and during the recession (Woodall and Klayman 2015). PEV sales over their first 34 months of having been intro- In the U.S. market, PEV sales increased from 0.62 per- duced to the U.S. market and indicates that PEVs are pen- cent in 2013 to 0.76 percent in 2014 (Cobb 2014, 2015); total etrating the market faster than HEVs. accumulated sales in the United States were about 291,000 The California market has been particularly important vehicles by the close of 2014 (Inside EVs 2015). To put the and accounts for over one-third of annual PEV sales. At the U.S. sales data in perspective, Figure 1-3 shows that North close of 2014, PEV sales in California were 3.2 percent of America accounted for almost half of the world PEV sales in new light-duty vehicle sales and 5.2 percent of new passen- 2013. Worldwide sales of PEVs were about 132,000 in 2012, ger vehicles (CNCDA 2015). California has a long history of 213,000 in 2013, and 318,000 in 2014 (Pontes 2015). PEV strong sales for new vehicle technologies, especially HEVs sales have not yet been reported for some countries so this as noted above. California is a favorable market for PEVs be- number could increase slightly. cause it has many wealthy buyers of new technology, broad The rate of market growth over the past 3 years has al- social support for PEVs in light of its history of air pollu- most doubled each year, but sales started at a very low level. tion, an active regulatory regime with purchase incentives By way of comparison, hybrid electric vehicles (HEVs) were and mandates for reducing carbon emissions and increasing introduced in 1997 in Japan and in 1999 in the United States. PEV sales, and favorable weather that is easy on battery life Although HEVs have been more successful in Japan than in and on charge available for vehicle miles. Furthermore, Cali- the United States—now at 20 percent of the total Japanese fornia has had a consistent, long-standing effort to provide light-duty vehicle market (Nikkei Asian Review 2012) and basic Web-based and printed information resources on low- over 50 percent of Toyota’s Japanese vehicle sales (Toyota and zero-emission vehicles and to hold some ride-and-drive 2014)—it took 13 years for HEVs to exceed 3 percent of an- events. Those activities have likely contributed to greater nual new light-duty vehicle sales in the United States (Cobb public awareness of PEVs. 2013). However, in certain markets, such as California and

FIGURE 1-1 U.S. BEV monthly sales data from 2010 to 2014. NOTE: BEV, battery electric vehicle. SOURCE: Based on data from Inside EVs (2015).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 26 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Introduction 11

FIGURE 1-2 U.S. PHEV monthly sales data from 2010 to 2014. NOTE: PHEV, plug-in hybrid electric vehicle. SOURCE: Based on data from Inside EVs (2015).

FIGURE 1-3 World PEV sales in 2012, 2013, and 2014. NOTE: PEV, plug-in electric vehicle. SOURCE: Based on data from Pontes (2015).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 27 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

12 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 1-4 The rate of PEV market growth in its first 34 months superimposed on the rate of HEV market growth during its first 34 months. NOTE: HEV, hybrid electric vehicle; PEV, plug-in electric vehicle. SOURCE: DOE (2014).

As shown in Figure 1-5, other strong PEV markets are the life of conventional brakes and thus reduce brake repair Washington, Oregon, Georgia, Maryland, Vermont, and Ha- and replacement costs. On a large scale, PEVs offer the po- waii. Those markets have also been driven primarily by social tential to reduce petroleum consumption and improve urban sentiment (an environmentally friendly population base), fi- air quality; the degree to which PEVs affect pollutant emis- nancial incentives, and regulatory mandates for reducing car- sions will depend on how the electricity that fuels a vehicle bon emissions. is generated, the degree to which charging of the vehicle is managed, and the degree to which emissions from power- Finding: HEV adoption, which entailed fewer technology generation sources are controlled (Peterson et al. 2011; see changes than PEVs, required 13 years to exceed 3 percent of further discussion below). PEVs might also act as an enabler annual new light-duty vehicle sales in the United States. for renewable power generation by providing storage or rapid demand response through smart-grid applications. Finding: PEVs have had higher sales than HEVs within the PEVs, however, also have important trade-offs. Current first 34 months of their introduction into the market, although limitations in battery technology result in restricted electric- the higher sales for PEVs could be the result of the various driving range, high battery cost, long battery-charging time, incentives that have been offered. and uncertain battery life. Concerns about battery safety, depending on the chemistry and energy density of the battery, PLUG-IN ELECTRIC VEHICLES: have also arisen. PEVs have higher upfront costs than their BENEFITS AND TRADE-OFFS conventional-vehicle counterparts and are available in only a few vehicle models. There is also a need to install a charging PEVs offer several benefits over conventional vehicles. infrastructure to support PEVs whether at home, at work, or The most obvious for the owner are lower operating cost, less in a public space. Beyond the technical and economic barri- interior noise and vibration from the power train, often bet- ers, people are not typically familiar with the capabilities of ter low-speed acceleration, convenient fueling at home, and PEVs, are uncertain about their costs and benefits, and have zero tailpipe emissions when the vehicle operates solely on diverse needs that current PEVs might not meet. If the goal its battery. BEVs have no conventional transmissions or fu- is widespread deployment of PEVs, it is critical to identify el-injection systems to maintain, do not require oil changes, and evaluate the barriers to their adoption. and have regenerative braking systems that greatly prolong

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 28 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Introduction 13

FIGURE 1-5 Projected annual light-duty PEV sales as a percentage of total light-duty vehicle sales. NOTE: PEV, plug-in electric vehicle. SOURCE: Data courtesy of Navigant Research in Shepard and Gartner (2014).

THE COMMITTEE AND ITS TASK ment of task, as shown in Box 1-1, and provides recommendations on ways to mitigate the barriers identified. The committee included experts on vehicle technology, The premise of the statement of task is that there is electric utilities, business and financial models, econom- a benefit to the United States if a higher fraction of miles ics, public policy, and consumer behavior and response (see driven in the United States is fueled by electricity rather Appendix A for biographical information). As noted above, than by petroleum and that PEV deployment will lead to this the committee was asked to identify market barriers that are desired outcome. Two reasons are commonly assumed for the slowing the purchase of PEVs and hindering the deployment benefit. First, a higher fraction of miles fueled by electricity of supporting infrastructure in the United States and to recom- would reduce U.S. dependence on petroleum. Second, a mend ways to mitigate the barriers. The committee’s analysis higher fraction of miles fueled by electricity would reduce

was to be documented in two reports: an interim report and the amount of CO2 and other air pollutants emitted into the a final comprehensive report. The committee’s interim report atmosphere. The committee was not asked to research and was released May 2013 and identified infrastructure needs for evaluate the premise for the statement of task, and it has not electric vehicles, barriers to deploying that infrastructure, and tried to do so. However, it is appropriate to summarize the possible roles for the federal government in overcoming the scientific case on which the premise is based and ask if any barriers. It did not make any recommendations because the recent developments might call the premise into question. committee was in its initial stages of gathering data. After re- U.S. energy independence and security have been long- lease of the interim report, the committee continued to gather term U.S. goals. Every administration from Richard Nixon’s and review information and to conduct analyses. This final onward has proclaimed its importance. A PEV uses no petro- comprehensive report addresses the committee’s full state- leum onboard when it is being fueled by electricity, and in

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 29 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

14 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

BOX 1-1 Statement of Task

An ad hoc committee will conduct a study identifying the market barriers slowing the purchase of electric vehicles (EVs, which for this study include pure battery electric vehicles [BEVs] and plug-in hybrid electric vehicles [PHEVs]) and hindering the deployment of supporting infrastructure in the United States. The study will draw on input from state utility commissions, electric utilities, automotive manufacturers and suppliers, local and state governments, the Federal Energy Regulatory Commission, federal agencies, and others, including previous studies performed for the Department of Energy (DOE), to help identify barriers to the introduction of electric vehicles, particularly the barriers to the deployment of the necessary vehicle charging infrastructure, and recommend ways to mitigate these barriers. The study will focus on light-duty vehicles but also draw upon experiences with EVs in the medium- and heavy-duty vehicle market segment. Specifically, the committee will:

1. Examine the characteristics and capabilities of BEV and PHEV technologies, such as cost, performance, range, safety, and durability, and assess how these factors might create barriers to widespread deployment of EVs. Included in the examination of EV technologies will be the characteristics and capabilities of vehicle charging technologies. 2. Assess consumer behaviors and attitudes towards EVs and how these might affect the introduction and use of EVs. This assessment would include analysis of the possible manner by which consumers might recharge their vehicles (vehicle charging behaviors, e.g., at home, work, overnight, frequency of charging, time of day pricing, during peak demand times, etc.) and how consumer perceptions of EV characteristics will impact their deployment and use. 3. Review alternative scenarios and options for deployment of the electric vehicle infrastructure, including the various policies, including tax incentives, and business models necessary for deploying and maintaining this infrastructure and necessary funding mechanisms. The review should include an evaluation of the successes, failures, and lessons learned from EV deployment occurring both within and outside the United States. 4. Examine the results of prior (and current) incentive programs, both financial and other, to promote other initially uneconomic technologies, such as flex-fuel vehicles, hybrid electric vehicles, and now PHEVs/BEVs to derive any lessons learned. 5. Identify the infrastructure needs for the electricity sector, particularly the needs for an extensive electricity charging network, the approximate costs of such an infrastructure, and how utility investment decision making will play into the establishment of a charging network. As part of this assessment, the committee will identify the improvements in the electricity distribution systems needed to manage vehicle charging, minimize current variability, and maintain power quality in the local distribution network. Also, the committee will consider the potential impacts on the electricity system as a whole, potentially including: impacts on the transmission system; dispatch of electricity generation plants; improvements in system operation and load forecasting; and use of EVs as grid-integrated electricity storage devices. 6. Identify the infrastructure needs beyond those related to the electricity sector. This includes the needs related to dealer service departments, independent repair and maintenance shops, battery recycling networks, and emergency responders. 7. Discuss how different infrastructure deployment strategies and scenarios might impact the costs and barriers. This might include looking at the impacts of focusing the infrastructure deployment on meeting the needs for EVs in vehicle fleets, where the centralization of the vehicle servicing might reduce the costs for deploying charging infrastructure or reduce maintenance issues, or focusing the infrastructure deployment on meeting the needs for EVs in multi-family buildings and other high-density locations, where daily driving patterns may be better suited to EV use than longer commutes from single family homes in lower density areas. This might also include looking to the extent possible of how the barriers and strategies for overcoming barriers may differ in different U.S. localities, states, or regions. 8. Identify whether there are other barriers to the widespread adoption of EVs, including shortages of critical materials, and provide guidance on the ranking of all barriers to EV deployment to help prioritize efforts to overcome such barriers. 9. Recommend what roles (if any) should be played by the federal government to mitigate those market barriers and consider what federal agencies, including the DOE, would be most effective in those roles. 10. Identify how the DOE can best utilize the data on electric vehicle usage already being collected by the department.

The committee’s analysis and methodologies will be documented in two NRC-approved reports. The study will consider the technological, infrastructure, and behavioral aspects of introducing more electric vehicles into the transportation system. A short interim report will address, based on presentations to the committee and the existing literature, the following issues: 1. The infrastructure needs for electric vehicles; 2. The barriers to deploying that infrastructure; and 3. Optional roles for the federal government to overcome these barriers, along with initial discussion of the pros and cons of these options.

The final report will discuss and analyze these issues in more detail and present recommendations on the full range of tasks listed in Items (1) to (10) for the full study. The final report will include consideration of the infrastructure requirements and barriers as well as technological, behavioral, economic, and any other barriers that may slow the deployment of electric vehicles, as well as recommendations for mitigating the identified market barriers. It is envisioned that the committee will hold meetings in different locations around the United States, as well as collect information on experiences in other countries, in order to collect information on different approaches being taken to overcoming the barriers to electric vehicle deployment and its supporting charging infrastructure.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 30 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Introduction 15 2013, less than 0.7 percent of the U.S. grid electricity was pro- fuel combustion. For a PEV, a well-to-wheels analysis would duced from petroleum.7 Thus, widespread adoption of PEVs include emissions associated with electricity generation, such would lead to a large decrease in petroleum use. There is a as extraction of fuels, their transportation, and the transmis- modest caveat, however, to that conclusion. U.S. petroleum sion of the electricity. For PHEVs, a well-to-wheels analysis consumption in the light-duty vehicle fleet is regulated by would be a weighted average of the emissions from electrici- National Highway Traffic Safety Administration (NHTSA) ty-fueled and petroleum-fueled operation. through its Corporate Average Fuel Economy (CAFE) pro- There are several (often conflicting) methods to evalu- gram (see Chapter 7 for a detailed discussion). CAFE stan- ate well-to-wheels GHG emissions of vehicles. One method dards are based on average fuel economy of a manufacturer’s is to use well-to-wheels emission factors produced by DOE. vehicle fleet, so reductions in fuel use attributed to the sale of Given that method, an analysis of the 30 mpg 2014 Chevrolet a single PEV could be offset by the sale of an ICE vehicle8 that Cruze (an ICE vehicle), the 50 mpg 2014 Toyota Prius (one consumes more fuel, resulting in no net fuel savings from PEV of the cleanest HEVs), and the Nissan Leaf BEV charged on deployment (Gecan et al. 2012). However, petroleum con- the 2010 U.S. average electricity-generation mix shows that sumption might still be reduced by PEV deployment because the Cruze, Prius, and Leaf produce GHGs of 369 g/mi, 222 the CAFE program underestimates the petroleum-reduction g/mi, and 200 g/mi, respectively.10 Accordingly, the opera- benefit of PEVs. Specifically, the factor used by theCAFE tion of the BEV is estimated to produce about 46 percent less program to calculate a fuel-economy rating for compliance is GHG than the ICE vehicle and 10 percent less GHG than equivalent to assuming that 15 percent of the electrical energy the best hybrid. If one considered cleaner electricity sources used by a PEV is generated from petroleum, which is clearly (for example, ones in California or Washington, where large an overestimate of the petroleum used by the U.S. electric sec- numbers of PEVs are purchased), the BEV would produce tor (EPA/NHTSA 2012, p. 62821). Moreover, successful de- only about half of the GHG of the best HEV (DOE 2015). ployment of PEVs would help to enable the implementation Well-to-wheels analyses of this type have been reported for of increasingly stringent CAFE standards, resulting in lower average GHG emissions within each grid subregion as de- petroleum consumption, as noted by the Congressional Bud- fined by the U.S. Environmental Protection Agency (EPA) get Office (Gecan et al. 2012). (Anair and Mahmassani 2012). In addition to reduced petroleum consumption, lower An alternative analysis examines the emissions attrib- GHG emissions are noted as a reason for PEV deployment. uted to PEV charging by taking into account not only the A series of authoritative scientific reports (IPCC 2014; NCA average emissions at a given location, but also the variation 2014; NRC 2014) stress that the emission of GHGs, par- in emissions due to time of day and the type of generation ticularly CO2, is contributing in a measurable way to global added to provide the additional electricity needed for charg- warming and urge the United States to reduce its CO2 emis- ing. Analyses of this type differ on the emissions resulting sions. Because light-duty vehicles were responsible for 17.4 from PEVs, depending on the modeling approach and the percent of total U.S. GHG emission in 2012 (EPA 2014a), time frame used. On the one hand, EPA in its latest rulemak- reducing GHG emissions from the light-duty vehicle fleet ing for light-duty CO2 standards found that the additional is seen as an important approach for reducing overall GHG power plants used to meet PEV load in the 2022-2030 time emissions. A vehicle completely powered by electricity from frame would have lower emissions than the national aver- the U.S. electric grid is often called a zero-emission vehicle age power plant at that time (EPA/NHTSA 2012, p. 62821). (ZEV) insofar as it emits no CO2 or other pollutants from its On the other hand, a model that attempts to simulate emis- tailpipe. However, whether PEVs reduce total U.S. emissions sions from today’s grid using older data from 2007 to 2009 of CO2 and other GHGs depends on the emissions associated suggests that the marginal emission rates for PEV charging with the production of the grid electricity that the vehicles use might be higher than the average power plant emissions and and, in the case of PHEVs, on tailpipe emissions. Estimation in the worst case might even be higher than emissions attrib- of the emissions attributed to a vehicle whether operating on uted to HEVs and ICE vehicles (Graff Zivin 2014). gasoline or electricity is often referred to as a well-to-wheels Another factor to consider is the treatment of GHG analysis.9 For a gasoline vehicle, a well-to-wheels analysis emissions from PEVs under the joint CAFE-GHG standards would consider emissions from fossil fuel extraction, refining, (see Chapter 7 for a more detailed discussion). Similar to and transportation, as well as tailpipe emissions from onboard the CAFE program requirement for a fleetwide average fuel economy, fleetwide average GHG emission rates are restrict- 7 Estimate calculated from data reported in EIA (2013), Short ed to a certain average grams of CO2 per mile. Therefore, Term Energy Outlook. lower PEV emission rates are averaged with higher emis- 8 For this report, ICE vehicle or conventional vehicle refers to a sion rates from ICE vehicles. If, however, standards become light-duty vehicle that obtains all of its propulsion from an internal- increasingly more stringent, PEV sales might be needed combustion engine. 9 A more complete analysis is a lifecycle assessment that, in addition to the well-to-wheels assessment, includes environmental 10 The latest data for ICE tailpipe emissions and for the “upstream impacts from vehicle production (all aspects), vehicle use, and dis- emissions” of GHGs (CO2 equivalent) to produce electricity from posal of the vehicle at the end of its life. the 2010 U.S. electricity grid are available at www.fueleconomy.gov.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 31 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

16 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

to meet them, and early deployment of PEVs encouraged the market could rapidly bring about substantial reductions through incentives might allow the implementation of more in petroleum use and emissions in the time that a comparable stringent GHG standards in the future. To encourage PEV variety of PEVs are brought to market. Accordingly, the fo- deployment in the near term, EPA temporarily allows the cus in this report on PEVs should not be misinterpreted so portion of PEV miles that are estimated to be driven on elec- as to keep policy makers from encouraging the switch from tricity to be treated as zero emissions and lets a single PEV ICE vehicles to HEVs. count as more than one vehicle. That favorable treatment creates a short-term trade-off in GHG emissions that is an- THE COMMITTEE’S APPROACH TO ITS TASK ticipated to bring long-term benefits from PEV deployment. Emissions attributed to PEV operation might change Ten meetings were held over the course of this study. over time with changes in emissions from electricity genera- Seven meetings included open sessions during which the tion. The United States has reduced its GHG emissions over committee heard from the sponsor and invited speakers rep- the last several years by converting some of its electricity resenting national laboratories, state agencies, university production from coal to natural gas. The result is that, on centers, vehicle manufacturers and dealers, and other pri- average, a PEV fueled by electricity is now responsible for vate industries and consultants (see Appendix B for a list of less GHG per mile driven. Well-to-wheels emissions must speakers from all the open sessions). Committee subgroups continue to consider the evolving understanding of upstream also visited several sites in this country and abroad, includ- methane emissions from coal and natural gas production and ing Texas, Japan, Germany, and the Netherlands, to gather distribution (EPA 2014b). The substantial reductions in U.S. information on electric-vehicle programs. On those trips, the GHG emissions from electricity generation are expected to committee members met with national and regional govern- continue for some time, especially if the proposed EPA GHG ment officials, automobile manufacturers, charging compa- regulations of new and existing power plants and oil and gas nies, and other relevant organizations. On the basis of infor- wells are enacted. Thus, PEVs will make further reductions mation received at its meetings, its on-site visits, and from in GHG emissions as the U.S. electric grid changes to lower the literature, the committee prepared this final report. carbon sources for its electricity—a fact that is sometimes As discussed above, the committee accepted its charge overlooked. And PEVs fueled on electricity have the poten- and is not debating the merits of promoting, enabling, or in- tial to produce no well-to-wheels emissions if the electricity creasing PEV adoption. This report focuses on ways to ex- is generated from carbon-free sources. That is not the case tend the market from “early adopters” to more mainstream for even the most efficient petroleum-fueled ICE vehicles. If customers. Early-market customers for PEVs tend to base the United States intends to reach low levels of GHG emis- their purchase decisions more on personal values and less on sions (80 percent reduction), large-scale adoption of PEVs is purchase price. In contrast, mainstream-market customers one viable option (NRC 2013b). tend to weigh price and overall vehicle utility more heavily The committee concludes that the premise for the state- in their purchase decisions. ment of task—that there is an advantage to the United States One final issue concerns the rapidly changing market if a higher fraction of the miles driven here is fueled from and the various factors that hinder the adoption of PEVs— the U.S. electric grid—is valid now. The advantage becomes particularly the price of gasoline. Wide fluctuations in gaso- even greater each year that the United States continues to line prices, as occurred over the course of this study, affected reduce the GHGs that it produces in generating electricity. the committee’s comparisons and conclusions about the cumulative costs of vehicle ownership. As discussed in Chap- Finding: The average GHG emissions for which PEVs are ter 7, gasoline prices are an important factor in determining responsible are currently lower than emissions from even the the benefits of PEV ownership and can provide an incentive cleanest gasoline vehicles and will be further reduced as the or a disincentive for purchasing a PEV. To address the issue electricity for the U.S. grid is produced from lower carbon of fluctuating gasoline prices, the committee decided that the sources. best approach was to use a range of gasoline prices, from $2.50 to $4.00, in its calculations, to present ranges as appro- Recommendation: As the United States encourages the priate throughout its report, and to draw conclusions based adoption of PEVs, it should continue to pursue in parallel on these ranges. the production of U.S. electricity from increasingly lower carbon sources. ORGANIZATION OF THIS REPORT

The committee notes that the use of HEVs rather than This final report is organized into seven chapters and ICE vehicles would provide a large reduction in U.S. petro- three appendixes. Chapter 2 discusses the current character- leum use and emissions. If their small market share could be istics and capabilities of PEV technologies. Chapter 3 pro- substantially increased, the many types of HEVs already on vides a brief assessment of consumer behavior and attitudes

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 32 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Introduction 17 toward PEVs and how they are affecting PEV deployment. DOE. 2013b. “State Laws and Incentives.” Alternative Fu- Chapter 4 discusses what can be done to improve institu- els Data Center. http://www.afdc.energy.gov/laws/state. tional support for PEV deployment. Chapter 5 provides an Accessed January 29, 2013. in-depth discussion of the charging infrastructure needed DOE. 2014. Presentation at the NEXTSteps Symposium, for PEV deployment, and Chapter 6 evaluates the ability of University of California, Davis. the electric infrastructure to meet the increased electricity DOE. 2015. “Greenhouse Gas Emissions for Electric and demand in light of the new charging infrastructure. Chap- Plug-In Hybrid Electric Vehicles.” http://www.fuelecon ter 7 discusses ways to motivate the consumer. Appendix A omy.gov/feg/Find.do?zipCode=90001&year=2014&ve provides biographical information for committee members, hicleId=34699&action=bt3. Appendix B lists the meetings and the presentations made in EIA (U.S. Energy Information Administration). 2013. Short- open sessions, and Appendix C provides some information Term Energy Outlook. http://www.eia.gov/forecasts/steo/ on international programs to support PEV deployment. archives/dec13.pdf. EIA. 2014. Annual Energy Outlook 2014. http://www.eia. REFERENCES gov/forecasts/aeo/. Accessed September 24, 2014. EPA (U.S. Environmental Protection Agency). 2014a. In- Anair, D., and A. Mahmassani. 2012. State of Charge: Electric ventory of U.S. Greenhouse Gas Emissions and Sinks: Vehicles’ Global Warming Emissions and Fuel-Cost Sav- 1990-2012. EPA 430-R-14-003. http://www.epa.gov/ ings across the United States. Union of Concerned Scien- climatechange/Downloads/ghgemissions/US-GHG-In- tists. http://www.ucsusa.org/assets/documents/clean_ve ventory-2014-Main-Text.pdf. hicles/electric-car-global-warming-emissions-report. EPA. 2014b. Regulatory Impact Analysis for the Proposed pdf. Carbon Pollution Guidelines for Existing Power Plants ANL (Argonne National Laboratory). 2011. “Hybrid Vehicle and Emission Standards for Modified and Reconstructed Technology.” http://www.transportation.anl.gov/hev/ind Power Plants. EPA-452/R-14-002. http://www2.epa.gov/ ex.html. Accessed March 14, 2013. sites/production/files/2014-06/documents/20140602ria- ANL. 2012. “Advanced Battery Research, Development, and clean-power-plan.pdf. Testing.” http://www.transportation.anl.gov/batteries/in- EPA/NHTSA (U.S. Environmental Protection Agency and dex.html. Accessed March 14, 2013. National Highway Traffic Safety Administration). 2012. ANL. 2013. “Argonne Leads DOE’s Effort to Evaluate Plug- 2017 and Later Model Year Light-Duty Vehicle Green- in Hybrid Technology.” http://www.transportation.anl. house Gas Emissions and Corporate Average Fuel Econ- gov/phev/index.html. Accessed March 14, 2013. omy Standards. Federal Register 77(199): October. http:// CNCDA (California New Car Dealers Association). 2015. Cal- www.gpo.gov/fdsys/pkg/FR-2012-10-15/pdf/2012- ifornia Auto Outlook: Fourth Quarter 2014. Volume 11, 21972.pdf. Number 1, February. http://www.cncda.org/CMS/Pubs/ Gecan, R., J. Kile, and P. Beider. 2012. Effects of Federal Tax Cal_Covering_4Q_14.pdf. Credits for the Purchase of Electric Vehicles. Congres- Cobb, J. 2013. “December 2012 Dashboard.” Hybrid Cars, sional Budget Office. http://www.cbo.gov/sites/default/ January 8. http://www.hybridcars.com/december-2012- files/cbofiles/attachments/09-20-12-ElectricVehicles_0. dashboard. pdf. Cobb, J. 2014. “August 2014 Dashboard.” Hybrid Cars, Graff Zivin, J.S., M.J. Kotchen, and E.T. Mansur. 2014. Spa- September 4. http://www.hybridcars.com/august-2014- tial and temporal heterogeneity of marginal emissions: dashboard/. implications for electric cars and other electricity-shift- Cobb, J. 2015. “December 2014 Dashboard.” Hybrid Cars, ing policies. Journal of Economic Behavior & Organi- January 6. http://www.hybridcars.com/december-2014- zation 107: 248-268. dashboard/. Inside EVs. 2015. “Monthly Plug-in Sales Scorecard.” http:// DOE (U.S. Department of Energy). 2009. Recovery Act—Elec- insideevs.com/monthly-plug-in-sales-scorecard/. tric Drive Vehicle Battery and Component Manufacturing IPCC (Intergovernmental Panel on Climate Change). 2014. Initiative. Funding Opportunity No. DE-FOA-0000026. Fifth Assessment Report. http://www.ipcc.ch/report/ar5/ http://www1.eere.energy.gov/vehiclesandfuels/pdfs/de- index.shtml. foa-0000026-000001.pdf. NCA (National Climate Assessment). 2014. “Climate Change DOE. 2011. One Million Electric Vehicles by 2015: February Impacts in the United States.” U.S. Global Change Re- 2011 Status Report. http://www1.eere.energy.gov/vehicles search Program, Washington, DC. http://nca2014.global- andfuels/pdfs/1_million_electric_vehicles_rpt.pdf. change.gov. DOE. 2013a. “Hybrid and Vehicle Systems.” Vehicle Tech- Nikkei Asian Review. 2012. “Hybrids 19.7% of New Cars Sold nologies Office. http://www1.eere.energy.gov/vehicle- in May 2012 in Japan.” Integrity Exports, June 8. http:// sandfuels/technologies/systems/index.html. Accessed integrityexports.com/2012/06/08/hybrids-19-7-of-new- March 14, 2013. cars-sold-in-ma-in-japan/. Accessed June 10, 2012.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 33 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

18 Overcoming Barriers to Deployment of Plug-in Electric Vehicles NRC (National Research Council). 2013a. Review of the Schiffer, M.B., T.C. Butts, and K.K. Grimm. 1994. Taking Research Program of the U.S. DRIVE Partnership: Charge: The Electric Automobile in America. Washing- Fourth Report. Washington, DC: The National Acad- ton, DC: Smithsonian Institution Press. emies Press. Shepard, S., and J. Gartner. 2014. Electric Vehicle Geographic NRC. 2013b. Transitions to Alternative Vehicles and Fuels. Forecasts. Navigant Research. Boulder, CO. Washington, DC: The National Academies Press. Toyota. 2014. “Worldwide Sales of Toyota Hybrids Top 6 NRC. 2014. Climate Change: Evidence and Causes. Wash- Million Units.” News release. January 14. http://corpo- ington, DC: The National Academies Press. ratenews.pressroom.toyota.com/releases/worldwide+to NREL (National Renewable Energy Laboratory). 2013. “Plug- yota+hybrid+sales+top+6+million.htm. In Hybrid Electric Vehicles.” Vehicle Systems Analysis. UCS (Union of Concerned Scientists). 2014. “Electric Ve- http://www.nrel.gov/vehiclesandfuels/vsa/plugin_hybrid. hicle Timeline: Electric Cars, Plug-In Hybrids, and Fuel html. Accessed March 14, 2013. Cell Vehicles.” http://www.ucsusa.org/clean_vehicles/ Peterson, S.B., J.F. Whitacre, and J. Apt. 2011. Net air emis- smart-transportation-solutions/advanced-vehicle-tech- sions from electric vehicles: The effect of carbon price nologies/electric-cars/electric-vehicle-timeline.html. and charging strategies. Environmental Science and Woodall, B., and B. Klayman. 2015. U.S. auto sales end Technology 45(5): 1792-1797. 2014 strong but slower growth looms. Reuters, January Pontes, J. 2015. “World Top 20 December 2014.” EV Sales, 5. http://www.reuters.com/article/2015/01/05/us-autos- January 31. http://ev-sales.blogspot.com/2015/01/world- sales-usa-idUSKBN0KE0WH20150105. top-20-december-2014-special.html.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 34 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies

As discussed in Chapter 1, the assigned task for the pres- tinues with a discussion of high-energy batteries, the critical ent report is to examine barriers to the adoption of plug-in and expensive components for all PEVs, and the possibil- electric vehicles (PEVs), which use electricity from the U.S. ity of increasing the energy densities of these batteries. A electric grid as their fuel. When powered by electricity from summary of current and projected battery costs is provided the grid, which uses little petroleum to produce electricity, because it is primarily higher battery costs that make PEVs such vehicles require essentially no petroleum, and they emit cost more than ICE vehicles. The chapter concludes with a

no carbon dioxide (CO2) or other harmful pollutants from discussion of vehicle charging and charging options. The the tailpipe (EPA 2012). The premise for the assigned task is committee’s findings and recommendations are provided that such vehicles have the potential for significantly lower- throughout this chapter. ing petroleum consumption and decreasing emissions now

and even more so in the future. The CO2 emissions advan- TYPES OF PLUG-IN ELECTRIC VEHICLES tage will grow as the United States continues to switch to lower-carbon-emitting sources of electricity by phasing out Essentially all U.S. vehicles today have an ICE that uses coal and natural gas combustion and replacing those energy gasoline or diesel fuel that is derived from petroleum and

sources with solar, wind, or nuclear energy, or alternatively produces CO2 and other harmful emissions as the vehicles by using carbon capture and sequestration for coal and natu- travel. Hybrid electric vehicles (HEVs) achieve a lower fuel ral gas plants if that technology ever proves to be practical. consumption than conventional vehicles of the same size and As described in more detail in this chapter, electric- performance. They typically have a smaller ICE and a high- ity from a battery powers the electric motor of a PEV and power battery and electric motor to increase the vehicle’s is thus the analog of the gasoline in a tank that powers the acceleration when needed and to power the vehicle briefly at internal-combustion engine (ICE) of a conventional vehicle. low speeds. Electric energy is provided to the battery when The hundreds of miles of range that is typical for a conven- the vehicle brakes and is produced by the ICE using power tional vehicle depends on how many gallons of fuel the tank that is not needed to propel the vehicle. The lower fuel con- can hold and on the fuel economy of the vehicle. Similarly, sumption that can be achieved is illustrated by the 50 miles the all-electric range (AER) of a vehicle (the distance that per gallon (mpg) of gasoline that is achieved by the Toyo- it can travel fueled only by the electricity that can be stored ta Prius, the best-selling HEV. There are many other HEV in its battery) depends on the size of the battery and the ef- models available in the market, most of which use much less ficiency of the vehicle. The AER, like the range of an ICE fuel than their ICE counterparts. Although HEVs still consti- vehicle, depends on such factors as the vehicle design, in- tute a small fraction of the U.S. vehicle fleet, the more rapid cluding its aerodynamics, rolling resistance, and weight; the adoption of efficient HEVs could be important for meeting driving environment, including road grade and outside tem- the increasingly stringent corporate average fuel economy perature; the amount of heating and cooling that is used; and (CAFE) and greenhouse gas (GHG) emission standards that how aggressively the vehicle is driven (NREL 2013). Some are helping to drive down the demand for petroleum and factors, such as outside temperature, will have a greater ef- to decrease vehicle tailpipe emissions. However, although fect on PEVs than ICE vehicles. The ranges quoted in the HEVs use batteries and electric motors, they derive all of present report are taken from the U.S. Environmental Protec- their electric and mechanical energy from their gasoline or tion Agency (EPA) data on results from the standard driving diesel fuel. Thus, HEVs are used as a point of comparison for cycle (EPA 2014). the present report, but they are not its primary focus. This chapter begins with a discussion of the capabili- As noted in Chapter 1, the PEVs that are the focus of ties and limitations of four classes of PEVs, each presenting the present report are often divided into two categories: bat- different obstacles to widespread consumer adoption. It con- tery electric vehicles (BEVs) and plug-in hybrid electric ve-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 35 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

20 Overcoming Barriers to Deployment of Plug-in Electric Vehicles hicles (PHEVs) that include an ICE and an electric motor. types and features are available compared with the types and This chapter uses vehicle AER to distinguish four classes of features available for conventional ICE vehicles and HEVs. PEVs. The reason is that the obstacles to consumer adoption Chapter 3 discusses the current paucity of choices as a pos- and the charging infrastructure requirements differ for the sible barrier to PEV adoption. As PEVs become more com- four classes of PEVs. BEVs are separated into long-range mon, however, the variety of choices will increase, and some BEVs and limited-range BEVs, and PHEVs are separated models could emerge that do not fit perfectly into one of the into range-extended PHEVs and minimal PHEVs. four categories described here. There are now examples in the market for each type of PEV, and the committee uses some of them to illustrate their Finding: The increasing number of PEVs entering the mar- capabilities (see Table 2-1). Despite the increasing number ket demonstrates the possibility of various types of electri- of PEVs entering the market, however, far fewer vehicle cally fueled vehicles, although far fewer vehicle types and

TABLE 2-1 Definitions and Examples of the Four Types of Plug-in Electric Vehicles Vehicle Battery Capacitya All-Electric Rangeb Type 1. Long-Range Battery Electric Vehicle. Can travel hundreds of miles on a single battery charge and then be refueled in a time that is much shorter than the additional driving time that the refueling allows, much like an ICE vehicle or HEV.

85 kWh nominal 265 miles

2014 Tesla Model S © Steve Jurvetson, licensed under Creative Commons 2.0 (CC-BY-2.0) Type 2. Limited-Range Battery Electric Vehicle. Is made more affordable than the long-range BEV by reducing the size of the high-energy battery. Its limited range more than suffices for many commuters, but it is impractical for long trips.

23 kWh nominal 76 miles

2014 Ford Focus Electric Image courtesy of Ford Motor Company Type 3. Range-Extended Plug-in Hybrid Electric Vehicle. Operates as a zero-emission vehicle until its battery is depleted, whereupon an ICE turns on to extend its range.

16.5 kWh nominal (~11 kWh usable) 38 miles

2014 Chevrolet Volt © General Motors Type 4. Minimal Plug-in Hybrid Electric Vehicle. Is mostly an HEV. Its small battery can be charged from the grid, but it has an all-electric range that is much smaller than the average daily U.S. driving distance.

4.4 kWh nominal (~3.2 kWh usable) 11 miles (blended) 6 miles (battery only)

2014 Toyota Plug-in Prius Image courtesy of Toyota Motor Corporation a Nominal battery capacities, reported by manufacturers in product specifications, are for a battery before it goes into a vehicle. Vehicle electronics restrict the usable battery capacity to what becomes the vehicle’s all-electric range. b The all-electric ranges noted are average values estimated by EPA. The motor size and design architecture of the Toyota Plug-in Prius require the use of its ICE to complete the Federal Test Procedure; therefore, its range is given for both blended, charge-depleting operation and battery-only operation. All other vehicle ranges are given only for fully electric, charge-depleting operation. NOTE: HEV, hybrid electric vehicle; ICE, internal- combustion engine. SOURCES: Based on data from Duoba (2012); DOE/EPA (2014a, 2014b, 2014c, 2014d, 2014e); DOE (2012, 2013); EPA (2014); Ford (2014); and Toyota (2014).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 36 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 21 features are currently available than are available for con- trips. A limited-range BEV is more affordable simply because ventional ICE vehicles and HEVs. a smaller high-energy battery is installed, giving it a shorter range. The 2014 Nissan Leaf, a midsize car, is the best-selling Type 1: Long-Range Battery Electric Vehicles example. It has a 24 kWh battery and an 84-mile range (DOE/ EPA 2014b). A more recent addition to the limited-range BEV Today’s drivers are accustomed to ICE and HEV vehicles market is the Ford Focus Electric compact car, which has a 76- that are able to drive for hundreds of miles and then be refu- mile range (DOE/EPA 2014c). As noted earlier in this chapter, eled at any gasoline station in several minutes. Extended trips the actual range of a BEV will depend on a variety of factors, are practical insofar as the refueling time is much shorter than including climate, road grade, and driver behavior. The differ- the additional driving time that refueling provides. The full- ence between the range, fuel economy, and emission perfor- size Tesla Model S is a demonstration that hundreds of miles mance estimated for regulatory compliance and what is actu- are also possible with a BEV that gets its energy entirely from ally experienced by drivers of all types of light-duty vehicles the electric grid. It has a range based on the EPA driving cycle continues to be controversial and is discussed in other NRC of 265 miles for a single charge of its 85 kWh battery (DOE/ reports (NRC 2011, 2013). EPA 2014a). Half of the charge of a depleted battery can be re- The ranges that are achievable by limited-range BEVs plenished in 20 minutes at any of the superchargers that Tesla are much longer than the 40 or fewer miles that 68 percent of is installing for its customers along major U.S. highways. That U.S. drivers drive in a day, making these vehicles adequate for charge would extend the driving distance by about 132 miles. normal commuting and the average daily use (FHWA 2011). Thus, the Tesla Model S is considered a long-range BEV be- However, drivers of ICE vehicles are accustomed to being cause it can drive for hundreds of miles on a charge and then able to travel well beyond the average daily distance when the be refueled in a time that is much shorter than the additional need arises and can add hours of additional traveling time by driving time that the refueling allows. Although filling a vehi- simply refilling a gasoline or diesel fuel tank in several min- cle with gasoline or diesel would be much quicker, the ability utes. For a limited-range BEV, however, a half hour of the to travel almost 400 miles stopping only once for a 20-minute fastest available charging will typically allow an hour or even recharge is a notable achievement for a BEV. With its high less of additional driving, making extended trips impractical. acceleration performance, low noise, high-end styling, and ex- For extended trips and driving distances much beyond the pected low maintenance, the Tesla Model S has earned several AER, the limited-range BEV driver needs to have access to a consumer performance awards (MacKenzie 2013; Consumer second vehicle that has no serious range limitations or to some Reports 2014). other transportation means. As discussed in Chapter 3, many The Tesla Model S is priced as a high-end luxury vehicle households have two or more vehicles, so trading vehicle util- comparable to a high-end BMW and is not affordable for most ity within a household is already common. For its customers, 1 U.S. drivers. Nonetheless, it is an important demonstration of BMW is experimenting with offering access to an ICE vehicle the possibility of a long-range BEV for consumers. For now, for the occasional long trip to see if this perk lowers the barrier however, high battery cost is a barrier to the mass adoption of to adoption of its vehicles. Rental companies like Hertz have the Tesla Model S and other BEVs. The fuel cost per mile and also indicated that they are interested in filling that same niche maintenance costs are much smaller for BEVs than for ICE (Hidary 2012). vehicles, but not enough to offset their higher purchase price at current U.S. petroleum prices. The situation can be quite Finding: Limited-range BEVs are the only type of PEV that different in countries where gasoline and diesel fuel cost 2 or have a considerable range limitation. However, the range 3 times as much as in the United States. that they do have more than suffices for the average daily travel needs of many U.S. drivers. Finding: The possibility of a long-range BEV that is powered by grid electricity rather than gasoline or diesel and that Finding: Given the substantial refueling time that would be meets consumer performance needs has been clearly demon- required, limited-range BEVs are not practical for trips that strated by the full-size Tesla Model S. would require more than one fast charge.

Type 2: Limited-Range Battery Electric Vehicles Type 3: Range-Extended Plug-in Hybrid Electric Vehicles The high cost of high-energy batteries leads to three types of more affordable PEVs. The first sacrifices driving range A range-extended PHEV2 is similar to a long-range or and the other two sacrifice zero tailpipe emissions for longer limited-range BEV in that the battery can be charged from

1 The cost of producing a Model S is currently offset somewhat in that Tesla is able to sell the zero-emission-vehicle (ZEV) credit it 2 The term range-extended PHEV is a general category based on earns for each vehicle to other vehicle manufacturers to allow them the all-electric range of the PHEV and should not be confused with to comply with the ZEV mandate. See Chapter 7 for a detailed dis- the term extended-range electric vehicle that General Motors uses cussion of the ZEV program. to describe the Chevrolet Volt.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 37 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

22 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

the electric grid. However, the battery is smaller than that in Recommendation: Minimal PHEVs should be treated as a BEV, and the vehicle has an onboard ICE fueled by gaso- HEVs with respect to financial rebates, HOV access, and line or diesel fuel that is able to charge the battery during a other incentives to encourage PEV adoption. trip. Although extended trips fueled only by electricity are not practical, the vehicle has a total range comparable with that of HIGH-ENERGY BATTERIES a conventional vehicle because of the onboard ICE. The 2014 Chevrolet Volt with an AER of 38 miles (DOE/EPA 2014d) The capacity, weight, and volume of the high-energy bat- is the best-selling example, and the 2014 Ford Energi models tery in a PEV largely determine its range, performance, and (Fusion Energi and CMax Energi) that have AERs of 20 miles cost relative to an HEV or an ICE vehicle. This section sum- are other prominent examples. The AERs are comparable to marizes the energy densities with respect to weight and vol- the average daily driving distance in the United States. ume that have been achieved with battery chemistries so far The consequence of eliminating the range restrictions of and considers possible improvements, despite the difficulty of a limited-range BEV is that the added ICE uses petroleum and precisely predicting future developments. Differences in cur- produces tailpipe emissions. Although the ICE can be oper- rent battery geometries and cooling strategies are discussed, ated to maximize efficiency and minimize emissions, the frac- along with the associated uncertainties about long-term bat- tion of miles traveled propelled by electricity depends on how tery durability. willing and able a driver is to recharge the battery during a trip longer than the AER. On the basis of data collected by DOE Energy Density and Battery Chemistry through its EV Project, early adopters of the Chevrolet Volt appear to be very motivated to minimize their use of the ICE The battery in a PEV is the counterpart to the fuel tank for engine by charging more frequently and logging more electric an ICE vehicle. Electric energy from the electric grid is stored miles per day than Nissan Leaf drivers (Schey 2013). Blanco in the battery until it is needed by the electric motor to turn (2014) reported that 63 percent of all miles traveled by the the wheels. The more energy stored in the battery, measured Chevrolet Volt are fueled by electricity. in kilowatt-hours (kWh), the longer the vehicle’s AER. An 80 kWh battery can propel a vehicle twice as far as can a 40 kWh Finding: The Chevrolet Volt demonstrates that if they become battery when the same vehicle is driven in the same way, just widely adopted, range-extended PHEVs with AERs compa- as 20 gallons of gasoline can provide the energy to propel an rable to or greater than the average U.S. travel distance offer ICE vehicle twice as far as 10 gallons of gasoline. The nomi- the possibility of significant U.S. petroleum and emission re- nal battery capacities for the PEVs in Table 2-1 are what the ductions without range limitations. batteries can store as their state of charge (SOC) goes from fully discharged (SOC of 0 percent) to fully charged (SOC of Type 4: Minimal Plug-in Hybrid Electric Vehicles 100 percent). Vehicle manufacturers use electronics to restrict how fully a battery can be charged and how far the vehicle is Minimal PHEVs are PEVs whose small batteries can be able to deplete the charge in its battery. They make different initially charged from the electric grid to provide electric pro- choices for the usable capacity of their vehicle batteries be- pulsion for an AER that is much less than the average daily cause it is known that this factor affects the degradation of the travel distance for the U.S. driver. Among many examples, the battery over time, even though the degradation has yet to be 2014 Plug-in Toyota Prius is a minimal PHEV in that its AER fully characterized or understood. is only 6 miles (DOE/EPA 2014e). It is an extreme example A battery’s energy density (see Figure 2-1) determines the of a car that is designed for minimum compliance with regu- mass and volume of the battery necessary to store the energy lations rather than to give good electric-drive performance. that a PEV requires. The vertical axis in Figure 2-1 is the en- Minimal PHEVs allow a manufacturer to comply with regula- ergy storage capacity per unit volume (Wh/L), and the hori- tions for obtaining PEV emission credits without the expense zontal axis is the energy storage capacity per unit mass (Wh/ of designing and producing a car that is optimized for using kg). Lead acid batteries have a relatively small energy density, electricity instead of petroleum. They allow their drivers to even though they provide starting, lighting, and ignition for comply with requirements for high-occupancy-vehicle (HOV) essentially all the ICE vehicles around the world. The Toyota lane access whether or not they bother to charge from the grid Prius was the first mass-produced vehicle to use nickel-metal (CCSE 2014). As might be expected, driver usage surveys of hydride (NiMH) batteries. Such batteries have about twice Plug-in Prius drivers show that a substantial fraction do not the energy density of lead acid batteries, and they proved to regularly charge their vehicles (Chernicoff 2014). Minimal be very reliable when they were used in all the early HEVs. PHEVs are essentially HEVs. However, there seems to be no prospect for the large increases in energy density that would be required to make them at- Finding: Minimal PHEVs with AERs much shorter than the tractive for use in PEVs. Lithium-ion batteries were invented average daily driving distance in the United States are es- in the 1970s (Goodenough and Mizushima 1981) and mass sentially HEVs. produced for the first time by Sony for laptop computers in 1991 (Yoshino 2012). In the following two decades, lithium-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 38 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 23

FIGURE 2-1 The volume energy density and the mass energy density for various battery types. NOTE: LIB, lithium-ion battery; LPB, lithium-polymer battery; Ni-Cd, nickel cadmium; Ni-MH, nickel-metal hydride; Ni-Zn, nickel zinc; Wh/kg, watt-hour per kilo- gram; Wh/L, watt-hour per liter. SOURCE: Amine (2010).

ion batteries took over the small electronics market in such The lithium-ion batteries in vehicles differ in the chem- devices as laptop computers and cell phones. In recent years, istries and materials that are used and in the energy densities they have also become the battery of choice for PEVs and for achieved (Table 2-2). In a lithium-ion battery (see Figure new HEV models. 2-2), the positive lithium-ions flow between the anode and An electrically powered vehicle needs only about one the cathode within the electrolyte, as do electrons in an ex- quarter of the stored energy that an ICE vehicle needs to de- ternal circuit connected between the anode and cathode. The liver the same energy to turn the wheels. Most of the energy cathodes used are described using chemical formulae that that combustion releases from the fuel within an ICE is wasted provide their composition. All anodes but one are carbon.

as heat that is dissipated through the radiator and exhaust. The All PEV batteries use an organic solution of LiPF6 as the large efficiency advantage of the PEV, however, is more than electrolyte. overcome by the much smaller energy density in a charged The committee notes that the design of a vehicle bat- battery compared with the energy density of gasoline. The re- tery is related not only to the battery chemistry but also to sult is that PEV batteries now weigh much more and occupy the power and energy requirements of the various applica- a much larger volume than a tank filled with gasoline. For ex- tions. For example, PHEVs require more power than BEVs; ample, the 85 kWh battery in a Tesla Model S, the largest pro- thus, BEVs can use thicker, cheaper electrodes. Furthermore, duction vehicle battery so far, weighs about 1,500 lb3 (Tesla PHEV batteries must be cycled more frequently than BEV 2014a). Delivering the same energy to the wheels of an ICE batteries, so PHEV batteries tend to use a smaller portion vehicle requires the combustion of slightly less than 9 gallons of the nominal battery capacity. Those two facts affect the of gasoline, which weighs about 54 lb. battery structure and cost per kilowatt-hour and are taken The increased weight (about that of seven extra pas- into account in various analyses of PEV battery costs (Dan- sengers) reduces the acceleration and the range that would iel 2014; Sakti et al. 2014) and in the EPA/NHTSA analysis otherwise be realized, although the powerful motor in the that informed the committee’s analysis of battery costs as Model S overcomes the acceleration problem. Accommodat- discussed below. ing large, heavy batteries makes it difficult to use an ICE or HEV platform for an electric vehicle. A vehicle designed Projected Energy Density Increases and from its beginning to have electric propulsion has more op- Possible New Battery Chemistries tions. The Model S, for example, was designed with a battery compartment under the vehicle’s entire floor board so that Lithium-ion batteries with increased energy density are the heavy batteries are used to keep the vehicle’s center of naturally the subject of research and development efforts. gravity low to improve handling. It is difficult to predict success or its timing, but three approaches that are being pursued are worthy of mention. 3 The estimate is based on Tesla’s reported energy density for the Model S battery of 121 Wh/kg (Tesla 2014a).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 39 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

24 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 2-2 Representation of a lithium-ion battery that shows lithium ions traveling between the anode and the cathode and electrons traveling through the external circuit to produce an electric current. SOURCE: Kam and Doeff (2012).

TABLE 2-2 Properties of Lithium-Ion Batteries in Four Plug-in Electric Vehicles on the U.S. Market PEV Cathode Anode Supplier Cell Type No. of Cells Energy (kWh) Power (kW)

Tesla Model S NCA = LiNi0.8Co0.15Al0.05O2 Carbon Panasonic Cylindrical ~8,000 85 270

Chevrolet Volt LMO = LiMn2O4 Carbon LG Chem Prismatic 288 16.5 111

Nissan Leaf LMO = LiMn2O4 Carbon Nissan/NEC Prismatic 192 24 90

Honda Fit NMC = LiNi1/3Mn1/3Co1/3O2 Li4Ti5O12 Toshiba Prismatic 432 20 92 NOTE: Al, aluminum; Co, cobalt; kWh, kilowatt-hour; Li, lithium; LMO, lithium manganese oxide; Mn, manganese; NCA, nickel cobalt aluminum oxide; NMC, nickel manganese cobalt oxide; Ni, nickel; O, oxygen; Ti, titanium.

• Increasing the number of lithium atoms in a layered The committee estimates that although there can be no cathode structure has been shown in the laboratory to guarantee, as much as a twofold increase in energy density increase the energy density (Julien et al. 2014). could come from some combination of the three approaches • Developing electrolytes that can operate at 4.8 V rather within the next decade. Such an increase would allow an im- than 4.2 V would increase the energy density (Pham et portant reduction in the volume and weight of high-energy al. 2014). batteries. Most important, however, the cost per kilowatt- • Replacing the carbon anode with one that includes hour needs to decrease; a battery having twice the energy silicon would improve the energy density (Ge et al. density at twice the cost would not make PEVs any more 2013). Theoretically, a pure silicon anode would have affordable. Nonetheless, even with such an improvement, an energy density 10 times that of a pure carbon anode. battery energy densities would still be much smaller than the However, pure silicon anodes are not practical because energy density of gasoline. they crumble during a charging cycle, being unable to On a longer time scale, other battery chemistries could sig- withstand having their volume changed by more than a nificantly increase the energy density. The theoretical energy factor of three. Mixtures of silicon and carbon with ap- density for a lithium-air battery is 5,200 Wh/kg (Rahman et al. propriate binders might minimize the volume change 2014), which is comparable to that of gasoline. Such a battery and yet provide an increased energy density. uses oxygen from air and therefore does not need to store an

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 40 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 25 oxidizer. PolyPlus (2009) claims to have a battery capable of time (see Figure 2-3) (Pesaran et al. 2013). Reports on short- 700 Wh/kg and expects to produce a rechargeable battery with er battery life for Nissan Leafs in Arizona seem consistent a higher energy density. Another promising approach is the with that observation (Gordon-Bloomfield 2013). As a - re development of a high-energy density lithium-sulfur battery. sult, Nissan has tested new battery pack designs to address Sion Power, the recipient of substantial Advanced Research the observed problem (Gordon-Bloomfield 2013), and press Projects Agency-Energy (ARPA-E) funding, claims that “over reports of the increased rate of battery deterioration have not 600 Wh/kg . . . and 600 Wh/L in energy density are achievable continued. However, it is not clear whether the problem has in the near future” (Sion Power 2014). Substantial challenges been solved. Although Figure 2-3 illustrates preliminary re- remain for both lithium-air and lithium-sulfur batteries, how- sults of studying the effect of temperature on battery capac- ever, particularly in producing batteries that survive frequent ity, battery life depends also on cycling at various depths of recharging, so it is difficult to predict if and when batteries charge, rate of charge and discharge, and likely many other with much higher energy densities will be available. variables besides temperature. Only long-term experience in hot climates will establish whether some manufacturers must Finding: Affordable batteries with higher energy densities improve battery temperature regulation, use different battery and longer useful lives could greatly increase the all-electric chemistries, or restrict sales in hot climates. range and presumably increase the adoption rate for PEVs. ICE vehicle manufacturers have a good understanding of how long their products will perform, and this knowledge Finding: Although there can be no guarantee, as much as a allows them to predict warranty costs. PEV manufacturers twofold increase in energy density from present values of 100- are still learning about battery longevity. As more PEVs en- 150 Wh/kg could come from some combination of current re- ter the market, vehicle manufacturers have the chance to ex- search efforts within the next decade. periment with various warranties and battery maintenance contracts as they look for affordable ways to reassure and Finding: Battery research is critical because more practical share risk among consumers that use these vehicles under re- vehicle batteries that have higher energy densities and longer al-world conditions. Vehicle leasing is becoming more popu- life are needed to address important concerns about battery lar and promoted by some manufacturers partly because this range and durability. option allows a consumer to avoid long-term liability for a battery if over time the battery performance degrades below Battery Geometry, Cooling, and Durability an acceptable level.

Just as there is no consensus on what is the best lithium- Finding: Concerns about the durability and performance of ion battery chemistry, there is also no consensus on what is the current lithium-ion batteries at extremely high and low the most stable or most economical battery geometry or on temperatures could be a barrier to PEV adoption, depending how much the battery temperature should be regulated for the on the durability observed as more vehicles are driven for sake of battery longevity. As more PEVs are driven, the early longer times. adopters are essentially testing both the various battery chemistries and the battery temperature regulation choices under RELATIVE COSTS OF PLUG-IN real-world conditions that are hard to duplicate in laboratories. ELECTRIC AND ICE VEHICLES Tesla connects many thousands of small cylindrical cells, each having the same physical shape and size as those that are Studies of current and projected costs of high-energy bat- commonly used in computer batteries, thereby profiting from teries and nonbattery components (EPA/NHTSA 2012) sug- the extensive manufacturing experience for cells with this gest that the difference in cost of producing a PEV and an geometry. All other manufacturers use many fewer but much ICE vehicle is (and will be) primarily due to the cost of the larger cells in so-called prismatic or pouch geometries. A Nis- high-energy battery. Those studies are part of the regulatory san Leaf air-cools its batteries, while the Chevrolet Volt and analysis performed by EPA and the National Highway Traf- the Tesla Model S use a liquid system and heat exchangers to fic Safety Administration (NHTSA) for the recent 2017-2025 regulate battery temperature. Over the next several years, the combined CAFE-GHG standards for light-duty vehicles. The real-world experience reported by early adopters should make comprehensive regulatory analysis includes vehicle-simu- clear the advantages and disadvantages of each strategy. lation modeling and detailed component cost analysis (cost Concerns about the durability and performance of the teardown studies) performed by external consultants to deter- current lithium-ion batteries at extremely high and low tem- mine cost and effectiveness of a wide range of technologies, peratures could be a barrier to PEV adoption, depending including conventional ICE vehicles, HEVs, and PEVs. Thus, on the durability observed as more vehicles are driven for for its assessment, the committee relied on the CAFE-GHG longer times (Steffke et al. 2008). One study that evaluated regulatory analysis (EPA/NHTSA 2012), as well as on pre- a PHEV with a 20 kWh battery showed that a hot climate sentations from vehicle manufacturers, suppliers, and market accelerates the normal degradation of battery capacity with analysts (Tamor 2012; Ward 2013; Woodard 2012; Sriramulu

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 41 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

26 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 2-3 Effect of ambient temperature on battery capacity on a 20 kWh battery in a PHEV. NOTE: DoD, depth of discharge; PHEV, plug-in hybrid electric vehicle. SOURCE: Pesaran et al. (2013).

and Barnett 2013; Anderman 2014), because a detailed in- more predictable. Until then, the price will likely depend dependent cost analysis was beyond its scope and resources. strongly on the availability of unused battery production ca- The committee also reviewed the cost information provided pacity and a manufacturer’s desire to be perceived as a tech- in Transitions to Alternative Vehicles and Fuels (NRC 2013). nology leader. It might further depend on the willingness of That committee estimated battery costs by assuming that fu- the vehicle manufacturer to set a price that allows it to gain a ture costs for Li-ion cells for vehicles would follow a similar, market share for its vehicles. although slower, cost reduction trajectory as that experienced Unfortunately, there are no definitive studies of battery by Li-ion 18650 cells. Although cost projections were some- costs from battery manufacturers given their need to protect what similar, this report makes use of the recent extensive proprietary information. The range of cost projections from analysis done specifically for the costs of vehicle Li-ion bat- studies of current and future battery costs is considerable. An teries. Costs of the batteries and nonbattery components are additional complication is that vehicle manufacturers make discussed below; vehicle price and cost of ownership are dis- different choices on how much of the total capacity of a bat- cussed further in Chapter 7. tery is made available for use; GM uses about 70 percent of the nominal capacity, and Nissan uses about 90 percent (see Lithium-Ion Battery Costs Table 2-1).4 To allow comparisons, the committee converted study results to be the projected costs per kilowatt-hour of the A high-energy battery costs much more than a sheet-met- total battery capacity rather than the available battery capac- al gasoline tank. Studies of current and projected battery costs ity. The costs estimated below are for complete battery packs, are summarized here to estimate the magnitude of the cost dif- excluding any cooling system. ferential and whether it is likely to continue. Cost refers to what a vehicle manufacturer would pay a supplier, which is • A 2012 Argonne National Lab study projected costs to known as the direct manufacturing cost (DMC) (EPA/NHTSA be between $251 and $280/kWh for a battery pack pro- 2012). What a consumer would pay for a battery (the retail duced in 2020 converted to 2012 dollars (Nelson et al. price equivalent) is expected in the automotive industry to 2011). be about 50 percent more than what a vehicle manufacturer 4 The values cited seem appropriate given that PHEV batteries would pay (NRC 2011). Large price fluctuations must be ex- could be cycled more times per trip than BEV batteries and that us- pected until battery supply and demand for PEVs becomes ing a smaller portion of the nominal capacity increases battery life.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 42 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 27

TABLE 2-3 Estimates of Dollars per Kilowatt-hour for a 25 kWh Battery Year Manufacturing Volume (packs/year) Cell Materials ($/kWh) Cell Price ($/kWh) Pack Price ($/kWh) 2013 25,000 110-150 275-325 400-500 2016 50,000 90-130 185-230 275-350 2020 100,000 85-110 140-190 225-275 SOURCE: Based on data from Anderman (2014).

• TIAX projected that direct material and direct labor more production capability than demand. Some reports sug- costs would amount to $310/kWh for an annual produc- gest that Tesla is paying much less for batteries from Panason- tion volume of 300,000, a large number compared with ic. In addition, Tesla has announced plans to build a $5 billion U.S. PEV sales to date (Sriramulu and Barnett 2013). battery factory and has stated that it believes it can substatially • DOE has estimated a current cost of $240/kWh (Howell reduce battery costs (Trefis Team 2014). The committee does 2013).5 not have any information about how the cost reductions will • EPA/NHTSA (2012) projected $540, $346, and $277/ be achieved, but the factory investment appears to be a strong kWh for a PHEV40 with a 16 kWh battery pack in 2017, indication that Tesla is confident that it can build high-energy 2020, and 2025, for an annual volume of 400,000. batteries more economically than has so far been possible. • A 2011 McKinsey study estimated the costs to be $350 to $420/kWh; it predicted that these costs would drop Finding: It is not possible to determine a completely reliable to about $140/kWh by 2020 and $112/kWh by 2025 projection of future battery cost. However, given the avail- (Hensley et al. 2012). able data, the committee assumed for this report a battery • Anderman (2012) predicted that the cost for a 24 kWh pack cost of $500/kWh in 2013 and a 50 percent lower cost battery pack in the 2015 time frame in volumes of in about 10 years. 100,000 units would be $340 to $450/kWh. • Anderman (2014) provided estimates of dollars per kilo- Finding: The high cost of high-energy batteries is primarily watt-hour for a 25 kWh battery (see Table 2-3). responsible for the higher initial cost of PEVs compared with HEVs and ICE vehicles and is a barrier to PEV adoption. An attempt has been made to convert study results to cost per kilowatt-hour of the total energy that can be stored in the Finding: Even if the higher initial battery cost drops as pre- battery and to 2013 dollars (see Table 2-4). dicted over the next 10 years, battery cost will remain a bar- The range of estimates in the current studies show that rier to PEV adoption. current costs are difficult to obtain and that the future projec- Nonbattery Costs tions are even more difficult, requiring, for example, an estimate of how many PEVs will be purchased. For the purposes An ICE vehicle includes an ICE, a radiator, a transmis- of this report, the committee decided to use the $500/kWh sion, and an oil system. A BEV has instead an electric motor; as the current cost of the lithium-ion battery pack and about power electronics that convert the direct current (dc) power $250/kWh as the cost in about 10 years. Thus, at $500/kWh, from the battery to the alternating current (ac) power needed the DMC of the Tesla battery would be $42,500, the DMC to drive the electric motor; and electronics needed to charge of the Leaf battery would be $12,000, the DMC of the Volt the battery. A PHEV includes both sets of components. The battery would be $8,250, and the DMC of the Plug-in Prius nonbattery costs of the PEV are primarily attributable to the battery would be $2,200. power electronic controls and the electric motor and genera- Figure 2-4 shows the decrease in costs of the Li-ion bat- tors. The committee reviewed and accepted the estimates for tery cell over the last 13 years and illustrates how Tesla has nonbattery costs from the EPA/NHTSA (2012) study that was profited from the reduced prices for the small cell package used to evaluate CAFE standards because it found that the cost used to power consumer electronics. The recent prices shown analysis performed by the agencies was thorough and compre- for Li-ion batteries in Figure 2-4 ($400/kWh) correspond to a hensive. cost of about $270/kWh if the assumption mentioned earlier The simplicity of a BEV compared with an ICE vehicle is used that price is 1.5 times the cost. Some care is required makes it somewhat surprising that the EPA/NHTSA (2012) in deducing cost from prices in recent years because battery study estimates that the direct manufacturing cost of the manufacturers might be reducing prices to cope with having nonbattery components for a BEV with a range of 75 miles 5 A current cost estimated to be $300/kWh becomes $240/kWh for is about $1,255 higher than the cost of the ICE power-train the total battery capacity, assuming that the original estimate was components it replaces. The increased cost includes $3,810 for an 80 percent utilization of the battery.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 43 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

28 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

TABLE 2-4 Summary of Estimated Costs of Total Energy from Various Sources (2013 U.S.$/kWh) Year Source Currenta 2017 2020 2022 2025 Argonne 2000 250-706 — — — — 2012 — — 50 kW = 336 — — 100 kW = 404 TIAX 2013 310 — — — — DOE 2013 300 — — 125 — EPA/NHTSA 2012 — 540 346 — 277 McKinsey 2011 350-420 — 140 — 112 Anderman 2012 340-450 — — — — 2014 400-500 — 220-275 — — a Current as defined in the respective studies.

1800.0 Change of the sales price of the battery（＄/kWh）

1600.0 NiMH

1400.0 Li‐ion NiCd

1200.0

1000.0 ） /kWh ＄（ 800.0 Cost

600.0

400.0

200.0

0.0 Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Jan Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep Sep May May May May May May May May May May May May May May 99 0 01 02 03 04 05 06 07 08 09 10 11 12

FIGURE 2-4 Change in the sales price of NiMH, Li-ion, and NiCd battery cells from 1999 to 2012. Prices are shown in 2012 dollars. The graph is based on data from a production survey conducted by the Ministry of Economy, Trade, and Indus- try, Japan. NOTE: kWh, kilowatt-hour; Li-ion, lithium ion; NiCd, nickel cadmium; NiMH, nickel-metal hydride. SOURCE: Maruyama (2013).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 44 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 29 for the electric motor, inverter, high-voltage wiring, and im- battery life. Although increasing the charging rate with high- provements in the climate-control system. Those component power chargers shortens the time needed to charge a vehicle’s costs are partially offset by the elimination of the ICE, trans- battery, an important technical issue now being researched is mission, and related components, which account for a sav- the extent to which faster charging at high power hastens the ings of $2,555 in direct manufacturing costs (EPA/NHTSA normal aging of a battery (Francfort 2013). 2012). The EPA/NHTSA estimates that the nonbattery costs The “pressure” with which an electric circuit in a home in 2025 will drop to 80 percent of their 2012 costs. However, or business can force electricity through wires into some de- even if the cost reduction is less, the cost of the high-energy vice is measured in volts (V). The amount of electricity flow- battery will still account for most of the difference in cost ing through various devices, the electric current, is measured between a BEV and an ICE vehicle. in amperes (A). The product of the two is the power in watts Because a PHEV has both an electric drive and an ICE, (W). Every circuit delivering electricity has a circuit breaker it has a higher nonbattery cost. The same study evaluated a or fuse that keeps the flow of electricity from exceeding the PHEV with a 40-mile AER and concluded that a PHEV has amperes that the circuit can safely provide. For example, a nonbattery cost that is $3,700 higher than the nonbattery cost 2014 Nissan Leaf is capable of accepting no more than 30 A of an ICE vehicle. Multiplying by 1.5 increases the price to of electric current when it is connected to a 240 V electric the consumer to $5,550 beyond the price of the battery. circuit, so its maximum power consumption is 7.2 kW. The A dramatic reduction in the price of power inverters vehicle will not accept more current or power even if the could potentially come from the replacement of silicon- circuit is able to provide it. The circuit is protected by a 40 based semiconductors by wide bandgap materials, such as A circuit breaker, resulting in what is referred to as a 240 V, SiC and GaAs, that would enable faster switching and lower 40 A service. resistance to improve the inverter efficiency. Those materi- As recommended by the National Electrical Code (NEC), als operate at much higher temperatures than the silicon used an apparatus known as the electric vehicle supply equipment in today’s power electronics, and that characteristic would (EVSE) is always connected between the charging circuit make cooling easier and thereby reduce the size of the power and the vehicle to protect the people and the vehicle during electronics package and possibly simplify the heat exchang- charging. The purpose of the EVSE is to create two-way com- ers (ORNL 2012). However, when such technology will be munication between vehicle and charger before and during far enough along to come to market is difficult to predict. charging to detect any anomalies that might affect safety or the equipment (Rawson and Kateley 1998). The NEC (2008) Finding: Because power electronics and large electric mo- defines the EVSE as “the conductors, including the unground- tors are new to the automotive industry, nonbattery costs will ed, grounded, and equipment grounding conductors and the likely drop substantially as new models come to market. electric vehicle connectors, attachment plugs, and all other fit- tings, devices, power outlets or apparatus installed specifically VEHICLE CHARGING AND CHARGING OPTIONS for the purpose of delivering energy from the premise’s wiring to the electric vehicle” (Section 625.2). Its ground fault in- Charging a PEV is analogous to filling a conventional terrupters—like those in bathrooms and kitchens—are safety vehicle’s fuel tank with gasoline. A gasoline-powered vehi- devices that can detect when a small electric current from the cle is attached to a pump that sends gasoline through a hose circuit has “gone missing” and disconnect the electric circuit into the fuel tank. A typical flow rate of 8 gal/min, for ex- and the current flow before anyone is injured. Furthermore, ample, means that typical gasoline tanks with capacities of the EVSE is able to communicate with a vehicle to ensure 10 to 20 gal will be filled in a few minutes. Similarly, a PEV that no current is provided before the vehicle is connected. is plugged into the electric grid so that electricity can flow The EVSE for slow charging via 120 V is typically a portable through wires into the battery. An energy flow rate of 6.6 kW, device that can be carried in the vehicle for possible use at re- for example, would fill an empty battery with a usable capac- mote locations. The EVSE for normal 240 V charging is typi- ity of 21 kWh in about 4 hours. cally mounted on a garage wall or on a purpose-built column. The maximum charging rate for residential charging is Fast chargers that use high dc voltages have the EVSE built limited by the size of the charger in the vehicle that changes into the substantial charger that is required. ac electricity into dc electricity. A fully discharged battery ini- For EVSEs connected to the single phase 120 V ac or the tially charges at the maximum rate that the onboard charger split-phase 240 V ac circuits that are commonly available in can manage and then charges more slowly as the battery nears U.S. homes and workplaces, a plug wired to the EVSE con- capacity. Thus, a vehicle battery does not charge at a constant nects to a socket on the vehicle. The circuit breaker or fuse rate, and that is why it takes about 4 hours to fill a 21 kWh sets the maximum current that the EVSE can provide, al- battery at 6.6 kW. For DC fast charging (discussed below), though individual vehicles will typically accept less current. the component that changes ac to dc is outside the vehicle and In the United States, there is one standard plug that is used is governed by control signals from the vehicle. Regulating to charge vehicles from the normal 120 V and 240 V circuits the charging rate is necessary to ensure safety and to protect found in residences, the SAE J1772 standard (SAE 2012).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 45 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

30 Overcoming Barriers to Deployment of Plug-in Electric Vehicles This interchangeability removes what otherwise could be 120 V circuit, the vehicle that is charging must typically be the a substantial barrier to the adoption of PEVs. However, for only device drawing current from the circuit to avoid exceed- faster charging options, fast chargers are being installed that ing the maximum current that the circuit breaker or fuse will have one or more of three incompatible plugs and protocols allow the circuit to provide. described below. PEVs are typically sold with a small and portable EVSE that can be carried in the car to allow AC level 1 charging AC Level 1 Charging from ubiquitous 120 V wall receptacles. A deficiency of the standard is that the portable EVSE is not secured to either the Most electric devices in the United States (for example, 120 V socket or to the vehicle to deter EVSE theft or vandal- lamps, small air conditioners, and computers) are plugged into ism. AC level 1 charging with this EVSE is the only charging single-phase 120 V ac electric circuits accessed via the wall option typically needed or available for the minimal PHEVs. sockets present in essentially every room of every building. Each hour of charging typically provides an additional elec- Circuit breakers or fuses switch off the electricity if the current tric range of about 4 to 5 miles, depending on the vehicle. flowing through the circuit exceeds 15 to 20 A to prevent fires For a range-extended PHEV, such as a Chevrolet Volt, some and other damage to the circuits. drivers use only AC level 1 charging, while others prefer to AC level 1 charging standard is for an EVSE that plugs charge about twice as fast using the AC level 2 charging that into a 120 V wall plug (Figure 2-5) and delivers up to 12 A to is discussed below. a SAE J1772 plug (Figure 2-6), which connects with a socket For charging the fully depleted batteries of PEVs with in the car. Most PEVs today have an onboard charger that large batteries, AC level 1 charging is too slow to be the changes the ac current into the dc current that charges the bat- primary charging method because charging times could be tery. The charger is able to accept only up to 12 A from the longer than the time that a car is parked at the home or work- EVSE and transfer energy at a rate of up to 1.4 kW. Much like place. For example, with an AC level 1 charger, the nomi- the largest window air conditioners that can be plugged into a nal time for fully charging the usable 21 kWh capacity of a

FIGURE 2-5 For AC level 1, a vehicle is plugged into a single-phase 120 V electric socket through a portable safety device called an electric vehicle supply equipment (EVSE).

FIGURE 2-6 The SAE J1772 plug that connects all PEVs to AC level 1 and level 2 is an agreed-on universal standard for 120 V and 240 V ac charging. SOURCE: © Michael Hicks, licensed under Creative Commons 2.0 (CC-BY-2.0).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 46 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 31 Nissan Leaf battery is more than 17 hours, and the nominal Tesla Model S also supplies a portable EVSE with adapters time to fully charge the 85 kWh battery of a Tesla Model S that allow it to be charged using most of the common 240 V is more than 61 hours. However, AC level 1 charging could wall sockets that deliver 24 or 40 A to electric dryers, stoves, be useful in some cases to merely extend the range of those and air conditioners. BEVs by a few miles if that is all that is needed. DC Fast Charging AC Level 2 Charging Faster charging is generally carried out by supplying a AC level 2 charging uses a 240 V, split-phase ac cir- high dc voltage directly to the battery. In this case, the char- cuit (Figure 2-7). Such circuits are available in essentially ger that turns the ac electricity available from the grid into all homes and workplaces and are used by electric dryers, the dc electricity required to charge the battery is located in electric stoves and ovens, and large air conditioners. Since the EVSE rather than within the car. Such charging is only 2009, the AC level 2 standard allows up to 80 A of current to useful for limited-range and long-range BEVs, such as the be delivered for an energy transfer rate of 19 kW, although Nissan Leaf and the Tesla Model S, and only BEVs are typi- the wiring in many houses will have trouble delivering that cally able to accept fast charging. much current, and only a long-range BEV is capable of ac- A proliferation of incompatible connector (and proto- cepting it. A Chevrolet Volt and a 2014 Nissan Leaf are able col) standards are used for the DC fast chargers. Four options to accept a maximum of 12 A or 30 A, respectively, which are being offered worldwide (Figure 2-8), three of which are corresponds to energy being transferred at maximum rates of becoming increasingly available in the United States. 3.3 and 7.2 kW, respectively. As noted, the 240 V EVSE for All fast chargers installed in the United States so far are AC level 2 charging is typically wall-mounted in a garage or CHAdeMO chargers with the exception of the Tesla super- on a post next to a parking spot, and in the United States, it is chargers.6 The Nissan Leaf accepts a CHAdeMO plug (Figure connected to the vehicle through the same SAE J1772 plug 2-8A), which provides the high voltage dc and control signals (Figure 2-6) used for AC leve1 1 charging. to the vehicle. A 44 kW CHAdeMO charger can charge a Nis- The 85 kWh battery of the Tesla Model S, much larger san Leaf to 80 percent of its capacity in 30 minutes (see Figure than the battery in any other PEV, is the only vehicle battery 2-9). so far that can accept the highest rated current and power The Tesla Model S accepts a proprietary fast-charging from an AC level 2 charging system. The normal home plug (Figure 2-8C), and charges are free at Tesla supercharg- charging recommendation is to deliver 40 A and nearly 10 ers for models with an 85 kWh battery (that is, such charging kW to a “single” charger installed in the Tesla Model S. If is included in the purchase price of the vehicle). Existing enough current is available in a home, a “double” charger 90 kW superchargers are being upgraded to 120 kW so that can instead be installed in the car to accept 80 A and 19 kW power for much faster charging. With that option, Tesla ad- 6 In October 2014, the total number of CHAdeMO chargers world- vertises that the car can travel an additional 58 miles for wide was 4,180, with the following breakdown: Japan, 2,129; Eu- each hour of charging (Tesla 2014b). For emergency use, the rope, 1,327; United States, 700; and other, 24 (CHAdeMO 2014).

FIGURE 2-7 For AC level 2 charging, a vehicle is plugged into a split-phase 240 V electric circuit like those used by electric dryers, stoves, and large air conditioners through a wall- or post-mounted safety device called an electric vehicle supply equipment (EVSE).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 47 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

32 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 2-8 Four plugs and control protocols are now being used for DC fast charging: (A) the CHAdeMO plug that is used for the Nissan Leaf; (B) the SAE J1772 combo standard that is used on the BMW i3 and the Chevrolet Spark. The upper part of the connector is the same as the SAE J1772 plug that is used universally in the United States for AC level 2 charging (see Figure 2-6); (C) the proprietary Tesla plug that is used for the Tesla supercharger network; and (D) the Mennekes plug recently adopted by the European Union for use in Europe. SOURCE: (A) © C-Car-Tom, licensed under Creative Commons 3.0 (CC-BY-3.0); (B) SAE (2012), reprinted with permission from SAE J1772 Feb2012 © 2012 SAE International; and (D) © loremo, licensed under Creative Commons 2.0 (CC-BY-2.0).

FIGURE 2-9 DC fast charging a Nissan Leaf. DC fast charging is able to charge a Nissan Leaf battery to 80 percent capacity in 30 min. The charge would typically allow a 2014 Nissan Leaf to travel about 67 miles. SOURCE: Copyright © 2010 by the eVgo Network, licensed under Creative Commons 2.0 (CC-BY-2.0).

the battery can be charged to 50 percent of its capacity in The variety of DC fast-charging plugs and communica- as little as 20 minutes. The announced goal is to install 250 tion protocols seems unfortunate. For long-range BEVs, the units so that 98 percent of U.S. drivers are within 100 miles future situation could be like having separate networks of of a supercharger by the end of 2015 (Tesla 2014c). The lo- gasoline stations for ICE vehicles made by different manu- cations of the superchargers are shown in Figure 2-10. Tesla facturers. It is not a big problem now in that the Tesla Model chargers will not be available to drivers of other long-range S is the only long-range BEV able to make long trips us- BEVs when these become available. ing the proprietary network of Tesla superchargers. As other The SAE added a dc and a ground lead to the SAE J1772 manufacturers introduce long-range BEVs, however, they plug universally used for AC level 2 charging (Figure 2-6) to might need to introduce their own charger networks to com- make a J1772 combo plug (Figure 2-8B). There are almost pete. The United States and proactive states like California no installed combo chargers in the United States to date and might be able to use their influence and incentives to make it few PEVs that are able to use them. However, the Chevrolet possible to fast charge any PEV at any fast-charging station. Spark and the BMW i3 that is just becoming available in the The United States could raise the issue of compatible charger United States use them. designs in free trade talks with the European Union and with The European Union recently adopted the Mennekes its trading partners in Asia. (Masson 2013) plug (Figure 2-8D) for its 240 V AC level 2 standard for charging rates up to 39 kW. That standard is not Finding: A network of fast-charging stations is currently discussed in detail because it is not expected to be used in being completed by Tesla without the use of public funds. the United States. However, it is a proprietary network that might not be avail-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 48 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 33

FIGURE 2-10 As of February 2015, Tesla had installed 190 units in the United States. SOURCE: DOE (2015).

able for the use of all drivers when more long-range BEVs Wireless charging would instead transfer the energy come to market. from the grid to the vehicle by using inductive coupling between a wireless transmitter located near the vehicle and a Finding: The various plugs and communication protocols wireless receiver attached to the vehicle (Miller et al. 2014). that are used across the world for charging PEVs are a bar- An alternating magnetic field produced by passing ac cur- rier to the adoption of PEVs insofar as they prevent all PEVs rent through coils in the wireless transmitter would induce a from being able to charge at any fast-charging station. voltage in the coil of the receiver. The latter currents would charge the vehicle battery. Static and dynamic wireless Recommendation: The federal government and proactive charging are possible. states should use their incentives and regulatory powers to Static wireless charging takes place when the vehicle is (1) eliminate the proliferation of plugs and communication not moving, as described. The energy transfer is less efficient protocols for DC fast chargers and (2) ensure that all PEV than using a charging cable, but there would be no cable to drivers can charge their vehicles and pay at all public charg- handle or keep clean. For publicly available charging, stan- ing stations using a universally accepted payment method dards would be needed to make it possible to charge most just as any ICE vehicle can be fueled at any gasoline sta- PEVs with most wireless charging systems. The opportu- tion. The Society of Automotive Engineers, the International nity for theft or vandalism of the cable or EVSE is greatly Electrotechnical Commission, and the Verband der Elektro- reduced because the transmitter could be embedded in the technik—companies that formed CHAdeMO—and Tesla parking space and controlled remotely. A safety standard to should be included in the deliberations on plugs and com- establish the acceptable levels of oscillating electromagnetic munication protocols. fields might also be needed. Dynamic wireless charging is a futuristic concept that is Wireless Charging being investigated to see if it might ever be feasible (Miller et al. 2014). The vision is that a vehicle would receive power So far, essentially all PEVs are charged by plugging a in its wireless receiver as it passed long series of wireless charging cable into the vehicle so that electricity can flow transmitters, so a BEV could be refueled on long trips with- from the EVSE to the battery. The process is simple and rap- out stopping to refuel. However, there are many technical id (less than a minute), and control electronics are included problems to overcome for dynamic wireless charging, one of to enhance safety. them being a very low charging efficiency.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 49 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

34 Overcoming Barriers to Deployment of Plug-in Electric Vehicles REFERENCES DOE/EPA (U.S. Department of Energy and U.S. Environmen- tal Protection Agency). 2014a. “Compare Side-by-Side: Amine, K. 2010. “Advanced Battery Chemistries for PHEV 2014 Tesla Model S (85 kW-hr battery pack).” http://www. and EV Applications.” Presentation at the U.S.- China fueleconomy.gov/feg/Find.do?action=sbs&id=34775. Electric Vehicle and Battery Technology Workshop, Accessed January 26, 2015. Argonne, IL, August 30-September 1. http://www.trans- DOE/EPA. 2014b. “Compare Side-by-Side: 2014 Nissan portation.anl.gov/batteries/us_china_conference/docs/ Leaf.” http://www.fueleconomy.gov/feg/Find.do?action roundtable1/adv_battery_chem_amine.pdf. =sbs&id=34699. Accessed January 26, 2015. Anderman, M. 2012. “Energy Storage for xEVs: Status, DOE/EPA. 2014c. “Compare Side-by-Side: 2014 Ford Focus Trends and Challenges.” Presentation at the 24th Inter- Electric.” http://www.fueleconomy.gov/feg/Find.do?act national AVL Powertrain Conference Engine and Envi- ion=sbs&id=34130. Accessed January 26, 2015. ronment, Graz, Austria, September 12. DOE/EPA. 2014d. “Compare Side-by-Side: 2014 Chevro- Anderman, M. 2014. “Li-Ion Technology Evolution for let Volt.” http://www.fueleconomy.gov/feg/Find.do?act xEVs: How Far and How Fast?” Presentation to the ion=sbs&id=33900. Accessed January 26, 2015. Committee on Overcoming Barriers to Electric-Vehicle DOE/EPA. 2014e. “Compare Side-by-Side: 2014 Toyota Deployment, Irvine, CA, February 25. Prius Plug-in Hybrid.” http://www.fueleconomy.gov/fe Blanco, S. 2014. “Chevy Volt Owners Log Half a Billion g/Find.do?action=sbs&id=34516. Accessed January 26, Electric Miles, 2015 Production Starts.” Autoblog- 2015. Green, June 19. http://green.autoblog.com/2014/06/19/ Duoba, M. 2012. “Evaluating Plug-In Vehicles (PHEV & chevy-volt-half-billion-miles-2015-production/. BEV) Using Standard Dynamometer Protocols.” Presen- CCSE (California Center for Sustainable Energy). 2014. tation at the 6th U.S.-China Electric Vehicles and Bat- February 2014 Survey Report. http://energycenter.org/ tery Technology Workshop, Boston, MA, August 22-24. clean-vehicle-rebate-project/vehicle-owner-survey/feb- http://www.cse.anl.gov/us-china-workshop-2012/pdfs/ 2014-survey. session4b_demos_standards/duoba_4B-2-Duoba-ANL- CHAdeMO. 2014. “CHAdeMO’s Fast Charging Station in Standardizing-Vehicle-Dyno-Test-Aug22-2012.pdf. the World.” http://www.chademo.com. Accessed Octo- EPA (U.S. Environmental Protection Agency). 2012. “Six ber 1, 2014. Common Air Pollutants.” http://www.epa.gov/oar/oaqps/ Chernicoff, W.P. 2014. Perspectives on Electric Vehicles urbanair. Accessed November 14, 2014. from Toyota. Presentation to the Committee on Over- EPA. 2014. “2014 Fuel Economy Datafile.” Office of Trans- coming Barriers to Electric-Vehicle Deployment, Irvine, portation and Air Quality. http://www.fueleconomy.gov/ CA, February 25. feg/download.shtml. Accessed January 26, 2015. Consumer Reports. 2014. “Tesla Model S 2013- 2014 Quick EPA/NHTSA (U.S. Environmental Protection Agency and Take” [video file]. http://www.consumerreports.org/cro/ National Highway Traffic Safety Administration). 2012. video-hub/cars/hybrids--alternative-fuel/tesla-model-s- Joint Technical Support Document, Final Rulemaking 20132014-quick-take/14786539001/2366240882001/. 2017-2025 Light-Duty Greenhouse Gas Emission Stan- Accessed March 28, 2014. dards and Corporate Average Fuel Economy Standards. Daniel, C. 2014. “Lithium Ion Batteries and Their Manu- Report No. EPA-420-R-12-901. facturing Challenges.” Presentation at NAE Frontiers of FHWA (Federal Highway Administration). 2011. Summary Engineering, Irvine, CA, September 11-13. http://www. of Travel Trends: 2009 National Household Travel Sur- naefrontiers.org/File.aspx?id=46415. vey. FHWA-PL-11-02. U.S. Department of Transporta- DOE (U.S. Department of Energy). 2012. 2011 Nissan Leaf tion. http://nhts.ornl.gov/2009/pub/stt.pdf. -VIN 0356 Advanced Vehicle Testing—Beginning-of- Ford Motor Company. 2014. “2014 Ford Focus Electric Pow- Test Battery Testing Results. Vehicle Technologies Pro- er and Handling.” http://www.ford.com/cars/focus/trim/ gram. http://www1.eere.energy.gov/vehiclesandfuels/av electric/. Accessed January 26, 2015. ta/pdfs/fsev/battery_leaf_0356.pdf. Francfort, J. 2013. “U.S. Department of Energy’s Vehicle DOE. 2013. 2013 Chevrolet Volt-VIN 3929 Advanced Ve- Technologies Program. Plug-in Vehicles and Charging hicle Testing—Beginning-of-Test Battery Testing Re- Infrastructure Usage Patterns: Lessons Learned from the sults. Vehicle Technologies Program. http://www1.eere. First Two Years.” Presentation at SAE Government/In- energy.gov/vehiclesandfuels/avta/pdfs/phev/battery_ dustry Meeting, Washington, DC, January 31. http://avt. volt_3929.pdf. inel.gov/pdf/prog_info/SAEGovtIndustryFeb2013. DOE. 2015. “Alternative Fueling Station Locator.” Alter- pdf. native Fuels Data Center. http://www.afdc.energy.gov/ Ge, M., X. Fang, J. Rong, and C. Zhou. 2013. Review of po- locator/stations/. rous silicon preparation and its application for lithium- ion battery anodes. Nanotechnology 24(42): 422001.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 50 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Plug-in Electric Vehicles and Charging Technologies 35 Goodenough, J.B., and K. Mizushima. 1981. “Electrochemi- http://www.cse.anl.gov/batpac/files/BatPaC%20ANL cal Cell with New Fast Ion Conductors.” U.S. Patent -12_55.pdf. 4,302,518, filed March 31, 1980, and issued November NRC (National Research Council). 2011. Assessment of 24, 1981. Fuel Economy Technologies for Light-Duty Vehicles. Gordon-Bloomfield, N. 2013. “Nissan Testing New Battery Washington, DC: The National Academies Press. Pack to Address Problems with LEAF in Hot Weather.” NRC. 2013. Transitions to Alternative Vehicles and Fuels. Plug-in Cars,August 26. http://www.plugincars.com/nissan- Washington, DC: The National Academies Press. testing-new-battery-pack-leaf-128088.html. NREL (National Renewable Energy Laboratory). 2013. Hensley, R., J. Newman, and M. Rogers. 2012. “Battery Tech- “Plug-In Hybrid Electric Vehicles.” Vehicle Systems nology Charges Ahead.” McKinsey Quarterly, July. http:// Analysis. http://www.nrel.gov/vehiclesandfuels/vsa/plu www.mckinsey.com/insights/energy_resources_materials/ gin_hybrid.html. Accessed March 14, 2013. battery_technology_charges_ahead. ORNL (Oak Ridge National Laboratory). 2012. Wide Band Hidary, J. 2012. “New Models of Mobility and EV Deploy- Gap Devices: Powering the Next Generation of Elec- ment.” Presentation to the Committee on Overcoming tric Traction Drive Systems. http://web.ornl.gov/sci/ees/ Barriers to Electric-Vehicle Deployment, Washington, transportation/pdfs/WBGBroch.pdf. DC, December 18. Pesaran A., S. Santhanagopalan, and G. Kim. 2013. “Ad- Howell, D. 2013. U.S. Battery R&D Progress and Plans. U.S. dressing the Impact of Temperature Extremes on Large Department of Energy. http://energy.gov/sites/prod/files/ Format Li-Ion Batteries for Vehicle Applications.” Pre- 2014/03/f13/es000_howell_2013_o.pdf. sentation at the 30th International Battery Seminar, Ft. Julien, C.M., A. Mauger, K. Zaghib, and H. Groult. 2014. Lauderdale, FL, March 11. http://www.nrel.gov/vehicles Comparative issues of cathode materials for Li-ion bat- andfuels/energystorage/publications.html. teries. Inorganics 2(1): 132-154. Pham, H., M. Lee, K. Nam, E. Hwang, Y. Kwon, and S. Kam, K., and M. Doeff. 2012. Electrode materials for lithium Song. 2014. “High-Voltage Electrolyte Additive for ion batteries. Materials Matters 7(4). http://www.sigma High-Energy Lithium-Ion Batteries.” Presentation at the aldrich.com/technical-documents/articles/material- 17th International Meeting on Lithium Batteries, Como, matters/electrode-materials-for-lithium-ion-batteries. Italy, June 13. https://ecs.confex.com/ecs/imlb2014/web html. program/Paper29242.html. MacKenzie, A. 2013. “2013 Motor Trend Car of the Year: PolyPlus Battery Company. 2009. “Advanced Lithium Bat- Tesla Model S Shocking winner: Proof Positive That tery Technology.” http://www.polyplus.com/liair.html. America Can Still Make (Great) Things.” Motor- Accessed October 25, 2013. Trend, January. http://www.motortrend.com/oftheyear/ Rahman, M.A., X. Wang, and C. Wen. 2014. A review of car/1301_2013_motor_trend_car_of_the_year_tesla_ high energy density lithium-air battery technology. model_s/. Journal of Applied Electrochemistry 44(1): 5-22. Maruyama, T. 2013. “Change of Sales Price of the Battery Rawson, M., and S. Kateley. 1998. “Electric Vehicle Charging from January 1999 to September 2012.” Presentation Equipment Design and Health and Safety Codes.” Cali- to the Committee on Overcoming Barriers to Electric- fornia Energy Commission. Sacramento, CA, August 31. Vehicle Deployment, Tokyo, Japan, December 9. http://www.energy.ca.gov/papers/98-09-23_KATELEY. Masson, L. 2013. “European Commission Backs Meenekes PDF. Type 2 Electric Car Plug.” Plug-in Cars, January 30. SAE (Society of Automotive Engineers). 2012. “SAE Elec- http://www.plugincars.com/european-commission- tric Vehicle and Plug in Hybrid Electric Vehicle Con- wants-act-help-evs-126265.html. ductive Charge Coupler.” Standard Code: J1772, Revi- Miller, J.M, O.C. Onar, C. White, S. Campbell, C. Coomer, sion B. SAE International, October 15. http://standards. L. Seiber, R. Sepe, and A. Steyerl. 2014. Demonstrating sae.org/j1772_201210/. dynamic wireless charging of an electric vehicle: The Sakti, A., J. Michalek, E. Fuchs, and J. Whitacre. 2014. A benefit of electrochemical capacitor smoothing. IEEE techno-economic analysis and optimization of Li-ion Power Electronics Magazine 1(1): 12-24. batteries for light-duty passenger vehicle electrification. NEC (National Electrical Code). 2008. NEC2008 National Journal of Power Sources 273: 966-980. Electrical Code, Article 625. Electrical Vehicle Charging Schey, S. 2013. The EV Project: Q2 2013 Report. Electric Trans- System Equipment, Section 625-2. http://www.freenec. portation Engineering Corporation. http://www.theevpro- com/T504.html. ject.com/cms-assets/documents/127233-901153.q2- Nelson, P.A., K.G. Gallagher, I. Bloom, and D.W. Dees. 2011. 2013-rpt.pdf. Modeling the Performance and Cost of Lithium-Ion Sion Power. 2014. “Technology Overview.” http://sionpow- Batteries for Electric-Drive Vehicles. No. ANL-12/55. er.com/technology.html. Accessed October 25, 2014.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 51 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

36 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Sriramulu, S., and B. Barnett. 2013. “Technical, Manufactur- Toyota Motor Corporation. 2014. “2014 Prius Plug-In.” http:// ing, and Market Issues Associated with xEV Batteries.” www.toyota.com/content/ebrochure/2014/prius-plug-in_ Presentation to the Committee on Overcoming Barriers ebrochure.pdf. to Electric-Vehicle Deployment, Washington, DC, Au- Trefis Team. 2014. “Gigafactory Will Cost Tesla $5Bil- gust 13. lion but Offers Significant Cost Reductions.” Forbes, Steffke, K., S. Inguva, D. Van Cleve, and J. Knockeart. 2008. March 11. http://www.forbes.com/sites/greatspeculation “Accelerated Life Test Methodology for Li-Ion Batter- s/2014/03/11/gigafactory-will-cost-tesla-5-billion-but- ies in Automotive Applications.” SAE Technical Paper offers-significant-cost-reductions/. 2013-01-1548, doi:10.4271/2013-01-1548. Ward, J. 2013. “DOE Electric Vehicle Activities Update.” Tamor, M. 2012. “Overcoming Barriers to Deployment of Presentation to the Committee on Overcoming Barriers Electric Vehicles NAE Committee Discussion.” Pre- to Electric-Vehicle Deployment, Washington, DC, Au- sentation to the Committee on Overcoming Barriers to gust 13. Electric Vehicle Deployment, Washington, DC, Decem- Woodard, T. 2012. “Overcoming Barriers to Electric-Vehi- ber 17. cle Deployment Workshop Barriers to Deployment, an Tesla. 2014a. “Increasing Energy Density Means Increasing OEM Perspective.” Presentation to the Committee on Range.” http://www.teslamotors.com/roadster/technolgy/ Overcoming Barriers to Electric-Vehicle Deployment, battery. Accessed on March 30, 2014. Washington, DC, December 17. Tesla. 2014b. “Charge Your Model S.” http://www.teslamotors. Yoshino, A. 2012. The birth of the lithiumion battery. An- com/charging#/basics. Accessed November 14, 2014. gewandte Chemie International Edition 51(24): 5798- Tesla. 2014c. “100 Supercharger Stations.” http://www.teslamot 5800. ors.com/blog/100-supercharger-stations.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 52 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles

The process of buying a vehicle is a complex, highly This chapter begins with a general discussion of models involved consumer decision (Solomon 2014). A vehicle is of adoption and diffusion of innovation. It presents evidence one of the most expensive purchases made by individuals on how new technologies are adopted by various categories or households, often equal to many months or even years of of customers and discusses the factors that affect the pace income, and will last for many years. As a result, consum- of adoption and diffusion of a new technology through soci- ers perceive the decision to be a relatively risky one and will ety. Next, the chapter discusses consumer demographics and strive to ensure a “safe” decision so that they are not stuck evaluates the implications of that information and other fac- with a poor purchase choice for years to come. In general, tors that affect adoption and diffusion of PEVs. The chapter consumers want vehicles that are affordable, safe, reliable, then reviews what motivates the purchases of mainstream and comfortable for travel and meet many practical needs, consumers and possible barriers for their adoption of PEVs. such as getting them to work, school, stores, and recreation Next, the chapter reviews strategies for addressing consumer and vacation areas. Some also want vehicles to meet their concerns and describes government efforts to familiarize the psychosocial needs; for example, vehicles can serve as sta- public with PEVs. Throughout the chapter, at the conclu- tus symbols that represent one’s success or self-image. For sions of the various sections, the committee highlights rel- all these reasons, consumers generally will undertake lengthy evant findings. Recommendations for addressing consumer research into their options to ensure a good choice that satis- perceptions (or misperceptions) and barriers to adoption are fies all their various needs. presented in a section dedicated to overcoming the challeng- Plug-in electric vehicles (PEVs) must compete effec- es. The committee notes that the chapter focuses primarily tively with internal-combustion engine (ICE) vehicles in on private (individual) new vehicle buyers, who are respon- meeting consumer needs. However, PEVs, many of which sible for about 80 percent of all new vehicle purchases. Fleet are in their first generation of deployment, add complexity sales, which average 20-22 percent of the U.S. market (Au- and uncertainty to the consumer’s multistep and potentially tomotive Fleet 2013), are addressed at the conclusion of this time-consuming process of purchasing a vehicle. Under chapter. conditions of uncertainty and perceived risk, consumers tend to gravitate to the known and familiar. That observation UNDERSTANDING AND PREDICTING THE is well-documented in the literature, particularly in Daniel ADOPTION OF NEW TECHNOLOGIES Kahneman’s (2013) work, Thinking Fast and Slow, which spurred much recent work in behavioral economics. Because Models for the Adoption of Innovative Products innovative products require a higher degree of learning than Developers of new technologies generally, and of PEVs existing products, the effort customers must put into the de- specifically, face challenges in developing a market and mo- cision process is greater than for more familiar products. To tivating consumers to purchase or use their products (Mohr unseat incumbent technologies, the new technology must et al. 2010). Incumbent technologies—in this case, ICE ve- offer advantages and benefits sufficient to offset any price hicles—can be difficult to unseat; they have years of pro- differential and the perceived risk and uncertainty of pur- duction and design experience, which make their production chasing an innovation (Aggarwal et al. 1998). Thus, the costs lower than those of emerging technologies and thus committee emphasizes that consumer considerations loom more affordable. In addition, ICE vehicle technology is con- large for the deployment of PEVs in the nation’s transporta- tinuously improving; many of these improvements, which tion mix, and understanding consumer perceptions, knowl- are being made to meet tighter fuel economy and greenhouse edge, and behavior are key to crafting viable strategies for gas emission standards (EPA/NHTSA 2012), are described successful commercialization of PEVs. in the NRC report Transitions to Alternative Vehicles and

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 53 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

38 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Fuels (NRC 2013a) . The necessary infrastructure—includ- tively inexpensive and practical item with no complicated ing dealerships, service stations, roadside assistance, and the infrastructure needs—took 15 years to reach just 50 percent ubiquity of over 100,000 gasoline stations across the United market penetration. Consumers did not have experience States (U.S. Census Bureau 2012)—is also well developed. with microwave ovens nor did they initially see the value Consumers know the attributes and features to compare to or usefulness of such a product; its means of cooking was evaluate their ICE-vehicle choices, and they are accustomed not understood, and it did a poor job of “baking” compared to buying, driving, and fueling these vehicles. Indeed, one with conventional ovens. Indeed, calling the microwave an of the main challenges to PEV adoption is how accustomed “oven” was probably an error, as that term confused con- people are to ICE vehicles. sumers about the microwave’s functions. Initial uses of the Traditional consumer-adoption models predict the dif- microwave were to heat water, thaw and heat frozen food, fusion of new innovations through society (Parasuraman and reheat leftovers—few of these tasks had much to do with and Colby 2001; Rogers 2003; Moore 2014). The models how conventional ovens were used. It took many years to are well established and empirically validated across many educate the consumer about exactly what a microwave could product categories (Sultan et al. 1990) and can help in under- do. Consumer knowledge, societal lifestyle changes, and standing the consumer purchase decision and market devel- lower prices due to volume production over decades resulted opment process for PEVs. As stated in Chapter 1, PEV sales in microwaves being a primary appliance in the household, reached about 0.76 percent of the U.S. market in 2014 (Cobb nearly 20 full years after they were first introduced. 2015). To put that in perspective, it took 13 years for hybrid One insight is that adoption and diffusion of new inno- electric vehicles (HEVs) to exceed 3 percent of annual new vations can be a long-term, complicated process that is espe- light-duty vehicle sales in the United States (Cobb 2013).1 cially slow for products that cost tens of thousands of dollars To compare various rates of market penetration, Figure and where consumers have questions about infrastructure 3-1 shows the consumer technologies with the fastest growth availability, resale value, and other variables. A further com- rates. As the figure shows, new products can take many years plication can be the innovation ecosystem, which includes to be adopted by a large percentage of the consumers in a all elements of the total customer solution. For PEVs, the market. For example, consider that the microwave—a rela- innovation ecosystem includes not only the vehicle but also the charging stations (whether at home, at work, or in public 1 More information on vehicle technologies, emissions, and fuel spaces) and the necessary permitting and installation, avail- economy trends is available in the EPA Trends Report (EPA 2014). ability of roadside assistance, and other ownership or main-

FIGURE 3-1 Years needed for fastest growing consumer technologies to achieve penetration (0-50 percent or 51-80 percent). SOURCE: Dediu (2012) © Horace Dediu, Asymco.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 54 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 39 tenance concerns. Accordingly, the innovation ecosystem for transportation needs with few, if any, changes in behavior PEVs has its own transition barriers that must be addressed (Consumers Union and the Union of Concerned Scientists for maximum market penetration to occur. Adner (2006) 2013). Of the drivers who could use a PHEV, 60 percent also suggests that wide-scale deployment of new technologies is fit the profile of those who could use a limited-range BEV, a function of three aspects of infrastructure development: (1) such as the Nissan Leaf, without major life changes. product technology—for example, viable, low-cost battery Therefore, market adoption and diffusion of PEVs, which technology; (2) downstream infrastructure—for example, are expensive, infrequently purchased, long-lasting products dealers, repair facilities, emergency roadside services, and with a complicated industry ecosystem, will be a slow process battery recycling options; and (3) complementary infrastruc- that will take decades. That insight is corroborated by the data ture—for example, charging stations (whether residential, presented early in Chapter 1 regarding early-market growth workplace, or public), knowledgeable electricians, and ame- rates and market shares of PEVs. nable zoning and permitting at the municipal level. Adner’s work on innovation ecosystems provides guid- Finding: Market penetration for new technology—particu- ance for how industry stakeholders might make investment larly expensive, infrequently purchased, long-lasting inno- decisions to encourage adoption of new technologies. For ex- vations with a complicated ecosystem—is typically a slow ample, if infrastructure is identified as the critical bottleneck process that takes 10-15 years or more to achieve even nomi- that affects customer adoption and use, industry stakeholders nal penetration. might decide to invest more in infrastructure development than in the product itself. Indeed, Japan has recognized that Finding: Market penetration rates are a function not only of need and has instituted a major initiative to build an extensive the product being purchased but also of the entire industry charging infrastructure to instill range confidence and ensure ecosystem. Hence, product technologies, downstream infra- a safety net for limited-range battery electric vehicle (BEV) structure, and complementary infrastructure all must be at- drivers (METI 2010). Brown et al. (2010) also emphasized tended to simultaneously during the development process. the importance of supporting infrastructure development and Finding: PEVs on the market as of 2014 are not a viable advocated for standardization of codes, training, and other option for all vehicle owners; rather, perhaps only about 40 aspects of infrastructure to facilitate the PEV market. percent of U.S. households exhibit lifestyles amenable to Given the complexity of the innovation ecosystem, main- owning and operating a PEV. stream consumers typically are unwilling to undertake what might be perceived as a risky purchase until all elements of Consumer Diffusion Models and Market Segments the requisite infrastructure are in place (Moore 2014). Indeed, if all aspects of the innovation ecosystem are not ready when Diffusion models categorize consumers (adopters) on the consumers are making purchase decisions, industry adoption basis of their propensity to adopt new technologies and iden- rates can be substantially lower than initial expectations. tify the factors that facilitate adoption and diffusion. Figure Adoption and diffusion models provide insight into what 3-2 illustrates that markets for innovation comprise five dis- might be considered realistic expectations about market pen- tinct categories of adopters; Table 3-1 describes each category etration rates. Given that about 16 million new vehicles are in terms of demographic and psychographic characteristics purchased each year, it would take at least 16 years to con- and buying motivations. Psychographics refer to values and vert the total U.S. fleet of 250 million passenger vehicles and lifestyles of consumers and can be determined empirically light-duty trucks if only PEVs were sold. In addition, not all through market research on their activities, attitudes, inter- households exhibit the demographic and lifestyle traits that ests, and opinions (Kahle and Chiagorous 1997; Wells 2011). make PEVs a viable purchase option. Specifically, when esti- Although demographics can explain who is buying particular mating the total addressable market for PEV sales, one must types of products, psychographics are more likely to explain consider what percentage of the total population would find why customers buy; therefore, psychographics generally are PEVs practical for their travel patterns and needs. A nation- more useful than demographics in understanding customer ally representative telephone survey of adult vehicle owners decisions. Major factors that affect diffusion include com- found that 42 percent of drivers—45 million households— munication (word-of-mouth) between consumers and social meet the basic criteria2 necessary to use a plug-in hybrid networks (Mahajan et al. 1990). electric vehicle (PHEV), such as the Chevrolet Volt, for their DEMOGRAPHICS AND IMPLICATIONS FOR 2 Basic criteria for PHEVs independent of pricing included access ADOPTION AND DIFFUSION OF VEHICLES to parking and an electric outlet at home or work, seating capacity for no more than four occupants, and no hauling or towing capabil- Demographic Traits of Buyers of Plug-in Electric Vehicles ity. A BEV was considered suitable not only when the PHEV criteria were met but also when the maximum weekday driving distance Demographic traits of PEV buyers are compared with was less than 60 miles and other household vehicles were available if weekend driving frequently exceeded the current BEV range those of ICE-vehicle buyers in Table 3-2, which shows that (Consumers Union and the Union of Concerned Scientists 2013). many characteristics of PEV buyers correspond to the traits

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 55 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

40 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Innovators Early Early Late Laggards Adopters Majority Majority

{ { { { { Technology Visionaries Pragmatists Conservatives Skeptics Enthusiasts

TABLE 3-1 Categories and Descriptions of Adopters Category Description Innovators or enthusiasts Are technology enthusiasts or lovers. Are willing to buy early release versions even if product quality or reliability are not yet proven or established. Want to work with developers and infrastructure providers to improve new products, a source of pride in their own techno-intelligence. Are important segments for endorsement about viability of the new innovation category. Are not a large enough market segment to be a long-lived or significant source of revenue. Early Adopters or Visionaries Are less concerned about price and more motivated by psychosocial benefits, such as visibility of their purchase in their peer group. Are more affluent, cosmopolitan, and, typically, younger than other categories. Are willing and motivated to address early market development problems, including service and infrastructure challenges, which when solved, become a source of pride. Are generally considering or comparing purchases not within the product category (for example, with a different vehicle make or model) but with some other major purchase. Early Majority or Pragmatists Are very concerned about value (benefits received relative to price paid). Want to evaluate several different models or options within the product category. Are willing to purchase only when all elements of the requisite infrastructure are in place. Want a hassle-free solution that performs as promised. Are not willing to tolerate anxiety or doubt. Are first sizable segment of the market by volume. Late Majority or Conservatives Tend to buy when there are a plethora of models and choices in the market and when prices have substantially decreased. Laggards or skeptics Would prefer not to buy anything designated as a new technology. Do so only when they can no longer avoid doing so. NOTE: Early and late are relative terms based on the time it takes to adopt.

of early market adopters. PEV buyers had a median income paid cash, and did not seriously consider purchasing another of nearly $128,000 to $148,000 whereas ICE-vehicle buy- vehicle, whereas Nissan Leaf buyers are younger with larger ers had a median income of about $83,000 (Strategic Vi- household sizes. Chevrolet Volt buyers exhibit lower educa- sion 2014). By way of comparison, HEV buyers had a me- tional levels than other PEV buyers. Toyota Plug-in Prius buy- dian household income of $90,204, and the average median ers have a higher percent of female buyers. Finally, PEV buy- U.S. household income was $51,017 (U.S. Census Bureau ers who considered other models of PEVs in their purchase 2013a). Consistent with traits of early adopters, PEV buyers process reported that the vehicle that they most seriously con- were better educated than ICE-vehicle buyers. sidered was the Chevrolet Volt. Table 3-3 lists demographic data for purchasers of a The data presented in Table 3-3 also show that leasing vehicle from each category of PEV as defined in Chapter 2 rates vary by PEV model. The Nissan Leaf and Chevrolet (long-range BEV, limited-range BEV, range-extended PHEV, Volt have higher lease rates than the Toyota Plug-in Prius and minimal PHEV) compared with data for all new-vehicle or Tesla Model S (Strategic Vision 2014). Furthermore, data buyers. The table shows that of the four types of PEVs, Tesla from the California Plug-in Electric Vehicle Survey indicate Model S buyers are primarily men who have higher incomes, that PEVs are leased at a rate of 28.8 percent in California,

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 56 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 41

TABLE 3-2 Comparison of New BEV Buyers, PHEV Buyers, and ICE-Vehicle Buyers Characteristic BEV Buyer PHEV Buyer ICE-Vehicle Buyer Gender 77% male 70% male 60% male Marital status 81% married 78% married 66% married Average age 48 years 52 years 52 years Education 86% college graduate 77% college graduate 59% college graduate Occupation 42% professional 37% professional 25% professional Median household income $148,158 $127,696 $83,166 Number of respondents 3,556 1,000 186,662 NOTE: BEV, battery electric vehicle; ICE, internal-combustion engine; PHEV, plug-in hybrid electric vehicle. SOURCE: Strategic Vision New Vehicle Experience Study of Vehicle Registrants, October 2013-June 2014.

TABLE 3-3 Comparison of All New-Vehicle Buyers to Buyers of Specific Plug-in Electric Vehiclesa Characteristic All New-Vehicle Buyers Tesla Model S Nissan Leaf Chevrolet Volt Toyota Prius Plug-in Gender (M/F) 61/39 82/18 77/23 74/26 66/34 Married or partnered 71 83 87 82 76 Age 50+ 56 68 37 61 39 Household size of 1 or 2 58 56 35 53 46 College grad or more 59 87 86 77 83 Income +$100K 40 88 66 63 62 Caucasian 79 86 70 82 56 Purchased/leased 78/22 95/5 14/86 56/44 68/32 Paid cash 14 36 5 12 2 Received special financial incentives 64 24 76 73 88 Did not seriously consider any other vehicle NA 62 50 42 48 Seriously considered other models NA Chevrolet Chevrolet Toyota Plug-in Chevrolet Volt (1%) Volt (10%) Prius (5%) Volt (8%) Number of respondents 237,235 285 2,257 556 169 a Entries are provided as percent of respondents. SOURCE: Strategic Vision New Vehicle Experience Study of Vehicle Registrants, October 2013-June 2014.

greater than the overall lease rate for light-duty vehicles in the figure represent U.S. data on all vehicles), their involve- the United States (Rai and Nath 2014; Tal et al. 2013). Al- ment in PEV purchases ranges between only 15 and 30 per- though many consumers have never leased a vehicle and cent (the bottom four bars in the figure represent data on Cal- are therefore unfamiliar with the process, leasing a PEV ifornia PEV buyers or lessees only) (Caperello et al. 2014). removes the risk to the consumer that is associated with The authors’ detailed interviews and focus groups find that unknown resale value, battery decay, and rapid technology men treat PEV purchases as “projects”—a classic feature changes. Moreover, leasing agencies are able to incorporate of early market adopters—whereas women in the study ex- the federal tax incentives into a shorter period of time. As a pressed more practical concerns and did not want to experi- result, attractive leasing deals have positively affected PEV ment, a buying trait more typical of mainstream adopters. sales (Loveday 2013a). Whether leases appeal differentially Hence, the gender data also are consistent with differences to early adopters or mainstream customers is unknown. between early adopters and mainstream adopters. To date, male buyers dominate the PEV market. Figure 3-3 shows that although women make between 50 and 60 Finding: PEVs to date have been sold primarily to custom- percent of vehicle purchases generally (the top two bars in ers in the early adopter segment of the marketplace whose

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 57 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

42 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Percent of vehicle transactions attributed to women 0% 20% 40% 60% 80% 100%

New (all)

Used (all)

CA PEVs

Leaf

Volt

Tesla S

FIGURE 3-3 Women’s rate of participation in the markets for all vehicles and for PEVs. The figure shows that women’s participation in vehicle purchases is much lower for PEVs than for vehicles as a whole. Data in blue represent the entire used and new-vehicle market for the entire United States. Data in red reflect California PEV purchasers. NOTE: PEV, plug-in electric vehicle. SOURCE: Image courtesy of Kenneth S. Kurani, University of California, Davis, Institute of Transportation. Data compiled from NBCUniversal, Center for Sus- tainable Energy, California Air Resources Board Clean Vehicle Rebate Project, and EV Consumer Survey Dashboard.

traits and buying motives are different from those of the nance in the initial beachhead can be used more efficiently mainstream market segment. and effectively to drive sales in related, adjacent segments. For example, word-of-mouth communication is easier and Selecting a Beachhead more effective between adjacent market segments (related geographically, by common lifestyles, or by common profes- Diffusion is a social and geographic process; at any sional circles) because people will find communicating with point in time, diffusion in one region of a large country can others who have similar traits more credible and relevant be ahead of diffusion in another, as is illustrated in Figure than with those who have dissimilar traits. Thus, rather than 3-4, which shows the variation in PEV deployment across attempting to succeed in the broad mass market, providers of the United States and provides the projected cumulative PEV new and complex technologies find it advantageous to focus volume in 2014 for the 100 largest urban areas. PEVs tend on a narrow segment of consumers for whom the innovation to be sold in states and municipalities where both the demo- offers a compelling reason to buy. Success in that initial seg- graphic and psychographic profiles of residents are consis- ment then can be leveraged powerfully in adjacent segments. tent with those of the early adopter category; these areas also For the PEV market, a beachhead approach logically tend to have a positive regulatory climate for PEVs. Califor- would focus on key geographic regions or regional corri- nia is one such area and has a long history of strong sales for dors where momentum has already been established; infra- new vehicle technologies. It has the highest proportion of structure is more readily available; word-of-mouth between HEVs in the United States, and the Toyota Prius hybrid was neighbors, friends, and co-workers can occur more readily; the best-selling vehicle in California in 2012 and 2013. where there is greater availability of PEV makes and models; To ensure that new technologies succeed with main- and where gasoline is expensive or electricity is cheap. As stream consumers, Moore (2014) suggests selecting a “beach- one might expect, California is a particularly attractive mar- head,” a narrow market segment of consumers for whom the ket; it accounts for over one-third of annual PEV sales in the new technology offers “a compelling reason to buy.” That United States, and sales of PEVs in California at the close of approach is in contrast to conventional thinking that a broad 2014 comprised 3.2 percent of new light-duty vehicle sales mass market is desirable. The logic behind a beachhead is and 5.2 percent of new passenger vehicles (CNCDA 2015). that, by offering a compelling value proposition specifically It also has a supportive regulatory environment with its zero- targeted to meet the needs of a narrow subset of consumers, emission-vehicle (ZEV) mandate, which has been a prime the technology stands a greater chance of dominance in a key contributor to the availability of PEV models in California. market segment. Then, the momentum gained through domi- States that have agreed to implement the multistate ZEV

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 58 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 43

FIGURE 3-4 Projected 2014 light-duty PEV volume in the 100 largest MSAs. Dot size is proportional to the projected total PEV volume in that MSA in 2014. It can be seen that large numbers of vehicles are located in Los Angeles, the San Francisco Bay Area, New York, San Diego, Seattle, and Atlanta. As noted in the figure, San Jose, San Francisco, Honolulu, Los Angeles, and San Diego have the largest per capita concentrations of PEVs (volume projected for 2014 per 1,000 people based on 2012 census projections). NOTE: MSA, metropolitan statistical areas; PEV, plug-in electric vehicle. SOURCE: Data courtesy of Navigant Research in Shepard and Gartner (2014).

action plan (modeled after California’s ZEV mandate) and ing a limited-range BEV if it also owns a PHEV or an ICE that have greater availability of PEV models are also favor- vehicle that can be used for long-distance trips. Many house- able places for the beachhead approach. They include Con- holds that have multiple vehicles, however, might not be able necticut, Maryland, Massachusetts, New Jersey, New York, to replace all their vehicles with PEVs because all the park- Oregon, and Rhode Island.3 Places where there is clean and ing spots might not have access to charging infrastructure. low-cost hydroelectric power are also favorable locales for the beachhead approach. One such example is Washington Finding: The PEV market is characterized by strong region- state, which has higher PEV per capita sales than California. al patterns that reflect certain key demographics, values, and One final segment that could constitute a favorable PEV lifestyle preferences and have favorable regulatory environ- market is the multiple-vehicle household. Most households ments for PEVs. have more than one vehicle. At the national level, of the roughly 75 million owner-occupied housing units, 3.4 per- Finding: Initial beachheads for PEV deployment are specific cent have no vehicle, 26.7 percent have one vehicle, 43.8 geographic areas, such as California, that have expensive gas- percent have two vehicles, and 26.1 percent have three or oline; key demographics, values, and lifestyles; a regulatory more vehicles (U.S. Census Bureau 2013b). Having multiple environment favorable to PEVs; a variety of PEV makes and vehicles offers the opportunity to choose among vehicles models available; and existing infrastructure or an ability to that have different utilities. For example, a multiple-vehicle readily deploy such infrastructure. household might be able to mitigate the challenges of own- Driving Characteristics and Needs of the Mainstream Consumer 3 Information on model availability by state was provided by representatives of vehicle manufacturers. Sources were Brian Brock- man, Nissan, September 8, 2014; William Chernicoff, Toyota, As discussed, selecting a beachhead plants the seeds August 22, 2014; Kevin Kelly, Joe LaMuraglia, and Shad Blanch, for the diffusion and adoption of PEVs, but PEVs will need GM, August 22, 2014; James Kliesch, Honda, September 2, 2014; to meet more consumer needs to gain greater market share Nancy Homeister, Ford, September 2, 2014; and Dan Irvin, Mit- or become widely adopted. To understand what mainstream subishi, September 10, 2014.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 59 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

44 Overcoming Barriers to Deployment of Plug-in Electric Vehicles consumers might want, it is important to consider how people Finding: Although there is substantial variability in vehicle use vehicles and how those driving habits intersect with the use, average daily travel or other routine use provides a met- four classes of PEVs defined in Chapter 2. As of 2013, there ric that can help evaluate the utility of a PEV. were more than 233 million light-duty vehicles registered in the United States, each traveling on average 11,346 miles per Finding: Aside from average or routine use, many consum- year (FHWA 2015). The Federal Highway Administration ers make a small number of long-distance trips that might provides more detailed information about household trips weigh heavily in their vehicle purchase decision. that might help to determine whether consumers would be interested in purchasing and using PEVs. In the most recent Changing Landscape of Vehicle Ownership and Use: data from 2009, households reported an average of 3.02 trips Implications for Adoption and Diffusion of Vehicles per vehicle per day and 28.97 miles per day per vehicle and an average vehicle trip length of 9.72 miles (FHWA 2011). Social and demographic changes are affecting the amount Changing trends in vehicle ownership and use are discussed that people drive and the demand for new vehicles; these in greater detail in the next section. changes have implications for PEV sales and their use. Vehi- Although averages provide some important information cle miles traveled (VMT) per capita generally increased from about how people use their vehicles, there is substantial vari- 1960 to 2007, outpacing growth in gross domestic product per ability in use among drivers and vehicle type and over time capita. After 2007, VMT peaked, and VMT per capita began (for example, from one day to the next), so that average use to decline as unemployment and gasoline prices rose, result- might not fully capture consumer needs over the life of the ing in fewer commuters, fewer driving vacations, and more at- vehicle. For example, the National Household Travel Survey tention to the cost of fuel (FHWA 2012a; U.S. Census Bureau shows that trips of fewer than 10 miles constituted 71 percent 2013a; Zmud et al. 2014). A 2014 report from the Transporta- of trips and accounted for 25 percent of miles traveled. Com- tion Research Board identified an aging and more ethnically muting is a common routine trip that averages lengths of 6 diverse population, along with changing patterns in the work- miles and represents 27.8 percent of miles. Routine trips are force, urban living, household formation, views on environ- important to consider because they represent an opportunity to mentalism, and use of digital technology, which are affecting electrify miles and maximize the value proposition for PEVs. total and per capita VMT (Zmud et al. 2014). Although VMT Long trips (over 100 miles) represented less than 1 percent of is on the rise again, it is still below 2007 levels (FHWA 2014) trips but 16 percent of miles traveled (FHWA 2011). As noted and is projected to grow at an average annual rate of only 0.9 in Chapter 2, long trips are an issue for BEVs because trips percent between 2012 and 2040 (EIA 2014a). that exceed the all-electric range become inconvenient. The demand for new vehicles also appears to be chang- Mainstream consumers consider what kinds of trips ing. First, the demand for new vehicles has decreased. they need to complete when purchasing (and using) a ve- Americans buy new vehicles every 6-8 years on average, hicle. Those considerations will affect their views on the util- as compared with every 3-4 years before the recession (Le- ity of the vehicle. Many consumers might not find the utility Beau 2012). Related research from J.D. Power (Henry 2012) of a long-range BEV to be substantially limited by trip dis- shows that the average trade-in vehicle at dealerships is tance. Some consumers might find that although a limited- now 6.5 years old, 1 year older than the average in 2007. In range BEV might meet their average travel needs, it does not contrast to almost all products, vehicles have a robust sec- meet their needs to make the occasional long trip. A high fre- ondary (used) market that is larger than the new market; in quency of those “inconvenient days” might greatly dissuade fact, two-thirds of all U.S. vehicle purchases are for used a consumer from purchasing a limited-range BEV. However, vehicles (35.7 million in 2013) (Edmunds 2013). Those data if consumers have multiple options for making longer trips, have implications for vehicle purchases generally and PEV such as public transportation or a second vehicle, they might purchases specifically. Given the length of time between find that a limited-range BEV best meets their routine needs. purchases, product options will have changed substantially, PHEVs can accommodate all possible trip lengths with easy particularly because of model and technology changes, and refueling, but they sacrifice electric miles for gasoline-fu- what the consumer might want or need in a new model might eled miles on longer trips. Average or routine travel needs, have changed substantially. Thus, the consumer likely will such as a commute, might also affect the PHEV range that undertake a lengthy and exhaustive process before purchase a consumer might choose because matching PHEV range to to research new options on the market; that research could average or routine use might improve the consumer value take as long as several weeks or months and involve many proposition. This discussion assumes that consumers under- hours of online research before even visiting dealerships stand their needs and the ability of various types of vehicles (Darvish 2013). The decreased demand for new vehicles and to meet those needs. Later, this chapter discusses misconcep- the lengthy research process will certainly affect the adop- tions and gaps in knowledge about PEVs that lead to con- tion and diffusion rates for PEVs. sumer misperceptions of range and vehicle utility, a barrier to PEV deployment.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 60 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 45 Second, the number of households without a vehicle has The societal changes noted are influencing the growth increased nearly every year since 2005; it was 8.87 percent in of alternative transportation methods, such as on-demand 2005 and 9.22 percent in 2012 (U.S. Census Bureau 2013a; transport services (for example, Uber and Lyft) and car- Sivak 2014). Fewer vehicle-owning households might mean sharing programs (for example, Car2Go and Zipcar). Ac- fewer households in the market for PEVs. The percentage of cording to Susan Shaheen and Adam Cohen (2013), there households without a vehicle also varies widely by geograph- are about 850,000 car-sharing members and 15,000 vehicles ical area. New York City, Washington D.C., Boston, Philadel- in North America (see Figure 3-5). Frost and Sullivan esti- phia, San Francisco, Baltimore, Chicago, and Detroit all have mate in their optimistic scenario that up to 7 million mem- more than 25 percent of households without a vehicle (Sivak bers and 155,000 vehicles could be part of car sharing by 2014). The geographic variation will affect where PEVs sell 2020 (Brook 2014). Given that car sharing and on-demand well. services are growing and tailoring their services to city liv- Another factor that is changing is household formation. ing, urban consumers are becoming reluctant to assume the In 2000, 68.1 percent of households were defined as “family” responsibility and expense of a vehicle. (married couples with children, married couples without chil- Car sharing could be a win or a loss for PEVs, depending dren, single parents with children, or other family). In 2010, on whether the programs use PEVs. If they do, they would that estimate decreased to 66.4 percent because single-person provide ways for drivers outside the new-vehicle market to households increased from 25.8 percent to 26.7 percent (U.S. use PEVs. However, if potential PEV buyers chose to use Census Bureau 2013a). As single-person households and ur- car sharing and car-sharing programs use only ICE vehicles, banization increase, charging vehicles at home could become PEV sales could be hurt and fewer miles electrified. Car- even more complicated as people move into apartments and sharing programs are discussed further later in this chapter. multifamily dwellings and away from single-family homes that have garages or dedicated parking. As urbanization con- Finding: Demographics, values, and lifestyles affect not only tinues to rise, people might use transportation modes other vehicle preferences but also the practicality of a given PEV for than the traditional ICE vehicles. Although urban dwellers a given individual. Different market solutions will be needed tend to log fewer VMTs (a characteristic favorable to PEV for different market categories and segments. ownership), they might face challenges in finding a reliable and regular place to plug in and recharge in a city.

FIGURE 3-5 Worldwide growth of car sharing in terms of vehicles and members. SOURCE: Shaheen and Cohen (2013), Transporta- tion Sustainability Research Center, University of California at Berkeley.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 61 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

46 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

TABLE 3-4 Factors That Affect Adoption and Diffusion of Innovation Factor Description Relative advantage The buyer’s perceived benefits of adoption (such as fuel savings) relative to the price paid (PEVs are expensive relative to ICE vehicles) and the nonmonetary costs (such as concerns about battery life, charging infrastructure, resale value, and vehicle range if a limited-range BEV).

Complexity Difficulty of using the new product. For example, what is involved in charging at home, at work, or at public stations? Are permits required for at-home installation? Is membership needed for a charging network? How much will the electricity to fuel the vehicle cost, and how is that cost calculated?

Compatibility How well does the new technology fit into the buyer’s lifestyle? For example, is the range of a limited-range BEV adequate? Consumer concerns about standards, for example, different plug types and charging networks with different communications protocols and payment methods; mainstream consumers take a wait-and-see attitude to avoid purchasing the wrong product that does not become the dominant design.

Trial-ability How easy is it for a potential customer to try the new technology? A typical test-drive for a PEV can demonstrate its acceleration speed and drivability but does not allow the buyer to experience charging or to resolve other concerns that inhibit purchase of a PEV.

Observability How observable are the benefits of the new purchase to the consumer, such as fuel savings relative to electricity costs; convenience of charging at home and not having to go to a gasoline station; and quiet driving experience. How observable are the new technology and its benefits to other consumers; for example, seeing neighbors or co-workers drive a PEV or seeing PEVs plugged in at a public location hastens diffusion, much like iconic white ear buds and wires were highly visible symbols of Apple products. NOTE: BEV, battery electric vehicle; ICE, internal-combustion engine; PEV, plug-in electric vehicle. SOURCE: Adapted from Mohr et al. (2010).

THE MAINSTREAM CONSUMER AND of Strategic Vision’s May 2013 survey results. Interestingly, POSSIBLE BARRIERS TO THEIR ADOPTION the average gasoline price per gallon was $4.02 at the time OF PLUG-IN ELECTRIC VEHICLES of the survey, and yet consumers still ranked such features as seating comfort above fuel economy as a purchase reason. Insights into strategies to diffuse new vehicle technolo- Just 5 percent of U.S. consumers who purchased a vehicle gies beyond early adopters can be gleaned from industry responded that they were willing to pay more for an environ- studies on what consumers consider when they make a pur- mentally friendly vehicle (Strategic Vision 2013). Additional chase and by examining general factors that affect adoption survey data from “rejecters” (people who considered buying and diffusion of new technologies (Rogers 2003). Five fac- a PEV but chose not to buy one) reveal consumer concerns tors typically affect the rates of adoption and diffusion for about the reliability of the technology and the durability of innovative products; these factors are shown in Table 3-4, the battery (Strategic Vision 2013). Those data suggest that which also provides implications specific to PEV deploy- consumers appear to use traditional criteria (reliability and ment. durability) in their PEV evaluations and that PEVs today As noted earlier, the characteristics and buying motiva- must compare effectively with ICE vehicles on traditional tions differ between categories of consumers. The charac- criteria to be competitive. teristics of PEV owners to date are consistent with those of Egbue and Long (2012) conducted a study to explore the early adopters. Because mainstream adopters (early and concerns about PEVs specifically. Interestingly, in their sam- late majority categories combined) comprise the bulk of the ple of respondents (a population of faculty, staff, and stu- purchases for any new technology (Rogers 2003), under- dents from a technically oriented university), battery range standing their purchase motivations is critically important to was the biggest concern expressed about PEVs, followed by increasing PEV deployment. the cost differential of PEVs compared with ICE vehicles. The top five reasons consumers give for their vehicle Battery range is not a question that is asked in typical ve- purchase choices generally (not specific to PEVs) are reli- hicle industry research studies.4 Corroborating the results of ability, durability, quality of workmanship, value for the 4 money, and manufacturer’s reputation (Strategic Vision Business experts note several caveats in conducting and inter- preting consumer research on new technologies (Leonard-Barton 2013). Although often assumed to be a key influential factor et al. 1995; Rayport and Leonard-Barton 1997; Seybold 2001; in vehicle purchases, fuel economy is a primary consider- McQuarrie 2008). First, consumers necessarily are constrained in ation for 45 percent of consumers (compared with reliabil- their responses by their knowledge of and familiarity with a given ity, a primary consideration for 68 percent of consumers). technology. Although they provide answers to research questions, In fact, fuel economy ranked 11 of 54 reasons on the basis the validity of their responses can be suspect. Moreover, the nature of the research protocols is similarly constrained by the known

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 62 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 47 the Strategic Vision study above, Egbue and Long (2012) Limited Variety and Availability of similarly found that environmental considerations carry less Plug-in Electric Vehicles weight in the purchase decision for PEVs than battery range and cost. Their study further suggests that even with incen- Consumers are accustomed to a dizzying array of ICE tives to subsidize the cost of PEVs, penetration rates are vehicle models and styles available from more than a dozen likely to remain low if consumers have low confidence in manufacturers. They include performance sports cars, mid- the technology. sized passenger cars, sport utility vehicles, crossovers, lux- Despite the fact that 55 percent of people shopping for ury sedans, compact and subcompact economy cars, sporty a vehicle have “favorable” or “very favorable” impressions compacts, pickup trucks, minivans, and full-sized vans. Be- of PEVs (versus 62 percent in 2009) (Pike Research 2012), cause consumers have a wide variety of needs and motiva- the purchase rates are still low.5 Importantly, consumers tions, a wide array of PEV makes and models are needed to make decisions on the basis of their perceptions rather than satisfy them. The rather limited choice of PEVs could slow factual data. Astute marketers realize that consumer percep- market development. tions form the basis of their reality—even if their perceptions Further complicating the rather limited variety of PEVs are factually inaccurate. Although objectively, PEVs might on the market is the fact that not all PEVs are available for exhibit a lower total cost of ownership than ICE vehicles, sale in all states. Two main considerations affect vehicle whether consumers actually compute a total cost of owner- availability. One is the availability of PEVs to the dealers, ship in making vehicle purchase decisions is not apparent.6 which is dictated by the vehicle manufacturers. Given the Ingram (2013) states that 75 percent of people in 21 of the questionable profit margins (Lutz 2012; Voelcker 2013a; largest cities in the United States were unaware of cost sav- Loveday 2013b), some vehicle manufacturers might not be ings and reductions in maintenance costs of PEVs. In fact, motivated to offer PEVs for sale in all 50 states. The other even for high-involvement purchase decisions, in which the consideration is the availability of PEVs to customers— assumption of a “rational consumer” is often made, psycho- specifically, the number of dealers in a given area actually social factors can be more important than rational consider- stocking the vehicle and the number of vehicles on the lot. ations. PEV availability is highly variable by dealer and by location. In addition to the price differential between PEVs and Lack of availability and the limited diversity of PEV options conventional vehicles and the range concerns for limited- are barriers to consumer adoption. range BEVs, the committee identified several additional barriers to PEV purchases—most of which are highly inter- Range of Plug-in Electric Vehicles related—that affect consumer perceptions and their decision process and ultimately (negatively) their purchase decisions. Range anxiety refers to the fear of running out of charge They include the limited variety and availability of PEVs; and being stranded. The driver’s experience of range anxi- misunderstandings concerning range of PEVs; difficulties in ety can be mild or strong and depends on the vehicle range, understanding electricity consumption, calculating fuel costs, charging routines, and driving patterns (Frank et al. 2011). and determining charging infrastructure needs; complexities As discussed in Chapter 2, range limitation should be an is- of installing home charging; difficulties in determining the sue only for limited-range BEVs. Yet, data collected from “greenness” of the vehicle; lack of information on incentives; people who considered a PEV but did not buy one (rejecter and lack of knowledge of unique PEV benefits. Those barriers data) reveal inaccurate perceptions about PEV range. For ex- are discussed briefly in the following sections. ample, some buyers who considered the Chevrolet Volt did not buy it because it “lacked range,” despite the fact that the Volt’s onboard ICE gives the vehicle a range similar to that and familiar. When replicating vehicle surveys to assess PEVs, the surveys do not include questions to assess consumer knowledge of a conventional vehicle. Specifically, after its 38 miles of of and preference for charging infrastructure, range, and other rel- all-electric range are depleted, it offers another 344 miles on evant factors. As a result of those and other limitations, innova- gasoline. Such observations show that a lack of familiarity tion experts recommend alternative methods of market research to with PEVs poses a barrier to vehicle deployment; this nega- complement traditional surveys and focus groups. 3M, Intel, HP, tive effect is corroborated by the modeling work of Lim et al. and other companies known for their culture of innovation rely on (2014), who found that range concerns, as well as concerns a variety of alternative research protocols, many of them more ob- servational in nature, to recognize such limitations. over unknown resale value, inhibit mass adoption of PEVs. 5 As a point of reference, the same survey showed 61 percent of consumers have “a favorable or very favorable” impression of HEVs Understanding Electricity Consumption that have sales of about 3-3.5 percent of new passenger vehicle sales. 6 Despite the lack of information, Eppstein et al. (2011) found in Drivers of ICE vehicles are accustomed to fueling with a simulation model that making available estimates of lifetime fuel gasoline and understand how much range they have left and costs associated with different vehicle types could enhance market where gasoline stations are located relative to that range. penetration substantially. That possibility is supported by marketing in Japan, where at least one PEV manufacturer is actively using PEV drivers, however, face a new experience—fueling with marketing messaging with information on total cost of ownership. electricity—and will need to understand the interaction be-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 63 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

48 Overcoming Barriers to Deployment of Plug-in Electric Vehicles tween several factors, including the storage capacity of the Determining Charging Infrastructure Needs batteries, access to charging infrastructure, and driving behavior. The amount of stored electricity is measured and then The charging infrastructure is a new part of the vehicle communicated through dashboard displays that provide an ecosystem that customers must navigate. Potential PEV pur- estimate of the remaining range of the battery, a measure- chasers need to know what type of charging infrastructure ment that not only is new but that can also be imprecise. they will need, how to get it installed at home, how to find PEV owners will experience consuming the electric energy charging stations when needed, and how to subscribe to or (depleting the battery) quickly or slowly, depending on driv- pay for access to the charging stations. Those issues must be ing speed (fast or slow), conditions (such as ambient air tem- considered by potential PEV customers when they consider perature and steepness of the road grade), and driving style purchasing a PEV. of the driver (light-footed or heavy-footed) (Turrentine et al. Unlike ICE vehicles, for which public fueling sta- 2011). A PEV on a cold day can consume its stored electric tions are the standard, PEVs may be fueled with electricity energy quickly because some portion of that energy goes to at home, at workplaces, or at public charging stations (see heat the vehicle interior; hence, drivers might see the bat- Chapter 5). In fact, early adopters have primarily satisfied tery energy on the dashboard display drop rapidly. For ex- their charging needs at home, and the majority of main- ample, the range of a Nissan Leaf is 84 miles on the EPA test stream PEV adopters are also likely to find home charging cycle, but if the owner drives 90 percent of his or her miles to be most convenient. The paradigm for fueling PEVs at at speeds above 70 mph and lives in a cold climate, the range the owner’s home is a fact not appreciated by many unfamil- could be as low as 50 miles. Thus, to feel comfortable pur- iar with PEVs, including many policy makers and presum- chasing a PEV, consumers generally must understand PEV ably many potential PEV customers who believe that public fuel consumption. charging stations are needed. Although home charging has been the primary method Calculating Fuel Costs of refueling, public charging does have an important role to play. The PEV-driver experience is shaped by the presence Determining electricity costs relative to gasoline costs (availability and visibility) of the charging network in his is yet another factor that affects consumer perceptions and or her region, and a perception of a lack of public charging purchase decisions.7 Box 3-1 shows how electricity cost infrastructure might hinder PEV deployment. For example, could be calculated. The committee was not able to find the United States has over 100,000 gasoline stations com- data on consumer perceptions of electricity costs compared pared with about 8,400 public charging stations (U.S. Cen- with gasoline costs. However, the calculations in Box 3-1 sus Bureau 2012). Japan has recognized the importance of are likely complex enough to be overwhelming for a typical public visibility and access to charging and has instituted a mainstream consumer and highlight the difficulty that con- major initiative to build an extensive charging infrastructure sumers face in computing fuel costs, particularly compared to instill range confidence and ensure a safety net for lim- with those for ICE vehicles. In fact, few consumers are likely ited-range BEV drivers (METI 2010). Drivers of all types to go into this level of detail to understand fuel costs when of PEVs can use their mobile phones or dashboard displays considering a vehicle purchase. The unknown costs repre- to navigate and find fueling stations. Apps for PEV owners sent yet another source of doubt and are therefore another to monitor their state of charge and to find fueling stations barrier. compatible with their vehicles might be particularly impor- Overall, the data indicate that energy costs for PEVs are tant to mitigate consumer concerns about location of fueling likely to be lower, even one-half of gasoline costs. Enrolling stations. in special rate plans, taking advantage of nighttime prices in The extent to which customers understand charging in- some markets, accessing some free electricity at workplaces, frastructure requirements and needs is unknown; however, it and relying on public charging could save PEV drivers even is reasonable to speculate that these considerations are new, more. It is important to note that PEV drivers experience and perhaps surprising, to mainstream consumers. The com- substantial variation and complexity in energy costs across mittee notes that the effect of public-charging availability on regions. Even within a given region, there is much local vari- PEV deployment is not well understood (Lim et al. 2014). ation because of local rates and special PEV rates offered by the thousands of electric companies in the United States, Installing Home Charging differences in prices charged at public charging stations, and in some cases free charging at public and work locations. Depending on regional variations, BEV and PHEV buyers might need to choose, acquire, permit, finance, and install a charger for their primary parking location even before purchasing the vehicle. The decision process will require the 7 Much of existing data about PEV driver behavior with respect to electricity prices are shaped by the high income of the initial buy- buyer to understand the differences in charging technolo- ers who are not as sensitive to gasoline or electricity costs as later gies and possibly to answer the following questions: Do they adopters are likely to be. want or need AC level 1 or level 2 charging? Are upgrades of

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 64 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 49

BOX 3-1 Calculating Electricity or Fuel Costs for Plug-in Electric and Other Vehicles

People shopping for a vehicle face difficulties in calculating fuel costs per mile, especially if they are trying to compare the fuel costs of vehicles operating on different fuels, including BEVs, PHEVs, HEVs, and ICE vehicles. A typical customer’s thought process might proceed as follows:

Possible Cost Calculation for a PEV Take as an example the five-passenger Nissan Leaf, which gets 3 or 4 miles per kilowatt-hour, depending on speed and the heating or cooling needs of the cabin interior. Assume that the average price of residential electricity in the United States is 12.5 cents per kilowatt-hour; this number is based on a range in the United States of 10 to 15 cents per kilowatt-hour (EIA 2014b), and this in turn translates to 3 to 5 cents per mile. On average, therefore, a Leaf owner who charges at home will pay about 4 cents per mile for electricity (these numbers average local taxes on electricity bills).

Possible Cost Calculation for an ICE Vehicle or an HEV Gasoline in the United States in August 2014 cost on average about $3.60 per gallon (regional averages ranged from $3.35 in the Gulf Coast region to $3.91 in the West Coast region) (EIA 2014c). An especially efficient HEV, the Prius, gets about 50 miles per gallon. Average ICE passenger vehicles have a fuel economy of 35 mpg. Thus, the Prius would have cost 7.2 cents per mile, and the average passenger vehicle would have cost about 10 cents per mile in the United States in August 2014.

Therefore, in most places in August 2014, BEVs and PHEVs operating in electric mode on stored electricity from the grid cost less than one-half as much per mile as a comparable-sized gasoline vehicle. Specifically, driving 10,000 miles in a gasoline-fueled compact vehicle would have cost around $1,000 for gasoline in 2014; a comparable-sized BEV would have cost less than $500 at that time.

Additional Considerations • The cost of electricity for a PHEV will vary greatly depending on driving patterns, the charging frequency, and the battery capacity. • Many PEV drivers might charge away from home, where prices vary. Some PEV drivers might be able to maximize their savings by charging for free at work and getting low off-peak or special PEV rates from their utility. • Some places, especially California, have tiered rates to discourage high consumption, or time-of-use rates to shift consumption peaks. Those pricing structures can make electricity rates vary for an individual household by time of day, by total monthly consumption, or by climate zone in which the house is located. • The cost of gasoline can also vary substantially, and that variation complicates the calculation of total fuel costs for PHEV drivers.

household circuits, panels, and even transformers required? formed partnerships to streamline the purchase and installa- How much will the changes cost? What permitting process- tion of home chargers. Three examples of partnerships (list- es, fees, and timing are involved? Will installing a charger ed below) are cited in the Federal Highway Administration require financing (most states require financing of the char- action plan (FHWA 2012b), but one has been discontinued ger to be separate from that of the vehicle)? How much will and another has been reworked. the extra cable for 240 V (level 2 charging) cost?8 Whether the vehicle is leased or purchased might have • Ford and Best Buy. Ford initially partnered with Best an effect on the home-charging decision; people who lease Buy to offer buyers an integrated process for purchas- might be less willing to commit to the expense and effort ing a vehicle and installing a home charger; Best Buy’s of installing home charging. In other cases, installation con- Geek Squad and third-party electrical contractors pro- cerns might be alleviated if PEV owners can use an existing vided installation services. The charging equipment outlet in their garage. The charging concerns for the 46 per- provided by Leviton could be removed, so that own- cent of new PEV buyers who do not have access to home- ers could easily take the charger with them when they charging because they park on the street or live in a multiunit moved. Ford estimated a cost of around $1,500 for the dwelling will be different, but they loom large nonetheless charging equipment and installation services. The pro- (Axsen and Kurani 2009). Barriers to home-based charging gram ended in 2013 when Ford partnered with AeroVi- for that market segment are discussed in Chapter 5. ronment (Motavalli 2013). To help mainstream PEV consumers navigate their • General Motors and SPX. General Motors initially of- home-charging needs, some vehicle manufacturers have fered an AC level 2 home-charging system through a partnership with SPX. The equipment costs were $490; 8 The 240 V cables are different from the 110 V cables that come installation costs varied depending on the existing home with the vehicle and represent an additional customer expense.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 65 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

50 Overcoming Barriers to Deployment of Plug-in Electric Vehicles wiring but were typically about $1,500 (GM Authority Lack of Information on Incentives 2010). General Motors appears to have discontinued its offer for an AC level 2 charging system because the As discussed in Chapters 2 and 7, the prices of PEVs Chevrolet Volt can recharge overnight using an AC level are higher than those of comparable ICE vehicles. However, 1 charger. various financial incentives for consumers can help offset the • Nissan and AeroVironment. For the Leaf, Nissan teamed difference. PEVs can also have nonfinancial incentives, such with AeroVironment to provide home charging; Nissan as access to high-occupancy-vehicle lanes (see Chapter 7 for estimates that a private contractor charges about $2,000 an extensive discussion of incentives). Consumer awareness on average for a typical installation. and perceptions of incentives influence their purchase decisions. In Norway and the Netherlands, for example, PEVs Charging decisions are unique to PEVs and can be over- are particularly popular because people are aware of and whelming. Indeed, until the purchase and use process is sim- want to take advantage of the generous incentives. In the pler—for example, a dealer helps the customer manage the United States, however, a study by Indiana University shows whole process—mainstream consumers simply might revert that 95 percent of the U.S. population in the 21 largest cities to the more familiar purchase of an ICE vehicle that does not is unaware of such incentives (Ingram 2013). A further com- have these added complications (Moore 2014). plication is that federal, state, and municipal incentives are often designed to start and stop at certain times or when cer- Greenness of Plug-in Electric Vehicles tain sales volumes have been achieved. The variability and inconsistency of incentives contribute to customer confusion Perceived favorable environmental impact (the green- in evaluating and purchasing PEVs. ness) of PEVs motivated some early adopters to purchase One study suggests that the effectiveness of PEV incen- PEVs, although environmental impacts appear to be less of a tives could be enhanced through greater consumer awareness motivator for mainstream market consumers given that just 5 (Krause et al. 2013). Dealers could be a source of informa- percent of U.S. vehicle purchasers stated a willingness to pay tion about incentives but are unlikely to have all the nec- more for an environmentally friendly vehicle (Strategic Vision essary information, as discussed below. Moreover, dealers 2013). Others also find that the impact of a green product on might not want to provide information on incentives for fear consumer purchases is usually a third trigger, behind price and of being held accountable if they provide inaccurate infor- quality (Esty and Winston 2009). Still, consumers might want mation (Cahill et al. 2014). Several Internet sources provide to know about the greenness of a PEV—if not for themselves, information on incentives, but the degree to which consum- then when friends, family, and colleagues inquire. ers are aware of and use them is unknown. Consumer might ask the following questions: Does driving a PEV actually benefit the environment? Are greenhouse Lack of Knowledge about the gas emissions and local pollutants decreased if I drive a PEV? Benefits of Plug-in Electric Vehicles Is my electric company a low- or high-carbon emitter? Is my electric company lowering its carbon emissions over time? PEV ownership offers benefits that are familiar to and Similar to computing electricity costs, assessing the greenness valued by their drivers but are probably unfamiliar to main- 9 of a vehicle is complicated; it includes not only the greenness stream consumers. For example, people discover on driving of the electricity supply used to charge the vehicle but also PEVs that they are “peppy” and provide smooth accelera- issues related to how batteries will be disposed of and their tion; moreover, they are quiet (Cahill et al. 2014). In addi- contribution to environmental degradation (see Chapter 4 for tion, PHEVs do not need oil changes as frequently as ICE a discussion of battery recycling). Greenness can be calculated vehicles, and BEVs do not require any oil changes (Voelcker on a well-to-wheels basis, which counts greenhouse gas emis- 2013b, 2014). Furthermore, regenerative braking and energy sions from a vehicle’s tailpipe (tank-to-wheels) and upstream recovery, which is novel to many new PEV drivers, provides emissions from the energy source used to power a vehicle a unique sensation. Whether engineered as part of the tradi- (well-to-tank).10 Although the factual details about the clean- tional braking system (as in the Toyota Prius) or integrated ness of the electric grid (see Chapter 1) might not be widely into the acceleration system (as in the BMW i3 and Tesla known, consumer uncertainty about how green PEVs actually Model S), or both, regenerative braking creates a unique are might cause customers to balk at purchasing one. driving experience. In contrast to systems that capture kinetic energy when the driver begins to brake, regenerative brak- 9 Take, for example, the greenness of the electric grid. Depend- ing integrated into the acceleration system begins to slow the ing on whether the power plants in a given area produce electricity vehicle and capture energy the moment drivers remove their from coal, nuclear, wind, hydropower, or other energy source, the foot from the gasoline pedal. Some drivers perceive the au- greenness can vary greatly. 10 A more complete analysis of vehicle greenness is a life-cycle tomatic braking as an advantage, especially in heavy traffic. assessment that, in addition to the well-to-wheels assessment, takes Thus, PEVs provide a driving experience that is differ- into account environmental impacts of vehicle production, vehicle ent from that of a traditional ICE vehicle. Such differences use, and disposal of the vehicle at the end of its life.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 66 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 51

TABLE 3-5 Consumer Questions Related to Plug-in Electric Vehicle (PEV) Ownership Vehicle Technology Vehicle Charging Other Concerns Questions PEV Customers Might Have Why are PEVs more expensive than Where can I charge the vehicle? What happens if I become stranded because I conventional vehicles? Can I charge it at home? run out of charge? What is the battery range? Do I need a special plug? How green is a particular PEV? How many years will the battery last? How much will the electricity cost? Do the batteries end up in landfills? Do I need to replace it? How long does it take to charge? Is there a risk of fire? Does the vehicle have sufficient power to drive on the highway? What will be the resale value? Questions Customers Might Not Know to Ask How much of my battery range will be sacrificed Where do I charge if I do not have a garage? How do I file for my tax credit? to interior heating and cooling in cold or hot Are there permitting fees to get a dedicated Does my state offer a tax credit? temperatures? charger installed in my home? Do I get free parking? Can my regular repair shop perform maintenance Does my state offer a rebate or incentive to Do I get access to a high-occupancy-vehicle and repair work on the vehicle? install charging equipment in my home? lane with a PEV? How does regenerative braking work? If I want AC level 2 charging, do I need additional Does my employer offer charging at work? Will my battery degrade over time if I use DC equipment, and how much will it cost? Because the car is so quiet, how do I know if fast charging? Do I need to inform my utility if I purchase a PEV? it is running? What are the savings in maintenance and fuel Do my rates for charging differ depending on relative to the purchase price of the vehicle? the time of day? If I belong to a charging network, can I use chargers from other networks? How do I find a charging location? Can I reserve a charging location? What if the charger I need to use is being blocked by another vehicle? Questions Friends, Neighbors, and Even Strangers Might Ask Did my tax dollars subsidize the purchase Do these vehicles put excessive demands on Why do people with PEVs get the most of your PEV? the electric grid? convenient parking spots? Do you pay any fuel fees for highway funds or road taxes?

might appeal more to early adopters than to mainstream con- the known and familiar (Mohr et al. 2006). Until they are suf- sumers, but mainstream consumers will never know whether ficiently informed and educated, they will likely continue to PEVs meet or even exceed their expectations unless they can prefer the relative safety, security, and familiarity of an ICE drive one, making the test-drive a critically important experi- vehicle. Therefore, mainstream adopters require additional ence, as discussed later in this chapter. encouragement, information, and incentives to overcome the barriers identified. Summary of Major Perceptual Barriers Finding: Lack of consumer awareness and knowledge about From the committee’s perspective, the factors discussed PEV offerings, incentives, and features is a barrier to the above pose major barriers to consumer adoption of PEVs. mainstream adoption of PEVs. Confusion continues to loom large in the consumer purchase decision for PEVs. Table 3-5 identifies questions that custom- Finding: The many perceptual factors that contribute to ers might have when contemplating a PEV purchase. Some consumer uncertainty and doubt about the wisdom of a PEV questions—Will my battery catch fire? How do I change my purchase combine with price and range concerns to nega- battery?—might seem nonsensical to a current PEV owner, tively affect PEV purchases. but they are questions that consumers have asked and demonstrate the extent of misinformation and the nature of the per- VEHICLE DEALERSHIPS: ceptual barriers that must be overcome before PEV deploy- A POTENTIAL SOURCE OF INFORMATION? ment becomes widespread. Uncertainty and perceived risk plague consumer willing- A well-known aspect of the vehicle-purchase process en- ness to purchase innovative products, particularly expensive, tails visits to various dealerships for test-drives and purchase long-lived ones, such as vehicles; consumers instead revert to negotiations. Vehicle dealerships traditionally have offered

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 67 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

52 Overcoming Barriers to Deployment of Plug-in Electric Vehicles many services to help with the complexity of the purchase In addition to the paucity of knowledgeable salespeople, process. In fact, in the early years of the automobile market, the Consumer Reports study (Evarts 2014) also found that dealers supported Americans with their vehicle purchases by dealers simply did not have PEVs in stock. Only 15 of 85 teaching them to drive and by providing financing, mainte- dealers in four states (California, New York, Maryland, and nance advice, and service to keep vehicles running. Accord- Oregon)12 had more than 10 PEVs on their lots; indeed, most ingly, dealers have always played a critical role in the decision- dealers had only 1 or 2 PEVs on their lots. That finding is sup- making process of people purchasing a vehicle (Ingram 2014). ported by a UC Davis study, which found that 65 percent of Given that 56 percent of PEV buyers make over three visits California dealerships had no PEVs for sale (UC Davis 2014). to dealerships, which is twice the number made by non-PEV Explanations for the lack of inventory on dealer lots buyers (Cahill et al. 2014), vehicle dealerships could serve as vary. Consumer Reports stated that of those with limited or an important source of information for potential PEV buyers. no stock, most (21) dealers attributed limited stock to “high Despite the importance of the consumer experience at demand;” the next most common explanation, however, the dealership, research on dealers and PEVs reveals sys- was a “lack of consumer interest” in PEVs, also expressed temic problems. A Center for Sustainable Energy survey as “nobody buys them.” Another possible reason for lack of of over 2,000 PEV buyers in California in December 2013 inventory on dealer lots is based on the financial returns from showed that 45 percent of those buyers were “very dissatis- selling PEVs. If vehicle manufacturers are losing money on fied” and another 38 percent were “dissatisfied” with their PEVs, they could limit availability deliberately (see, for ex- purchase experience (Cahill et al. 2014). In the same survey, ample, Beech 2014; Lutz 2012; Voelcker 2013a; Loveday PEV buyers were asked “how valuable was it to have deal- 2013b). The strategy of limiting inventory further hurts sales ers knowledgeable about various topics.” The responses in and makes it harder to generate economies of scale to drive Table 3-6 show that PEV buyers expect dealer salespeople down manufacturing costs. If production costs are not re- to be informed about much more than just the vehicle char- duced, prices will remain high. acteristics. However, a Consumer Reports mystery shopper Additional pressure on dealerships comes from Tesla’s recently went to 85 dealerships in four states and found that challenges to the vehicle dealership franchise laws. In the salespeople were not very knowledgeable (Evarts 2014). United States, direct manufacturer-to-consumer vehicle sales The dealer and salesperson motivation to sell PEVs var- are prohibited by franchise laws that require new vehicles to ies. As noted in the committee’s interim report (NRC 2013b), be sold only by licensed, independently owned dealers (Quin- salespeople take three times longer to close a PEV sale than land 2013). Such licensed dealers sell new and used cars, in- an ICE vehicle sale—time for which they are not differentially cluding certified preowned vehicles, employ trained vehicle compensated. Furthermore, because dealership revenues in- technicians, and offer financing. The National Automobile clude charges for after-sales service and support and because Dealers Association has challenged Tesla’s showroom model PEV maintenance requirements are lower than those for ICE in a handful of states (Colorado, Massachusetts, Illinois, New vehicles, that service revenue is missed. Moreover, sales staff York, Oregon, Texas, and North Carolina).13 So far, because at dealerships often turnover rapidly; thus, technically savvy Tesla does not actually sell vehicles through its showrooms sales staff who are knowledgeable about PEVs are not always (rather, orders are placed online to the factory), courts have available at a given dealer on a given day. Given such turn- generally upheld its model.14 Distribution issues are an on- over, sales training on new products is not always a good in- going area of dispute and although Tesla has advocated for a vestment for the dealership (Darvish 2013). federal law to overturn the state franchise laws, to date it has To address those issues, some dealers in California have been unsuccessful. hired PEV advocates to sell PEVs specifically. Rather than train the entire sales force, high-volume PEV dealerships Finding: Knowledge of PEVs at dealerships is (at best) un- have one or two PEV gurus. Moreover, some dealerships even and (at worst) insufficient to address consumer ques- now separate floor (personal) sales from Internet sales, and tions and concerns. in some situations, 100 percent of PEV sales come from In- ternet inquiries (UC Davis 2014).11 The PEV gurus usually Finding: Dealers are generally less motivated to sell PEVs are part of the Internet sales team for the dealer; social me- than to sell ICE vehicles, and a further complication is that dia are used to steer buyers to those individuals. Partially the inventory of PEVs on dealer lots is limited. because dealership salespeople might lack the ability, time, or incentives to educate customers adequately about PEVs, Tesla decided to operate its own dedicated showrooms in 12 Oregon was not named in the article, but personal communica- which specially trained employees focus exclusively on tion on October 17, 2014, with the author provided the name of the fourth state. educating customers about Tesla vehicle ownership. Tesla 13 Interestingly, NADA is also starting a communications cam- showrooms are typically styled like boutiques in high-traffic paign to counter criticisms that the franchise laws are outdated and locations, such as a mall, much like Apple stores. that dealers are not willing to change (Nelson 2014). 14 See, for example, the situation in New Jersey (Friedman 2014; 11 Nissan initially sold Leafs only on the Internet. Gilbert 2014).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 68 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 53

TABLE 3-6 Ratings of Dealer Knowledge about Various Topics Topic Percentage of Respondents Who Assigned the Dealer a Rating of “Very Valuable” (Highest Ranking) Financial incentives 62

Vehicle performance 62

Nonfinancial incentives 48

Cost of ownership 46

Home charging 42 PEV Smartphone applications 40

Away from home charging 33

Electricity rates for PEVs 32 NOTE: PEV, plug-in electric vehicle. SOURCE: Based on data from Cahill et al. (2014) and CSE (2014).

STRATEGIES TO OVERCOME BARRIERS TO ciently responsive to warrant larger advertising expenditures DEPLOYMENT OF PLUG-IN ELECTRIC VEHICLES is questionable. Regardless of the underlying reasons for what would appear to be a limited effort by any one com- This chapter began by describing diffusion models for pany to advertise its PEVs, the lack of promotion creates a new technologies, including PEVs, and then discussed the self-fulfilling prophecy. demographics and behavior of early adopter and mainstream One strategy that might be used to overcome barriers market segments and their implications for adoption and dif- to vehicle manufacturers advertising their PEVs is coopera- fusion. A number of barriers consumers face in the PEV pur- tive advertising or joint promotion efforts to communicate chase process were discussed next. The sections that follow the advantages of PEVs. Cooperative advertising is a shared consider ways to address consumer barriers to PEV deploy- campaign whereby companies work together to achieve an ment, including advertising strategies to educate consumers, important goal. For example, trade associations for the fish- greater use of Internet resources to disseminate information, ing industry15 and the recreational-vehicle industry16 run and more opportunities for test-drives. campaigns to stimulate demand for the product category as a whole. Rather than any one company having to incur Advertising of Plug-in Electric Vehicles the expense of stimulating consumer demand, an industry trade association or other third party undertakes such efforts Given the complexity of the issues consumers face in on behalf of its members. Other examples here include the the PEV purchase decision, the committee examined vari- “Beef: It’s What’s for Dinner” and “Got Milk” campaigns. ous information sources that consumers commonly rely on Cooperative advertising campaigns typically exist under for decision making. One such source is the advertising and at least one of the following conditions: marketing of the vehicle manufacturer. Advertising historically has been a way to stimulate interest in new products • The industry as a whole is facing a decline in demand and to steer customers to dealers. Because advertising and because of competitive threats (when, for example, marketing plans are critical aspects of a vehicle manufac- consumers spend more time with technology gadgets turer’s strategy, they are proprietary, and the committee did than in going outdoors). not receive information on individual company efforts, such • A new technology is attempting to overcome an incum- as how much they spend on advertising to promote PEVs. bent technology, and the combined efforts of the new Although PEVs are advertised in traditional media, casual technology providers might be able to educate con- observation suggests that company efforts to promote PEVs sumers in a synergistic fashion. are not nearly as aggressive as their efforts for traditional • The products are commodities; thus, the goods of any ICE vehicle makes and models. one provider are indistinguishable from those of the Reasons for limited PEV advertising could include the others (for example, milk). fact that the market is small. A lack of profitability also could be a reason companies do not want to advertise (see, for example, Beech 2014; Lutz 2012; Voelcker 2013a; Loveday 15 See, for example, the Recreational Boating and Fishing Founda- 2013b). Companies want to maximize the return from their tion (http://takemefishing.org/). advertising budgets, and whether the PEV market is suffi- 16 See, for example, Go RVing, Inc. (http://gorving.com/).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 69 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

54 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Although the third situation does not apply in the automotive credits and other incentives, the value proposition for PEV industry, the second one is critical. When the motives of an ownership, and who could usefully own a PEV. individual stakeholder are insufficient, a collective approach to stimulating awareness and action can be effective. Thus, Internet Resources for Information vehicle manufacturers, suppliers of vehicle components, and on Plug-in Electric Vehicles charging providers, if united by a strong enough common interest and a capable third-party organization to manage the For the motivated and savvy consumer, a plethora of campaign, could find value in banding together in a manner online resources are available to research PEVs (see Table similar to other industries. 3-7) and the other components of the purchase decision. Another partnership that might prove advantageous is Online research can provide make and model availability, between vehicle dealerships and electric utilities. Together prices, technical specifications and reviews; describe the they could work to promote the vehicles by emphasizing the charging infrastructure, including locations of public charg- convenience and affordability of electric fuel. For example, ing stations; list incentives by state or zip code; and even Austin Energy works with local vehicle dealerships to provide give estimates of total cost of ownership. Traditional car- prepaid unlimited public fueling cards for $50 per year (K. buying websites, such as Kelley Blue Book and Edmunds, Popham, Austin Energy, personal communication, December have areas dedicated to PEVs. Manufacturers of PEVs have 18, 2014). The program allows salespeople to offer “free un- information on their websites. Many automotive enthusiasts limited public charging” on every new vehicle for a year. also provide information on various other websites. Because In addition to the marketing activities of industry stake- of the importance of electricity and charging to their busi- holders, the federal government has played an important role ness models, most utilities have a section of their websites in sponsoring advertising campaigns to support socially ben- dedicated to PEVs. Many nonprofit environmental organiza- eficial behaviors. The Ad Council (2014), founded in 1941, is tions have sections for PEVs. Consumers looking for infor- a federally subsidized advertising program that partners with mation on charging infrastructure specifically can find many national nonprofits or federal agencies on multimedia mar- resources through the various private companies offering keting campaigns. It selects important public issues (such as public charging. Finally, federal and state websites offer use- the Smokey the Bear “Only You Can Prevent Forest Fires,” ful resources, including calculators for electricity costs. the United Negro College Fund “A Mind Is a Terrible Thing The plethora of online information provides an op- to Waste,” and “Friends Don’t Let Friends Drive Drunk”), portunity to overcome the lack of consumer awareness and partners with a sponsoring organization, and then stimulates knowledge about PEVs, but two potential problems arise. action on those issues through advertising programs. Partner First, the sheer number of Web resources might cause con- organizations (such as the Department of Energy [DOE] in sumers to become overwhelmed and confused. Studies of the case of PEVs) are considered the issue experts and, as consumer decision making show that information overload such, sponsor the campaign and are responsible for produc- is negatively associated with purchase (Herbig and Kramer tion and distribution costs (research, multimedia production, 1994); too many options create confusion (Schwartz 2005; multimedia distribution, social media, and public relations); Scheibehenne et al. 2010). Despite the wealth of information the media space or air time is donated. The Ad Council asks or perhaps because of it, consumer knowledge about PEVs is for a campaign commitment of at least 3 years, which is con- not as great or as sophisticated as it could be, and mispercep- sistent with models of consumer learning and engagement tions certainly continue to exist. for risky, durable purchases. For an Ad Council campaign, Second, finding an easy-to-use source of credible, repu- DOE could work with marketing experts to craft appropriate table information can be difficult. For example, an online messaging, including accurate information about federal tax search to find information related to purchasing a PEV yields credits and other incentives; the value proposition for PEVs a wide array of links, such as sponsored advertisements for generally, including lower operating costs; and the identifi- PEVs, vehicle-manufacturer websites, news articles about cation of people who could usefully own PEVs. More broad- PEVs, blog posts from PEV enthusiasts, buyer guides, infor- ly, the government’s objectives (energy security and clean mation from nonprofits encouraging PEVs, information on transportation) could also be part of the message. tax credits, and even paid Google AdWords campaigns for fuel-efficient ICE vehicles and technologies.17 The confus- Finding: The federal government has a mechanism to coming array of results—including misinformation on PEVs— municate messages to the general public for issues deemed emphasizes the need for a central, credible (unbiased), easy- to be in the public interest. to-use resource to simplify consumer information needs. If consumers are lucky, they will find the useful fed- Recommendation: To provide accurate consumer informa- eral government websites for PEVs. The Alternative Fuels tion and awareness, the federal government should make use of its Ad Council program, particularly in key geographic 17 Search results on any given day and computer are conditioned markets, to provide accurate information about federal tax by the cookies on an individual user’s computer, search engine marketing at the time, and other factors.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 70 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 55

TABLE 3-7 Websites with Information on Plug-in Electric Vehicles Category URL Type of Information Available Vehicle reviews http://www.edmunds.com/hybrid/ Reviews, technical specifications, http://www.kbb.com/electric-car/?vehicleclass=newcar&intent=buy-new&filter=hasincentives make and model availability http://www.consumerreports.org/cro/cars/hybrids-evs.htm http://www.cars.com/guides/all/all/?prop63=Electric%20Powered&highMpgId=1836&sf1Dir=ASC

Vehicle industry blogs http://www.greencarreports.com/ Market trends, including sales and websites http://www.epri.com/Our-Work/Pages/Electric-Transportation.aspx volumes, PEV news, reviews http://www.electrificationcoalition.org/ http://www.plugincars.com/ http://www.howtoelectriccar.com/is-an-electric-car-right-for-me/ https://www.aepohio.com/save/ElectricVehicles/EVRight.aspx http://www.electricdrive.org/ http://www.electriccarbuyer.com/guide/ http://insideevs.com/ http://www.pluginamerica.org/ http://driveelectricweek.org/ http://green.autoblog.com/ http://evsolutions.avinc.com/electric_vehicles/ http://cleantechnica.com/category/clean-transport-2/electric-vehicles/ http://chargedevs.com/ http://www.thecarconnection.com/category/new,electric-car http://www.huffingtonpost.com/news/electric-cars/ http://www.tva.com/environment/technology/electric_transportation.htm https://www.alamedamp.com/types-of-electric-vehicles http://transportevolved.com/

Nonprofit organizations http://www.nrdc.org/energy/vehicles/green-car-tech.asp Environmental impacts of PEVs, http://www.edf.org/transportation/fuel-economy-standards incentives, policy, dispelling myths http://content.sierraclub.org/evguide/

Charging-infrastructure http://www.plugshare.com/ Maps and search tools to find locators http://www.afdc.energy.gov/fuels/electricity_infrastructure.html charging infrastructure, availability of chargers, subscription plans http://www.nrgevgo.com/ http://www.chargepoint.com/ www.juicebarev.com

Cost of ownership http://www.afdc.energy.gov/calc/ Calculators for cost of ownership calculators http://energy.gov/maps/egallon of PEVs based on local and individual variables http://www.electrificationcoalition.org/sites/default/files/EC_State_of_PEV_Market_Final_1.pdf

Federal government http://avt.inel.gov/ Incentive information, regulation resources http://avt.inel.gov/hev.shtml information, data on PEVs, government research, and www.fueleconomy.gov deployment initiatives http://energy.gov/maps/egallon http://www.evroadmap.us/ http://www.afdc.energy.gov/vehicles/electric.html http://www1.eere.energy.gov/cleancities/ http://energy.gov/eere/vehicles/vehicle-technologies-office-hybrid-and-vehicle-systems http://energy.gov/eere/vehicles/vehicle-technologies-office-information-resources http://energy.gov/eere/vehicles/vehicle-technologies-office-ev-everywhere-grand-challenge http://www.epa.gov/greenvehicles

State government https://energycenter.org/ State-specific incentives and resources http://www.westcoastgreenhighway.com/electrichighway.htm policies, consumer guides, resources for advocates, state, http://www.in.gov/oed/2675.htm local and regional charger maps http://www.plugandgonow.com/

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 71 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

56 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Data Center (DOE 2013a) provides comprehensive infor- a critical aspect of the purchase decision, vehicle manufactur- mation on vehicle and fuel characteristics and infrastructure ers, vehicle dealers, nonprofit organizations, and various DOE and useful fuel cost calculators (DOE 2013b, 2014a). The initiatives have experimented with a variety of events to draw DOE Energy Efficiency and Renewable Energy Clean Cit- customers to experience PEVs. For the EV1 launch, GM took ies website (DOE 2014b) also provides valuable information it to several U.S. cities for month-long tests. In 2012, it offered specific to localities and regions where a Clean City Coali- 3-day test-drives of its Chevrolet Volt in major cities. More tion operates. Unfortunately, those websites do not appear recently, Fiat took its 500e BEV to 30 corporate campuses consistently high in search results for PEV information. in California and offered lunch and test-drives to employees Moreover, compared with other shorter and catchier Web ad- (Anders 2012). Plug-in America has been organizing National dresses—such as greencarreports.com, plugincars.com, and Drive Electric Week (formerly National Plug-in Day) since plugandgonow.com—consumers might find it difficult to re- 2011. In 2013, 80 events sponsored by corporations, nonprof- member the URLs for the government sites (see Table 3-7). its, and PEV enthusiasts across the country hosted over 33,000 The committee did not have access to data on the extent to participants and gave over 2,700 test-drives (Plug-in America which car shoppers relied on government website resources. 2013). Furthermore, given the lack of evidence on how consum- DOE recognized the importance of consumer demon- ers use objective information (versus their perceptions) in strations in its July 2014 call for proposals through the Clean purchase decisions, the potential effect of calculators for fuel Cities program (DOE 2014c). It is offering funding for 7 to cost and total cost of ownership on a customer’s evaluation 15 deployment projects in three areas: on-the-road demon- of PEVs is unknown. One common strategy used to evaluate strations, safety-related training, and emergency prepared- the responsiveness of website visitors to various types of and ness. On-road demonstrations will allow people to have first- formats for online information is called A-B testing. A-B test- hand PEV experience for extended periods of time. Whether ing presents version A of the information for a period of time the experience is through car sharing, rental car, or com- and tracks visitor activity and then presents version B for a mercial fleet leasing programs, more drivers will understand similar period of time. The two information strategies are then the benefits of PEVs and be more prepared to evaluate them compared to reveal the differential impact of the information knowledgeably and perhaps more likely to purchase them. presentation. Vehicle manufacturers—including Cadillac, BMW, and Porsche—also are developing regional experience centers Finding: Government websites provide useful information (Colias 2014). To adapt to shifting shopping habits, vehicle for motivated PEV shoppers; however, the degree to which manufacturers are offering customers an opportunity to look they are easy to find, remember, and share is unknown, as is at vehicles in a less sales-oriented environment. For ex- their actual impact on consumer perceptions and behavior. ample, “BMW’s new retail sales model includes plans for regional pools of test cars with a wider range of models, Recommendation: The federal government should engage giving dealers access to more demo models than any store a knowledgeable, customer-oriented digital marketing con- could stock” (Colias 2014). Because the regional facilities sultant to market its online resources and then evaluate their will supplement, not supplant, the existing dealer networks impact. Marketing activities could include purchasing a us- and because they address a different point in the consumer er-friendly, memorable domain name, running various A-B decision-making journey before the actual purchase deci- tests, optimizing search engine marketing to allow shoppers sion, the regional centers are (so far, at least) not in conflict to find useful resources more easily, using sharing tools to with dealer franchise laws. facilitate dissemination among online networks, and iden- The committee finds that such regional centers could be tifying key partners to use application protocol interfaces to a useful strategy to help mainstream customers gain more promote greater consistency of information. hands-on experience with PEVs. For customers who want to compare and contrast different types of PEVs, doing so Test-Drive Events and Regional Experience Centers at a central location would be much easier than having to visit three or four dealerships, especially given that dealer Test-drives are critically important for potential PEV buy- salespeople might not be as knowledgeable as desired and ers because they allow customers to assess the driving char- given the dearth of PEV models available on the lots. There 18 acteristics of PEVs. Because driver experience with PEVs is might also be a business model whereby vehicle manufactur- 18 The committee notes that a test-drive will not allow the driver ers hire a third party to provide ride-and-drive opportunities to experience the effect of driving factors on range and most likely at workplaces and community events. will not provide an opportunity to recharge the vehicle. One could argue that drivers also do not experience the true range or refueling Finding: The test-drive experience, including an opportu- of an ICE vehicle during a test drive. However, ICE-vehicle buy- nity to become familiar with vehicle range and charging, is a ers have enough experience to make an informed decision about those topics to alleviate concern. Potential PEV buyers, on the other critical aspect of the consumer decision-making process for hand, will likely lack information on those topics and will have to PEVs. Thus, more initiatives that offer “ride and drives” for trust the information provided by the dealer. a range of PEVs at a single location would be helpful.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 72 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 57 Recommendation: The federal government should explore leave the car at the destination where they want to go. Other opportunities for a vehicle-industry effort to provide a re- members find the car through radio-frequency identification gional PEV experience center to provide important test-drive technology and use it if the location fits their needs. opportunities. To the extent that car-sharing fleets use PEVs, they also represent a way for the public to experience PEVs and might Other Opportunities to Experience represent an important means for introducing PEVs to mar- Plug-in Electric Vehicles ket segments that might not be traditional new-vehicle buyers. For example, Car2Go indicated that its customers tend to Another opportunity for people to experience PEVs can be young, low-to-moderate income earners, often students, come from fleets. For example, some municipalities in Ja- and urban dwellers who live in areas that have high conges- pan allowed citizens to reserve a city-owned PEV to drive tion and limited parking (Cully 2014). on weekends (H. Matsuura, Kanagawa Prefectural Govern- ment, personal communication, December 11, 2013). Em- FEDERAL GOVERNMENT EFFORTS ployers that have PEVs in their corporate fleets could con- TO FAMILIARIZE CONSUMERS WITH sider a similar idea, either as a perk for employees or as a PLUG-IN ELECTRIC VEHICLES: way to promote an environmental image. Rental car fleets CLEAN CITIES COALITION could also provide an opportunity for customers to experience PEVs. Hertz has its “green collection,” which allows DOE’s major effort in PEV deployment is the Clean Cit- drivers to experience PHEVs (not BEVs). ies program, which is managed by the the department’s Ve- As noted previously, car sharing is a growing trend in hicle Technology Office and stems from the Energy Policy the vehicle market, particularly in large cities, where person- Act of 1992. The program goal is to support local actions for al vehicle ownership is less necessary and less convenient. reducing petroleum use in transportation by promoting alter- PEVs seem like a good choice of vehicle for many car-shar- native-fuel vehicles. Over 80 local Clean Cities coalitions rep- ing enterprises given the often short distances traveled per resent about 80 percent of the U.S. population (Frades 2014). rental and the environmental values that motivate some car In 2011, an additional funding stream of $8.5 million was al- sharers. However, companies that want to use PEVs in car- located for 16 PEV community-readiness projects to support sharing fleets face barriers, such as vehicles that might be public-private partnerships in deployment of PEVs and their more expensive and have a limited range, which might make associated infrastructure. Figure 3-6 shows the Clean Cities them inconvenient for customers. The companies will also coalitions that received funding for PEV readiness. need charging stations and creative strategies for managing Although the coalitions act locally, one of the most use- the operation of the fleets.19 ful and comprehensive resources for PEV owners and policy A successful car-sharing program that uses PEVs has makers from the Clean Cities program is the DOE Vehicle been implemented in Madrid, where Respiro Car Sharing, Technology Office Alternative Fuels Data Center website Nissan Leaf, and NH Hotels have collaborated to develop in Table 3-7. Particularly useful is the station locator (DOE a sustainable mobility plan for the city, where many vehi- 2014a), which allows searching for PEV charging by loca- cles travel fewer than 50 km per day (EnergyNews 2013). tion, by charging technology, by station type, and by pay- In Paris, AutoLib was introduced in 2012 with 250 vehicles ment method accepted. and 250 stations. Eventually, the company plans to grow to Although PEVs are only a small part of the Clean Cit- 3,000 vehicles and 1,000 stations at an investment cost of ies initiative, the 2011 Clean Cities strategic plan describes €235 million. PEV car sharing has also been successful in the key areas for PEV deployment efforts, including strong urban centers in places like in the Netherlands, where there is coalitions and partnerships, infrastructure deployment in a scarcity of parking spaces and having a reserved PEV park- vehicle-manufacturer target markets, information provision ing location is valuable. Car2Go has three all-BEV fleets and data collection, and training (DOE 2011). In their readi- worldwide (San Diego, Amsterdam, and Stuttgart); its 27 ness plans, the 16 Clean Cities coalitions identified barriers other locations offer PEVs and gasoline vehicles. And BMW to PEV adoption and infrastructure deployment and imple- has its i-Drive initiative in the San Francisco Bay Area, where mented plans to overcome the barriers. One of the primary it has a point-to-point service similar to Car2Go. Members barriers identified was lack of awareness of and information of the service can access a car from one of the multiple loca- about PEVs on the part of many stakeholders, including ve- tions in the cities where the service is available and can then hicle purchasers and the government. To overcome the barriers, the coalitions supported outreach, education, training, 19 As noted, car-sharing programs are allowing point-to-point and marketing efforts, including hosting events for people to rides, but that freedom also makes managing an all-BEV fleet dif- experience PEVs. They also produced templates, guides, and ficult because members could drain the battery and leave a vehicle tools for outreach to the public, businesses, and local gov- stranded. Given the few public charging stations in San Diego, Car2Go has managed an all-BEV fleet by building and operating ernment officials (Frades 2014). The committee found that a charging barn where it charges its entire fleet on average every Clean Cities coalitions were vital to increased deployment 2 days.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 73 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

58 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 3-6 Clean Cities coalitions funded for community-readiness and planning for PEVs and PEV charging infrastructure. The grants are to Clean Cities coalitions with a focus on PEV deployment. SOURCE: Frades (2014).

in some places, such as Atlanta. Aside from DOE’s research cent. Fleet managers are looking to alternative-fuel vehicles, funding, the Clean Cities coalitions that are working on PEV including PEVs, to meet societal responsibilities to lower deployment represent DOE’s most prominent efforts in PEV greenhouse gas emissions, to lower fuel and operating costs, deployment. and to maintain an environmentally friendly image. As Table 3-8 shows, information resources for fleet man- Finding: Clean Cities coalitions have been vital to increased agers who are tasked with greening their fleets are plentiful. deployment in some localities and represent DOE’s most However, PEV deployment in fleets has not been strong. Ac- prominent efforts in PEV deployment aside from its research cordingly, this final section provides a brief overview of the funding. three main classes of fleet buyers and an assessment of the barriers to and opportunities for facilitating PEV deployment FLEET PURCHASES in this segment.

One method to increase PEV deployment is purchasing Rental Fleets them for fleets, particularly fleets where vehicles leave and return to the same base and have similar daily routes. Vehicle Although rental companies comprise the largest fleet fleet sales make up 20-22 percent of the U.S. market (Auto- buyers, the viability of PEVs in their fleets is constrained by motive Fleet 2013). The exact size of the fleet market is hard not knowing a typical customer’s driving range and the need to measure because not all purchasers identify themselves as for charging and the difficulty of gauging the resale value of a a business purchaser, and some fleet vehicles are driven for PEV (El-Moursi 2013). Rates for renting PEVs are generally private use. Figure 3-7 shows that the fleet category includes higher than for conventional vehicles, and their availability is an array of buyers, including rental companies, which ac- harder to ascertain. When coupled with uncertainties about count for over 80 percent of fleet purchases. Governments charging and how far customers will drive the rental vehicle, comprise the smallest category of fleet buyers, about 4.1 per- the business proposition for PEVs in rental fleets is unclear.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 74 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 59

FIGURE 3-7 Fleet sales for passenger vehicles for 2012 by fleet purchase agency. SOURCE: Based on data from Automotive Fleet (2013).

TABLE 3-8 Information Resources for Fleet Managers Resource URL EV Solutions: Operation Audits http://evsolutions.avinc.com/yourbusiness/fleets/gov

Vehicle Trends & Maintenance Costs Survey http://www.fleetanswers.com/sites/default/files/Dow%20Kokam%20 Survey%20Report_0.pdf

Plug‐In Vehicle Strategic Planning/Feasibility Study http://www.denvercleancities.org/Plug-In%20Vehicle%20Assessment- Template, Denver Metro Clean Cities Coalition %20Denver%20Metro%20Clean%20Cities%20with%20logo.pdf

Demand Assessment of First-Mover Hybrid and Electric http://www.calstart.org/Libraries/Publications/Demand_Assessment_of_First- Truck Fleets, CALSTART Mover_Hybrid_and_Electric_Truck_Fleets.sflb.ashx

Fleet Electrification Roadmap, Electrification Coalition http://www.electrificationcoalition.org/sites/default/files/EC-Fleet-Roadmap-screen.pdf

PG&E in California: PEV Case Study: It’s Electrifying: http://fleetanswers.com/sites/default/files/PGE%20case%20study%20Final.pdf Positive Returns in PEV Deployment, Electrification Coalition

FedEx: EV Case Study, The Electric Drive Bellwether?, http://www.fleetanswers.com/sites/default/files/FedEx_case_study.pdf Electrification Coalition

Joint Procurement of EVs and PHEVs in Sweden, Clean Fleets http://www.clean-fleets.eu/fileadmin/files/CF_case_study_sweden_04.pdf

Despite the uncertainty, Hertz is experimenting with PEVs plus the capital costs balance out in three years or less,” an in its rental fleets in key locations and where partners, such unlikely scenario for PEVs in the near term. as nearby hotels, are equipped to address charging (Hidary Electric utilities across the country provide an interest- 2012). ing opportunity for PEV fleet deployment. Given that many utilities are actively working with the vehicle industry in Corporate or Business Fleets PEV deployment, they should be one of the main fleet owners transitioning to PEVs. Their lessons could help to inform General Electric made big news when it announced in other fleet managers. The DOE Clean Cities program dis- 2010 that it would convert half of its fleet to PEVs (25,000 cussed above includes incentives to convert business fleets vehicles), of which one-half would be Chevrolet Volts (Rich- and offers information for fleet managers (DOE 2012); these ard 2010; Antich 2011). Many other companies, including resources will need to be updated as PEV deployment occurs Pepsi, Frito-Lay, and Cisco, have also stated objectives to and lessons are learned. green their fleets. For business fleets, issues related to limited choice of models, charging infrastructure, and higher initial Government Fleets prices compared with ICE vehicles pose barriers to adoption by fleets. Furthermore, fleet managers face challenges in try- The federal government has a vehicle fleet comprised ing to manage routes (Westervelt 2012; Hanson 2013). Un- of more than 600,000 vehicles and is, therefore, the nation’s like consumers, who do not appear to consider total cost of single largest fleet operator (GSA 2011). The General Services ownership when deciding whether to purchase a PEV, fleet Administration procures about 65,000 vehicles each year and managers attend carefully to such issues. As noted by Wolski owns and leases about 210,000 vehicles to federal agencies. (2013), “the real tipping point [for broad implementation of State, county, and municipal governments also have their own PEVs in commercial fleets] is when the total operating costs fleets.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 75 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

60 Overcoming Barriers to Deployment of Plug-in Electric Vehicles In 2011, President Obama ordered that by the end of best practices for dissemination to the private sector and 2015, all new light-duty vehicles purchased or leased by other government fleets. federal agencies be alternative-fuel vehicles, which include flex-fuel vehicles, HEVs, PEVs, compressed natural gas and REFERENCES biofuel vehicles (Obama 2011). The order also encouraged the agencies to support the development of alternative-fuel- Ad Council. 2014. “Become an Ad Council Campaign.” ing infrastructure. Through 2013, the majority of alternative- http://www.adcouncil.org/Working-With-Us/Become- fuel vehicles purchased by federal fleets have been flex-fuel an-Ad-Council-Campaign. Accessed October 23, 2014. vehicles that can operate on E-85 rather than vehicles fueled Adner, R. 2006. “Match Your Innovation Strategy to Your Inno- by electricity, natural gas, or other fuels (GSA 2013). Specif- vation Ecosystem.” Harvard Business Review, R0604F- ically, about half of reported federal fleet purchases in 2013 PDF-ENG, April. http://hbr.org/2006/04/match-your- were flex-fuel vehicles, and 36 percent were conventional innovation-strategy-to-your-innovation-ecosystem/ar/1. gasoline vehicles, some of which might even satisfy the Aggarwal, P., T. Cha, and D. Wilemon. 1998. Barriers to the mandate in areas where alternative fuels are not considered adoption of really-new products and the role of surrogate to be available. Few PHEVs or BEVs have been purchased; buyers. Journal of Consumer Marketing 15(4): 358-371. PEVs represent about 1.2 percent of reported federal fleet Anders, G. 2012. “Test Drive a Volt, Please!—How GM’s Pitch vehicle purchases. DOE itself has purchased few PEVs; only Backfired.” Forbes, July 18. http://www.forbes.com/sites/ 0.73 percent of its fleet are PEVs. georgeanders/2012/07/18/volt-test-drive/. Like individual consumers, the government faces bar- Antich, M. 2011. “Half of GE’s 30,000-Unit Fleet to be Elec- riers in adopting PEVs. The price of PEVs is a particularly tric Vehicles by 2015.” Automotive Fleet, February. high barrier for the federal agencies; tax incentives avail- http://www.automotive-fleet.com/channel/leasing/ar- able to consumers are not available for government fleets. ticle/story/2011/02/half-of-ge-s-30-000-unit-fleet-to-be- Another barrier is that vehicle purchases come from the electric-vehicles-by-2015.aspx. capital budgets, and fuel expenses come from operating bud- Automotive Fleet. 2013. “J.D. Power and LMC Automotive gets (DOE 2014d). The need to provide charging infrastruc- say U.S. Fleet Sales Improve in September.”September ture at its fleet facilities poses yet another barrier. Finally, 19. http://www.automotive-fleet.com/channel/vehicle-re- government procurement practices have been described as search/news/story/2013/09/j-d-power-and-lmc-automo- excessively complicated and lead some to wonder whether tive-say-u-s-fleet-sales-improve-in-september.aspx. government fleet sales are a realistic way to demonstrate the Axsen, J., and K.S. Kurani. 2009. Early U.S. market for suitability of PEVs in fleets. plug-in hybrid electric vehicles: Anticipating consumer The committee notes that although the total number of recharge potential and design priorities. Transportation vehicles in government fleets is small compared with the Research Record 2139: 64-72. total number of vehicles in the overall market, converting Beech, E. 2014. “Chrysler CEO: Please Don’t Buy Fiat some portion of the fleets to PEVs is important. First, people 500e Electric Car.” Reuters, May 21. http://www.re- expect leadership from their government. Given the man- uters.com/article/2014/05/21/chrsyelr-ceo-evs-idUSL- dates for energy security and clean transportation—the very 1N0O71MS20140521. motivations for this committee’s work—the symbolic im- Brook, D. 2014. “Should You Become a Carsharing Opera- portance of the government’s own efforts lend authenticity to tor?” Metavera, March 13. http://www.slideshare.net/Met the mandates. Second, the large number of people working at avera/should-you-become-a-car-sharing-operator-intro- all levels of government, particularly in the federal govern- duction. ment, could play a role in information diffusion and the edu- Brown, S., D. Pyke, and P. Steenhof. 2010. Electric vehicles: cation of friends and neighbors. Third, given that DOE is the The role and importance of standards in an emerging main government agency working to deploy PEVs, it should market. Energy Policy 38(7): 3797-3806. serve as a model by deploying PEVs in its own fleets. To Cahill, E., J. Davies-Shawhyde, and T. Turrentine. 2014. “Ze- explore ways to remove barriers to PEV deployment across ro-emission Vehicles and Retail Innovation in the U.S. the private sector while not removing barriers in its own or- Automotive Sector: An Exploration of the Consumer ganization is poor policy. Purchase Experience for Plug-in Electric Vehicles.” Uni- versity of California, Davis Institute of Transportation Recommendation: To lend authenticity to the federal gov- Studies Working Paper, August 2014. ernment’s initiative and to enhance the visibility of PEVs Caperello, N., J. Tyree-Hageman, and K. Kurani. 2014. “En- generally, the federal government should demonstrate lead- gendering the Future of Electric Vehicles: Conversa- ership by adding PEVs to its fleets and offering charging in- tions with Men and Women.” University of California, frastructure at its facilities. Davis Institute of Transportation Studies Working Paper UCD-ITS-WP-14. http://www.its.ucdavis.edu/research/ Recommendation: DOE should itself serve as a model by publications/publication-detail/?pub_id=2095. adding PEVs to its fleets and use its experience to discern

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 76 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 61 CNCDA (California New Car Dealers Association). 2015. DOE. 2014c. “$4.5 M in Funding Available for Alternative California Auto Outlook: Fourth Quarter 2014. Volume Fuel Market Readiness.” http://www1.eere.energy.gov/ 11, Number 1, February. http://www.cncda.org/CMS/ cleancities/news_detail.html?news_id=21555. Pubs/Cal_Covering_4Q_14.pdf. DOE. 2014d. Presentation at the NEXTSteps Symposia, Cobb, J. 2013. “December 2012 Dashboard.” Hybrid Cars, University of California at Davis. January 8. http://www.hybridcars.com/december-2012- Edmunds. 2013. 4th Quarter-2013 Used Market Quarterly dashboard. Report. http://www.edmunds.com/industry-center/data/ Cobb, J. 2015. “December 2014 Dashboard.” Hybrid Cars, used-car-market-quarterly-report.html. January 6. http://www.hybridcars.com/december-2014- Egbue, O., and S. Long. 2012. Barriers to widespread adop- dashboard/. tion of electric vehicles: An analysis of consumer atti- Colias, M. 2014. “Cadillac’s European Retail Strategy May Be tudes and perceptions. Energy Policy 48: 717-729. Adapted for U.S.” Automotive News, March 24. http:// EIA (U.S. Energy Information Administration). 2014a. An- www.autonews.com/article/20140324/RETAIL/30324 nual Energy Outlook 2014 with Projections to 2040. 9978/cadillacs-european-retailstrategy-may-be-adapted- Washington, DC: U.S. Department of Energy, April. for-u.s. http://www.eia.gov/forecasts/aeo/pdf/0383(2014).pdf. Consumers Union and the Union of Concerned Scientists. EIA. 2014b. “Table 5.6.A. Average Retail Price of Electricity 2013. Electric Vehicle Survey Methodology and Assump- to Ultimate Customers by End-Use Sector.” Electric Pow- tions: American Driving Habits, Vehicle Needs, and Atti- er Monthly. http://www.eia.gov/electricity/monthly/epm tudes Towards Electric Vehicles. December. http://www. _table_grapher.cfm?t=epmt_5_6_a. Accessed December ucsusa.org/assets/documents/clean_vehicles/UCS-and- 30, 2014. CU-Electric-Vehicle-Survey-Methodology.pdf. EIA. 2014c. “Petroleum and Other Liquids: Weekly Retail CSE (Center for Sustainable Energy). 2014. “California Air Gasoline and Diesel Prices.” http://www.eia.gov/dnav/ Resources Board Clean Vehicle Rebate Project, EV Con- pet/pet_pri_gnd_a_epm0_pte_dpgal_m.htm. sumer Survey Dashboard.” http://energycenter.org/clean- El-Moursi, S. 2013. “Rent Green: Will Electric and Hybrid vehicle-rebate-project/survey-dashboard. Car Rentals Take Off?” Buzz: The Power2Switch Blog, Cully, M. 2014. Presentation to the Committee on Overcom- March 1. https://power2switch.com/blog/rent-green-willing Barriers to Electric-Vehicle Deployment, Washing- electric-and-hybrid-car-rentals-take-off/. ton, DC, May 6. EnergyNews. 2013. “Madrid City Council Goes for Car Darvish, T. 2013. Presentation to the Committee on Over- Sharing.” http://www.energynews.es/english/nh-hoteles coming Barriers to Electric-Vehicle Deployment, Wash- -se-apunta-a-la-movilidad-el-ctrica-1365489521/. ington, DC, December 3. EPA (U.S. Environmental Protection Agency). 2014. Trends Dediu, H. 2012. “When Will Smartphones Reach Saturation Report. Light-Duty Automotive Technology, Carbon in the US?” Asymco, April 11. http://www.asymco.com/ Dioxide Emissions, and Fuel Economy Trends: 1975 2012/04/11/. Through 2014. EPA-420-R-14-023. October. http://epa. DOE (U.S. Department of Energy). 2011. Clean Cities Five- gov/fueleconomy/fetrends/1975-2014/420r14023.pdf. Year Strategic Plan: Reducing Petroleum Dependence EPA/NHTSA (U.S. Environmental Protection Agency and in On-Road Transportation in the United States. https:// National Highway Traffic Safety Administration). www1.eere.energy.gov/cleancities/pdfs/strategic_plan. 2012. “2017 and Later Model Year Light-Duty Vehicle pdf. Greenhouse Gas Emissions and Corporate Average Fuel DOE. 2012. Plug-in Electric Vehicle Handbook for Fleets. Economy Standards.” Federal Register 77(199): Octo- Alternative Fuels Data Center. http://www.afdc.energy. ber. http://www.gpo.gov/fdsys/pkg/FR-2012-10-15/pdf/ gov/pdfs/pev_handbook.pdf. 2012-21972.pdf. DOE. 2013a. “Alternative Fuels and Advanced Vehicles.” Eppstein, M.J., D.K. Grover, J.S. Marshall, D.M. Rizzo. Alternative Fuels Data Center. http://www.afdc.energy. 2011. An agent-based model to study market penetra- gov/fuels/. tion of plug-in hybrid electric vehicles. Energy Policy DOE. 2013b. “Vehicle Cost Calculator.” Alternative Fuels 39(6): 3789-3802. Data Center. http://www.afdc.energy.gov/calc/. Esty, D., and A. Winston. 2009. Green to Gold: How Smart DOE. 2014a. “Alternative Fueling Station Locator.” Alterna- Companies Use Environmental Strategy to Innovate, tive Fuels Data Center. http://www.afdc.energy.gov/loca- Create Value, and Build Competitive Advantage. Hobo- tor/stations/. ken, NJ: Wiley. DOE. 2014b. “Building Partnerships to Reduce Petroleum Evarts, E. 2014. Dealers not always plugged in about electric Use in Transportation.” http://www1.eere.energy.gov/ cars, Consumer Reports’ study reveals. Consumer Reports, cleancities/. April. http://www.consumerreports.org/cro/news/2014/ 04/dealers-not-always-plugged-in-about-electric-cars- secret-shopper-study-reveals/index.htm.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 77 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

62 Overcoming Barriers to Deployment of Plug-in Electric Vehicles FHWA (Federal Highway Administration). 2011. Summary Herbig, P.A., and H. Kramer. 1994. The effect of information of Travel Trends: 2009 National Household Travel Sur- overload on the innovation choice process: Innovation vey. FHWA-PL-11-022. U.S. Department of Transpor- overload. Journal of Consumer Marketing 11(2): 45-54. tation. http://nhts.ornl.gov/2009/pub/stt.pdf. Hidary, J. 2012. “New Models of Mobility and EV Deploy- FHWA. 2012a. “Public Road Mileage-VMT-Lane Miles ment.” Presentation to the Committee on Overcoming 1920-2012.” Office of Highway Policy Information Barriers to Electric-Vehicle Deployment, Washington, Highway Statistics. http://www.fhwa.dot.gov/policyin- DC, December 18. formation/statistics/2012/vmt421c.cfm. Ingram, A. 2013. “Many Consumers Still Unaware Of Elec- FHWA. 2012b. “An Action Plan to Integrate Plug-in Elec- tric Car Incentives.” Green Car Reports, November 19. tric Vehicles with the U.S. Electrical Grid.” http://www. http://www.greencarreports.com/news/1088583_many- fhwa.dot.gov/environment/climate_change/energy/pub- consumers-still-unaware-of-electric-car-incentives. lications_and_tools/pev_action_plan/index.cfm. Ingram, A. 2014. “Electric Car ‘Guest Drive’ Service Launch- FHWA. 2014. “New Data Show U.S. Driving at Highest Lev- es in CA To Get ‘Butts In Seats’.” Green Car Reports, May el in Six Years.” https://www.fhwa.dot.gov/pressroom/ 2. http://www.greencarreports.com/news/1091852_elec fhwa1430.cfm. tric-car-guest-drive-service-launches-in-ca-to-get- FHWA. 2015. “Highway Statistics 2013. Annual Vehicle Dis- butts-in-seats. tance Traveled in Miles and Related Data-2013-by High- Kahle, L., and L. Chiagouris. 1997. Values, Lifestyles, and way Category and Vehicle Type.” Office of Highway Pol- Psychographics. Mahwah, New Jersey: Lawrence Erl- icy Information. http://www.fhwa.dot.gov/policyinforma baum Associates, Inc. tion/statistics/2013/vm1.cfm. Kahneman, D. 2013. Thinking, Fast and Slow. New York: Frades, M. 2014. A Guide to the Lessons Learned from the Clean Farrar, Straus and Giroux. Cities Community Electric Vehicle Readiness Projects. Krause, R. M., S.R. Carley, B.W. Lanea, and J.D. Graham. http://www.afdc.energy.gov/uploads/publication/guide 2013. Perception and reality: Public knowledge of plug- _ev_projects.pdf. in electric vehicles in 21 U.S. cities. Energy Policy 63: Frank, T., F. Neumann, F. Butler, P. Cocron, and J. Krems. 2011. 433-440. Experiencing range in an electric vehicle: Understand- LeBeau, P. 2012. “Americans Buying Fewer New Cars in ing psychological barriers. Applied Psychology 61 (3): Lifetime.” CNBC, October 22. http://www.cnbc.com/id/ 368-391. 49504504. Friedman, M. 2014. “Fate of Bill to Allow Tesla to Restart Sales Leonard Barton, D., E.Wilson, and J. Doyle. 1995. Com- Could Hinge on Pro-Dealership Bill.” NJ.com, July 5. mercializing Technology: Understanding User Needs. http://www.nj.com/politics/index.ssf/2014/07/tesla_sh In Business Marketing Strategy, edited by V.K. Rangan owrooms_previously_banned_may_return_to_new_jer- et al., 281-305. Chicago: Irwin. sey.html. Lim, M., H. Mak, and Y. Rong. 2014. “Toward Mass Adop- Gilbert, B. 2014. “Why Tesla Motors Can’t Sell Cars in Most tion of Electric Vehicles: Impact of the Range and Resale of the United States.” Engadget, July 17. http://www.enga Anxieties.” http://papers.ssrn.com/sol3/papers.cfm?abs dget.com/2014/07/17/tesla-motors-us-sales/. tract_id=2236560. GM Authority. 2010. “Chevrolet Announces Voltec 240V Loveday, E. 2013a. “Consumer Reports Says Leasing is the Home Charger For $490.” http://gmauthority.com/blog/ Right Financial Choice for Today’s Electric Vehicles.” 2010/10/chevrolet-annouces-voltec-240v-home-char- InsideEVs, June 28. http://insideevs.com/consumer-re ger-for-490. ports-says-leasing-is-the-right-financial-choice-for-to- GSA (U.S. General Services Administration). 2011. “Elec- days-electric-vehicles/. tric Vehicle Pilot Program.” http://www.gsa.gov/portal/ Loveday, E. 2013b. “Are Electric Vehicles Being Sold at content/281581?utm_source=OCM&utm_medium= a Loss? Kelley Blue Book Seems to Think So.” In- print-radio&utm_term=electricvehicles&utm_ sideEVs, July 18. http://insideevs.com/are-electric-ve- campaign=shortcuts. hicles-being-sold-at-a-loss-kelley-blue-book-seems-to- GSA. 2013. Federal Fleet Report. http://www.gsa.gov/por- think-so/. tal/category/102859. Lutz, B. 2012. “The Real Story On GM’s Volt Costs.” Hanson, S. 2013. “Frito-Lay Electric Vehicles: Why, How Forbes, September 10. http://www.forbes.com/sites/bo- and What Is Next.” Presentation to the Committee on blutz/2012/09/10/the-real-story-on-gms-volt-costs/. Overcoming Barriers to Electric-Vehicle Deployment, Mahajan, V., E. Muller, and F.M. Bass. 1990. New product Washington, DC, December 3. diffusion models in marketing: A review and directions Henry, J. 2012. “Average Car in the U.S. Now Over 10 Years for research. The Journal of Marketing 54: 1-26. Old, a Record.” Forbes, January 20. http://www.forbes. McQuarrie, E. 2008. Customer Visits: Building a Better com/sites/jimhenry/2012/01/20/average-car-in-the-u-s- Market Focus. 3rd edition. Armonk, NY: Sharpe. now-over-10-years-old-a-record/.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 78 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Understanding the Customer Purchase and Market Development Process for Plug-in Electric Vehicles 63 METI (Ministry of Economy, Trade, and Industry). 2010. Richard, M.G. 2010. “GE to Purchase 25,000 Electric Cars Next Generation Vehicle Plan. http://www.meti.go.jp/ by 2015!” TreeHugger, November 11. http://www.tree- english/press/data/pdf/N-G-V2.pdf. hugger.com/cars/ge-to-purchase-25000-electric-cars- Mohr, J., S. Sengupta, and S. Slater. 2010. Marketing of by-2015.html. High-Technology Products and Innovations. 3rd edi- Rogers, E. 2003. The Diffusion of Innovations. 5th edition. tion. Upper Saddle River, NJ: Prentice Hall. New York: The Free Press. Mohr, J., S. Slater, and S. Sengupta. 2006. Foundations for Scheibehenne, B., R. Greifeneder, and P.M. Todd. 2010. Can Successful High-Technology Marketing. In Managing there ever be too many options? A meta-analytic review Technology and Innovation: An Introduction, edited by of choice overload. Journal of Consumer Research 37(3): R. Verburg, R. J. Ortt, and W. M. Dicke, 84-105. Lon- 409-425. don: Routledge Publishing. Schwartz, B. 2005. Paradox of Choice: Why More Is Less. Moore, G. 2014. Crossing the Chasm: Marketing and Sell- New York: Harper Perennial. ing Technology Products to Mainstream Customers. 3rd Seybold, P. 2001. Get inside the lives of your customers. edition. New York: Harper Business. Harvard Business Review 79(5): 81-89. Motavalli, J. 2013. “Ford Switches to AeroVironment for Shaheen, S., and A. Cohen. 2013. “Transportation Sustain- Chargers, Drops Best Buy.” Plug-in Cars, May 10. http:// ability Research Center: North American Carsharing www.plugincars.com/best-buy-out-aerovironment- 2012-2013 Outlook.” 2013 Carsharing Association Con- fords-charging-partner-127183.html. ference. Toronto, Canada, September 16-17. http://www. Nelson, G. 2014. “NADA Moves to Counter Franchise Sys- slideshare.net/AlanWoodland/dr-susan-shaheen-north- tem Critics.” Automotive News, June 16. http://www.au- american-carsharing-20122013-outlook. tonews.com/article/20140616/RETAIL06/306169944/ Shepard, S., and J. Gartner. 2014. Electric Vehicle Geo- nada-moves-to-counter-franchise-system-critics. graphic Forecasts. Navigant Research. Boulder, CO. NRC (National Research Council). 2013a. Transitions to Sivak, M. 2014. Has Motorization in the U.S. Peaked? Part Alternative Vehicles and Fuels. Washington, DC: The 4: Households without a Light-Duty Vehicle. Report No. National Academies Press. 2014-5. University of Michigan Transportation Research NRC. 2013b. Overcoming Barriers to Electric-Vehicle De- Institute, Ann Arbor, MI. http://deepblue.lib.umich.edu/ ployment, Interim Report. Washington, DC: The Na- bitstream/handle/2027.42/102535/102988.pdf. tional Academies Press. Solomon, M. 2014. Consumer Behavior: Buying, Having, Obama, B. 2011. “Federal Fleet Performance.” Presidential and Being. 11th edition. Boston: Prentice Hall. memorandum. May 24. https://www.whitehouse.gov/the- Strategic Vision. 2013. New Vehicle Experience Studies of press-office/2011/05/24/presidential-memorandum-fed- Vehicle Registrants. San Diego, CA. eral-fleet-performance. Strategic Vision. 2014. New Vehicle Experience Studies of Parasuraman, A., and C. Colby. 2001. Techno-Ready Mar- Vehicle Registrants. San Diego, CA. keting: How and Why Customers Adopt Technology. Sultan, F., J. U. Farley, and D. R. Lehmann. 1990. A meta- New York: The Free Press. analysis of applications of diffusion models. Journal of Pike Research. 2012. Energy & Environment Consumer Survey Marketing Research 27(1): 70-77. Consumer Attitudes and Awareness about 13 Clean En- Tal, G., M.A. Nicholas, J. Woodjack, and D. Scrivano. 2013. ergy Concepts: Published 1Q 2012. http://www.navigan Who Is Buying Electric Cars in California? Exploring tresearch.com/wordpress/wp-content/uploads/2012/02/ Household and Vehicle Fleet Characteristics of New EECS-12-Pike-Research.pdf. Plug-in Vehicle Owners. Research Report UCD- ITS- Plug-in America. 2013. “National Drive Electric Week 2013 RR-13-02. Institute of Transportation Studies, Univer- Stats.” https://driveelectricweek.org/stats.php?year=2013. sity of California, Davis. Quinland, R.M. 2013. Has the traditional automobile fran- Turrentine, T.S., D.M. Garas, A.H. Lentz, and J.F. Woodjack. chise system run out of gas? The Franchise Lawyer 2011. The UC Davis MINI E-Consumer Study. UCD- 16(3). http://www.americanbar.org/publications/franchi ITS-RR-11-05. Institute of Transportation Studies, Uni- se_lawyer/2013/summer_2013/has_traditional_auto- versity of California, Davis. mobile_franchise_system_run_out_gas.html. UC Davis. 2014. “Plug-in Hybrid and Electric Vehicle Re- Rai, V., and V. Nath. 2014. “How the Interaction of Supply search Center of the Institute of Transportation Studies.” and Demand Shapes Patterns of New Technology Adop- PEV Market Briefing, May 2014. http://policyinstitute. tion: Plug-In Electric Vehicles in California.” Paper pre- ucdavis.edu/files/PEV-Market-Briefing-Turrentine- sented at the International Association for Energy Eco- 28May2014.pdf. nomics, New York, NY, June 16. U.S. Census Bureau. 2012. “2012 Economic Census, 2012 Rayport, J., and D. Leonard Barton. 1997. Spark innovation Economic Census of Island Areas, and 2012 Nonem- through empathic design. Harvard Business Review ployer Statistics.” Industry Statistics Portal NAICS 4471. 75(6): 102-113. http://www.census.gov/econ/isp/.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 79 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

64 Overcoming Barriers to Deployment of Plug-in Electric Vehicles U.S. Census Bureau. 2013a. Income, Poverty, and Health Wells, W. 2011. Life Style and Psychographics. Chicago: Insurance Coverage in the United States: 2012. https:// Marketing Classics Press. www.census.gov/prod/2013pubs/p60-245.pdf. Westervelt, A. 2012. “Electric Vehicles Lead Pack in Green- U.S. Census Bureau. 2013b. “Physical Housing Character- ing Corporate Fleets.” Green Biz, March 30. http://www. istics for Occupied Housing Units: 2009-2013 Ameri- greenbiz.com/blog/2012/03/30/electric-vehicles-corpo- can Community Survey 5-year Estimates.” American rate-fleet. FactFinder. http://factfinder.census.gov/bkmk/table/1.0/ Wolski, C. 2013. “What Is the Future of Electric Vehicles in en/ACS/13_5YR/S2504/0100000US. Accessed January Fleet?” Green Fleet, September. http://www.greenfleet 22, 2015. magazine.com/article/51755/what-is-the-future-of- Voelcker, J. 2013a. “Next Chevy Volt $10,000 Cheaper to electric-vehicles-in-fleet/p/2. Build, Profitable for GM?” Green Car Reports, May 2. Zmud, J., V. Barabba, M. Bradley, J.R. Kuzmyak, M. Zmud, http://www.greencarreports.com/news/1083890_next- and D. Orrell. 2014. Strategic Issues Facing Transporta- chevy-volt-10000-cheaper-to-build-profitable-for-gm. tion: Volume 6, The Effects of Socio-Demographics on Voelcker, J. 2013b. “Why Some Dealers Are Inept at Selling Future Travel Demand. NCHRP Report 750. Transpor- Plug-In Electric Cars.” Green Car Reports, December 12. tation Research Board, Washington, DC. http://online- http://www.greencarreports.com/news/1089055_why- pubs.trb.org/onlinepubs/nchrp/nchrp_rpt_750-v6.pdf. some-dealers-are-inept-at-selling-plug-in-electric-cars. Voelcker, J. 2014. “Many Car Dealers Don’t Want to Sell Electric Cars: Here’s Why.” Green Car Reports, February 14. http://www.greencarreports.com/news/1090281_ma ny-car-dealers-dont-want-to-sell-electric-cars-heres- why.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 80 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles

The successful deployment of plug-in electric vehicles differential between PEVs and internal-combustion engine (PEVs) involves many entities and will require the resolution (ICE) vehicles. Battery cost will need to decrease substan- of many complex issues. The present report focuses on indi- tially to allow PEVs to become cost competitive with ICE vidual strategies for overcoming barriers related to purchas- vehicles (see Chapters 2 and 7). Thus, the current goal of bat- ing and charging PEVs, and this chapter specifically explores tery research and development is to increase the energy den- how federal, state, and local governments and their various sity of PEV batteries and to lower their cost. The improved administrative arms can be more supportive and implement battery technology can then be used to lower vehicle cost, policies to sustain beneficial strategies for PEV deployment. increase vehicle range, or both, and those improvements Although electric utilities can also provide institutional sup- would likely lead to increased PEV deployment. port for PEV deployment, they and their associated poli- As in many areas of fundamental research and develop- cies are discussed in Chapter 6. Where opportunities exist ment, the federal government has an important role to play. to improve the viability of PEVs but no single institution is Although basic science and engineering research is funded clearly positioned to capitalize on the opportunity, the com- by both government and the private sector, the government mittee highlights possible partnerships that might fill these role is to fund long-term, exploratory research that has the voids. The committee’s findings and recommendations are potential for positive national impact. Stable funding for ex- provided throughout this chapter. ploratory research allows investments in research facilities and human capital that are necessary for the research to bear FEDERAL GOVERNMENT RESEARCH fruit. The federal government has directly supported battery FUNDING TO SUPPORT DEPLOYMENT research and development for electric vehicles since 1976 OF PLUG-IN ELECTRIC VEHICLES (Electric and Hybrid Vehicle Research, Development, and Demonstration Act 1976, Pub. L. 94-413). Past investment Funding research is one of the most important ways the in research and development contributed to the development federal government can lower barriers to PEV deployment. of the NiMH batteries used in early hybrid electric vehicles Research is needed in two areas in particular. As discussed in (HEVs) and to the lithium-ion battery technology used in the Chapter 2, the first is basic science and engineering research Chevrolet Volt (DOE 2008). to lower the cost and improve the energy density and other The largest funder of energy storage research in the fed- performance characteristics of batteries. The second critical eral government is DOE, followed by DOD. From 2009 to area concerns PEV deployment, especially the role of infra- 2012, across all areas of the federal government, investment structure in spurring vehicle sales and increasing electric in energy-storage research, development, and technology de- vehicle miles traveled (eVMT). Fundamental and applied ployment totaled $1.3 billion, which includes batteries for all science and engineering research for vehicle energy stor- applications, not only vehicles (GAO 2012a). In Fiscal Year age is being undertaken by vehicle manufacturers and in the (FY) 2013, the DOE Vehicle Technology Office funded $88 laboratories of the U.S. Department of Energy (DOE), the million for battery research and development focused on ve- Department of Defense (DOD), and academic institutions. hicle applications (DOE 2014a). Much of the funding is for Research into the deployment of PEV infrastructure and grants or cooperative research agreements with government, markets is much less developed. Both areas are discussed industry, or university laboratories, but a growing proportion below. is also funding loan guarantees to deploy new technologies. Worthy DOE goals for battery storage improvements include Engineering Research and halving the size and weight of PEV batteries and reducing Development of Battery Science the production costs to one quarter of its 2012 value by 2022 (DOE 2013a). Recently, DOE has initiated and supported As discussed in Chapter 2, the battery is the most costly several collaborative research programs with even more component of PEVs and represents the majority of the cost ambitious goals to accelerate basic and applied research,

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 81 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

66 Overcoming Barriers to Deployment of Plug-in Electric Vehicles development, and deployment. They include Energy Fron- an important consideration in the United States, there were tier Research Centers, several Advanced Research Projects clearly limitations on the tracking data that could be shared Agency-Energy (ARPA-E) programs in energy storage, and with researchers. Data collection ended as of December the Battery and Energy Storage Hub, which is funded at up 2013, but data analysis continues. to $25 million per year for 5 years and aims to increase battery energy density five times and reduce cost by 80 percent Finding: Research is critically needed in understanding the (DOE 2013b). relationship between infrastructure deployment and PEV adoption and use. Finding: Investment in battery research is critical for producing lower cost, higher performing batteries that give Recommendation: The federal government should fund re- PEVs the range consumers expect from ICE vehicles. search to understand the role of public charging infrastructure (as compared with home and workplace charging) in Recommendation: The federal government should continue encouraging PEV adoption and use. to sponsor fundamental and applied research to facilitate and expedite the development of lower cost, higher performing Recommendation: A new research protocol should be de- vehicle batteries. Stable funding is critical and should focus signed that would facilitate access to raw charging data to on improving energy density and addressing durability and relevant stakeholders within the confines of privacy laws. safety. INSTITUTIONAL SUPPORT FOR PROMOTING Research on Deployment of Plug-in Electric Vehicles PLUG-IN ELECTRIC VEHICLE READINESS

In contrast to the substantial investment in battery re- The concept of PEV readiness refers to an entire ecosys- search and development, research on PEV deployment is tem of automotive technology, including its supporting infra- much less advanced. A critical research need is understanding structure, regulations, financial incentives, consumer informa- the relationship between PEV deployment and infrastructure tion, and public policies, programs, and plans that can make deployment. Supporting that research is an appropriate role PEVs a viable choice for drivers. Several tools have been cre- for the federal government given that it might be motivated ated to assess whether a given organization, community, state, to deploy infrastructure if by doing so it encourages PEV or even country has in place the essential elements to be con- deployment and increases eVMT. sidered PEV ready. Examples of assessment tools include the The primary DOE effort to understand PEV vehicle and Rocky Mountain Institute’s Project Get Ready (Rocky Moun- infrastructure deployment is the EV Project, an infrastructure tain Institute 2014), DOE’s Plug-in Electric Vehicle Readiness deployment and evaluation program managed by the Idaho Scorecard (DOE 2014a), Michigan Clean Energy Coalition’s National Laboratory (INL) in partnership with ECOtality. Plug-in Ready Michigan (Michigan Clean Energy Coalition Around the time of the most recent wave of PEVs in 2009, 2011), California PEV Collaborative’s PEV Readiness Toolkit DOE awarded in 2009 a $99.8 million grant for deployment (CAL PEV 2012a), and the Center for Climate and Energy of charging infrastructure in private residences and in public Solutions’ State DOT PEV Action Tool (C2ES 2014). Further- areas in 20 of the target launch markets of the Nissan Leaf more, $8.5 million has been provided through the DOE Clean and the Chevrolet Volt, including San Francisco, Seattle, Cities program to 16 projects across 24 states to assess PEV San Diego, Los Angeles, Portland, and Nashville. The pro- readiness and develop specific plans to enable the communi- gram has grown with an additional $15 million grant from ties to become PEV ready (DOE 2014a). Table 4-1 indicates DOE and partner matches from the vehicle manufacturers the many common factors that constitute PEV readiness and and charging providers to a total of $230 million (ECOtal- the different institutions or organizations that might have a ity 2013; INL 2014a). When it concluded collecting data role to play. in December 2013, over 8,200 vehicles were participating State governments will be particularly important actors and over 8,200 residential chargers, 3,500 public AC level 2 in supporting PEV deployment. Most supportive PEV actions chargers, and 107 DC fast chargers had been installed (Smart at the state level can be carried out by various administrative and White 2014; INL 2014b). agencies, including environmental and clean air agencies, The EV Project included data collection on where and utility commissions, departments of energy, transportation when the vehicles in the project charged so that DOE could agencies, licensing and inspection agencies, general services learn more about how drivers were using the vehicles and the agencies, and workforce training or education agencies. In the associated charging infrastructure. Thus, the data provided committee’s interim report, the committee noted several areas important information about early adopters of PEVs in large where the federal government could play a convening role to metropolitan areas, including location of charging, eVMT, coordinate state and local government activities in support of impacts on utilities, impact of workplace charging, and re- the emerging PEV sector (NRC 2013, pp. 2, 4, 52). gional variations in charging behavior. Because privacy is

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 82 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 67

TABLE 4-1 Factors Determining PEV Readiness and Organizations Involved Readiness Feature Federal Government State Government Municipal Government Electric Utility Private Industry Permit streamlining __ Environmental and Building and __ __ archeological electrical codes Utility regulatory policies __ PUC regulation of Muni-owned cost __ __ cost recovery and recovery policies retail markets Building code requirements Model ordinances Model state ordinances Local ordinances __ PEV-ready buildings Infrastructure DOE funding, Interregional and Regional and Distribution network Strategic investment deployment plans assistance, and interstate plans metropolitan area plans and capacity plans and sites dissemination Land use and Federal regulations State regulations Comprehensive plans __ __ uniform signage and policies and zoning Electricity pricing policies NIST metering and State laws and PUC Muni-owned policies Smart grid and EVSE pricing strategies pricing standards rate regulation and technology metering technologies Training personnel __ Workforce training First-responder __ Skilled trades and permits safe practices

Vehicle financial incentives PEV subsidies Rebates, tax exemptions Utility taxes, parking Rebates Equity investments, from registration, tolls fees financing Infrastructure financial Equipment subsidies Equipment subsidies Equipment subsidies, Cost sharing in Workplace and fleet incentives land gifts any upgrades, charging equipment subsidies

Energy policies Clean energy programs Zero-emission-vehicle TOU or special TOU or special Green power programs standards PEV rates PEV rates Dealership franchise laws __ State laws and __ __ Vehicle manufacturers’ regulations policies and practices Environmental policies EPA regulations Clean air laws Carbon reduction plans Clean power __ and regulations generation Procurement policies GSA regulations State purchasing Purchasing cooperatives __ Bulk purchase discounts and goals and policies and bulk orders Business policies and Research and State-backed Municipal-owned Own or Innovative financing permissible models demonstration projects financing assistance infrastructure operate EVSE NOTE: DOE, U.S. Department of Energy; EPA, U.S. Environmental Protection Agency; EVSE, electric vehicle supply equipment; GSA, General Services Administration; NIST, National Institute of Standards and Technology; PEV, plug-in electric vehicle; PUC, public utility commission; TOU, time of use.

TRANSPORTATION TAXATION AND to be widespread misunderstanding about the extent to which FINANCING ISSUES RELATED TO PEVs currently pay transportation taxes and the resulting PLUG-IN ELECTRIC VEHICLES fiscal impacts to transportation budgets both now and into the future. This section explores the issue in depth, attempts One potential barrier for PEV adoption that is solely to bring more clarity to current tax policy and impacts, and within the government’s direct control is taxation of PEVs, 1 makes recommendations for how transportation-tax policy in particular, taxation for the purpose of recovering the cost might be harmonized with a transportation innovation policy of maintaining, repairing, and improving the roadways. As for PEVs. described below, the paradigm for roadway taxation in the United States has depended on motor fuel taxes, which are Current State of Transportation Taxation indirect user fees. The advent of PEVs poses a dilemma for public officials responsible for transportation-tax policy be- Motor fuel taxes have been the most important single cause battery electric vehicles (BEVs) use no gasoline and source of revenue for funding highways for nearly a century plug-in hybrid electric vehicles (PHEVs) use much less than and have also been an important source of transit funding ICE vehicles.2 To further complicate matters, there appears since the 1980s (TRB 2006, pp. 24-36). The state of Oregon instituted the nation’s first per-gallon tax on gasoline in 1919 1 Chapter 7 addresses the issue of tax incentives; this chapter dis- (ODOT 2007). Within 10 years after that, every state had en- cusses tax disincentives. acted a fuel tax. The federal government did not enact a fuel 2 The amount of gasoline used by a PHEV depends on the all- tax until 1932 and did not dedicate the tax to transportation electric range and the frequency with which the vehicle is charged.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 83 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

68 Overcoming Barriers to Deployment of Plug-in Electric Vehicles projects until 1956 (FHWA 1997, Chapter IV). At the time of Fuel consumption depends on both the number of miles their introduction, fuel taxes were viewed as the most eco- driven and the fuel economy of the vehicle fleet. Therefore, nomical method of collecting a fee for roadway construction any decrease in the number of miles driven or increase in and maintenance from those who directly benefited: motor- the fuel economy of the vehicle fleet will result in less tax vehicle operators. However, the share of highway spending revenue generated for a cost-per-gallon tax. One of those covered by fuel-tax revenues has been declining. In 2012, factors can offset the other and moderate the negative effect fuel taxes accounted for 59 percent of all federal, state, and on the revenue stream. For example, the fuel economy of the local highway-user revenues (fuel taxes, fees, and tolls) used light-duty vehicle fleet has been increasing since 2005 (EPA for highways and 28 percent of total government disburse- 2013). From 2005 to 2007, light-duty vehicle miles traveled ments for highways (FHWA 2014, Table HF-10). (VMT) also increased, which helped mask the negative ef- For most of the past century, the fuel tax has been viewed fect on the revenue stream of improving fleet fuel economy. as a reasonably fair and reliable tax revenue to fund transpor- However, VMT and fuel consumption both declined with the tation. The fuel economy of most vehicles remained fairly recession in 2007 and 2008 and have remained flat since then consistent across different models (NSTIFC 2009) as there (Figure 4-3). Without the revenue-bolstering effect from in- were no strong incentives (such as increasing gasoline prices creasing VMT, transportation budgeters and policy makers or stricter government regulation) to improve fuel economy. have become acutely aware of how rising fleet vehicle econ- However, the 1973 Yom Kippur War and resulting oil Arab omy affects transportation fund balances. embargo served as the marker for the U.S. policy shift to reduce the nation’s petroleum dependence by improving ve- Federal and State Concerns hicle fuel economy. In later years, the federal government enacted Corporate Average Fuel Economy (CAFE) regula- With the recent increases in federal CAFE standards,3 tions, which essentially mandated improved fuel economy in the flattening of VMT, and political opposition to raising passenger vehicles (see Figure 4-1). the tax rate itself, federal and state officials are increasingly Both the federal government (see Figure 4-2) and the concerned with the potential effects of high-mpg vehicles on states rely heavily on motor-fuel tax revenue, which includes their transportation budgets. The poster child for their wor- taxes on gasoline and diesel, to maintain the transportation ries is the BEV, which uses no gasoline and whose drivers system. At the federal level and in the vast majority of states, therefore pay no fuel tax. fuel taxes are based on a flat cents-per-gallon tax levied on A recent survey of 50 state departments of transportation motor fuel; the extent of reliance on the fuel taxes varies (DOTs) reflected the strong sentiment that PEVs threaten loss from state to state (Rall 2013). For example, gasoline taxes of revenue for transportation. The majority of state DOTs re- range from $0.08 per gallon in Alaska to $0.53 per gallon in sponded that they would support federally led field tests of California (the nationwide average is $0.31 per gallon) (Rall mileage fees for PEVs to improve the equity and sustainability 2013). Of all government tax and fee revenues used for high- of Highway Trust Fund revenues (GAO 2012b, p. 45). ways in 2012, 20 percent came from the federal government, 3 49 percent from state governments, and 31 percent from lo- 49 CFR Parts 523, 531, 533 et al., 2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Av- cal governments (FHWA 2014, Table HF-10). erage Fuel Economy Standards; Final Rule.

FIGURE 4-1 Corporate Average Fuel Economy requirements by year. SOURCE: DOE (2013c).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 84 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 69

FIGURE 4-2 Sources of revenue for the federal Highway Trust Fund, FY 2010. These revenue sources exclude transfers from the general fund because those are not considered revenues in the federal nomenclature. SOURCE: GAO (2012b, p. 6).

A common refrain is that “PEVs pay nothing to use the The committee recognizes that PEVs and current trans- highways” because they use little if any gasoline (Battaglia portation tax policies raise the following important questions: 2013). That is not, however, the case. At the federal level, the highway trust fund has relied on transfers of general tax • Is the difference in transportation taxes collected from revenues to maintain sufficient balances to meet its transpor- PEVs and ICE vehicles significant in the context of fed- tation funding obligations (GAO 2011). Therefore, all U.S. eral or state transportation budgets, either now or in the taxpayers—including PEV drivers—are paying for the fed- near future? eral transportation system from their general tax payments, • Even if the amount of unrealized revenue is negligible, in addition to the 18.4 cent per gallon federal gasoline tax. do PEVs raise issues of fairness in the user-pays prin- That misunderstanding is even more acute at the state ciple underlying the U.S. transportation tax system that level, where many states and local governments levy a myr- has been in place for almost a century? iad of taxes and fees that are dedicated to transportation, • To remedy the issues inherent in the first two questions, including roadway funding.4 Specifically, most local trans- should PEVs be a focus for new methods of taxation, portation funding comes from property taxes, general fund considering that the unrealized revenue from high-mpg appropriations, and fares for mass transit; at the state level, vehicles will dwarf that of PEVs? motor fuel taxes are significant, but motor vehicle taxes, • Are there other intervening policy considerations that fees, and other revenue, such as sales taxes, play important might trump the general transportation tax paradigm of roles. Washington State recently estimated that, on average, user pays, at least for a period of time? BEV drivers pay $210 per year in transportation-related state and local taxes and fees even though they pay no fuel taxes Finding: It is not true that PEV drivers pay nothing for the (WSDOT 2013).5 That equates to 44 percent of what is paid maintenance and use of the transportation system given vari- by the average gasoline-powered passenger vehicle in that ous transportation-related taxes and fees that must be paid state. Figure 4-4 compares transportation-related taxes paid by all vehicle drivers. It is true that BEVs pay no federal or by Washington state drivers of different classes of vehicles. state gasoline taxes, and it is also true that PHEVs, such as the Chevrolet Volt, might pay proportionately very little in 4 For a breakdown of transportation funding sources at the federal, gasoline taxes. state and local levels, see http://www.transportation-finance.org/ funding_financing/funding/. Recommendation: Governments (federal, state, and local) 5 Calculations are based on the 11,489 miles per year driven, on average, by drivers residing in the greater Seattle metro area. should fully and fairly disclose all transportation-related tax-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 85 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

70 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 4-3 U.S. annual light-duty fuel consumption and VMT. NOTE: VMT, vehicle miles traveled; LDV, light-duty vehicle. SOURCE: DOE (2014b).

es and fees currently paid by all vehicles, including average road by 2015 were met with BEVs, the resulting unrealized passenger vehicles, alternative-fuel vehicles (such as com- revenue from motor fuel taxes would be less than 0.5 percent pressed natural gas), HEVs, PHEVs, and BEVs. Providing of the total projected revenue shortfall for the federal Highway that information to elected officials and the public will give Trust Fund (CAL PEV 2012b). them an accurate baseline against which policy discussions and choices can be made. Finding: For the next few years, the fiscal impact of not collecting a fuel tax from PEVs is negligible. Impacts on Transportation Budgets Fairness and Equity in Transportation Taxes As noted, the first policy question is whether, from a fiscal viewpoint, the lack of fuel taxes paid by PEVs is having The second policy question is whether PEV drivers who a negative effect on federal or state transportation budgets, pay little or no fuel taxes raise issues of fairness, given the either now or within the next 10 years. At the federal level, strong user-fee paradigm for funding the expenses of the estimates can be made of the unrealized fuel tax revenues highway infrastructure in the United States. Even though the from PEVs; the results are shown in Table 4-2. On the basis government would only derive an extremely small share of of the number of PEVs sold through 2013, an additional $14 revenue by taxing PEVs, the sentiment among elected offi- million annually could be generated for the federal Highway cials and the general public remains that PEV drivers should Trust Fund if each PEV was required to pay $96 per year, the be paying the fuel tax (or its equivalent) as their fair share same amount paid by a driver of a 22 mpg gasoline-powered for maintaining and improving the roadways on which they sedan. To put that amount in context, the federal Highway drive. Although its study did not focus on equity issues re- Trust Fund collects fuel-tax revenues of about $33 billion lated to taxation of PEVs, TRB (2011) did identify strongly each year (CBO 2013). held notions of fairness and equity that are inherent in the PEV industry analysts have also examined the impact of transportation tax system and that are important for public PEVs on transportation budgets. The California PEV Collab- policy making; they are summarized in Table 4-3. orative—a public-private consortium of governments, private The fairness issue in the tax treatment of PEVs appears businesses, vehicle manufacturers, and nongovernment orga- to be more acute at the state and local levels, where many nizations allied to promote PEVs—recently found that if the elected officials are actively considering fuel-tax increases Obama administration goal of putting 1 million PEVs on the to reduce the backlog of roadway maintenance and improve-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 86 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 71

FIGURE 4-4 Annual transportation-related taxes paid by Washington state drivers. SOURCE: WSDOT (2013).

ment projects. As noted by the TRB (2011, p. 103), for politi- considering the much larger impact other high-mileage ve- cians and other decision makers, one of the first hurdles to hicles will have on transportation funding levels, particularly overcome in embarking on a new transportation initiative— once the 2025 CAFE standards (54.5 mpg) take effect. The which will require financing, perhaps through an increase in fuel economy of the entire light-duty passenger vehicle fleet is the fuel tax—is to gain public support. Decision makers go increasing and will continue to increase in the coming decades to great lengths to ensure that the burdens (taxes) and bene- largely due to federal CAFE standards (see Figure 4-1). The fits (capital projects) are allocated in ways that are perceived Congressional Budget Office (CBO) estimated that the new as fair. It is in trying to rally public support for tax increases CAFE standards would gradually lower federal gasoline-tax that some politicians have sought to remedy the perceived revenues, eventually causing them to fall by 21 percent. The unfairness concerning unrealized revenue from PEV driv- CBO analysis demonstrated that from 2012 through 2022, ers (Vekshin 2013). Washington, Virginia, North Carolina, which is before the most stringent CAFE standards take ef- South Carolina, New Jersey, Indiana, Arizona, Michigan, fect, there will be a $57 billion drop in revenues (CBO 2012). Oregon, and Texas have all considered or, in some cases, In addition to federal consideration of the impacts of enacted legislation that imposes a fee or tax on PEVs. Many high-mileage vehicles, many states are now actively explor- of the efforts were undertaken as part of, or coincident with, ing potential solutions to the forecasted revenue shortfalls (see proposals to increase the fuel tax on all motorists. Figure 4-5). At least one state (Washington) has forecast the potential transportation-revenue shortfalls attributable to im- Finding: Perceptions of fairness and equity are important proving fuel economy and to alternative-fuel vehicles, such factors to consider in PEV tax policies, even though the ac- as PEVs, and found that the potential drop in revenues ranges tual revenue impact of PEV taxation is negligible in the short from 10 to 28 percent over the next 25 years (WSTC 2014). run and likely to remain minimal over the next decade. Both federal and state policy makers and the public are becoming increasingly aware of the impact that high-mile- Government Responses to Plug-in Electric age and alternative-fuel vehicles will have on roadway fund- Vehicles and High-Mileage Vehicles ing (Weissmann 2012). The Texas Transportation Institute recently convened several focus groups to better understand The third policy question raised is the extent to which public sentiment. Participants strongly preferred mileage PEVs should be a specific focus for new methods of taxation, fees for vehicles that might only pay state vehicle registration and title fees for their road use (GAO 2012b).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 87 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

72 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

TABLE 4-2 Comparison of Unrealized Revenue from Battery Electric Vehicles and Plug-in Hybrid Electric Vehicles Fuel Economy Annual Gallons Federal Gas Annual Unrealized Vehicle Type U.S. Total 2013a Average Annual VMT (MPG or MPGe) Consumed Tax Rate Revenue Avg. Sedanb — 11,489 22 522 gal $0.184 $96 per vehicled

BEV 72,028 11,489 — — $0.184 $6.9 million

PHEVc 95,589 11,489 98 117 gal $0.184 $7.1 million a Electric Drive Transportation Association Sales Dashboard, Totals from December 2010 to December 2013. b The data comprising the base case are adapted from GAO (2012b, p. 9). c Because PHEV models vary widely, the Chevrolet Volt was used as the reference case as it has the longest all-electric range of the PHEVs on the market. d This estimate is the baseline annual gasoline tax paid per vehicle, not annual unrealized revenue. NOTE: BEV, battery electric vehicle; MPG, miles per gallon; MPGe, miles per gallon gasoline equivalent; PHEV, plug-in hybrid electric vehicle; VMT, vehicle miles traveled.

TABLE 4-3 Types of Equity and Examples in the Transportation Tax System Type of Equity Simple Definition Transportation Example Benefits received I get what I pay for People who use a facility the most pay the most. Ability to pay I pay more because I have more money A project is financed through a progressive tax that is disproportionately paid by higher income people. Return to source We get back what we put in Transit investment in each county is matched to that county’s share of metropolitan tax revenues used for transit. Costs imposed I pay for the burden I impose on others Extra expense required to provide express bus service for suburb- to-city commuters is recovered by charging fares for this service. Process (or participation) I had a voice when the decision was made Public outreach regarding proposed new high-occupancy-toll lanes provides transparent information and seeks to involve all affected parties in public hearings and workshops. SOURCE: TRB (2011, p. 41).

Whether the concern is limited to PEVs or more broad- tests of mileage-based fees for PEVs; none reported that they ly centered on high-mileage vehicles, states are beginning would be opposed to such tests for PEVs. The Road Usage to take action. Several states have enacted special taxes on Fee Pilot Program Act of 2013 was introduced in the U.S. PEVs or are considering how to tax them. Other states are House of Representatives to authorize, fund, and partner exploring new transportation-tax methods to address not with states to conduct VMT pilot projects across the nation. only PEVs but all high-mileage vehicles (see Figure 4-6). Separate from the federal government efforts, over 20 Two congressionally chartered transportation funding states are actively studying, testing, or, in the case of Or- and financing commissions—the National Surface Transpor- egon, implementing some version of a mileage-based fee, tation Policy and Revenue Study Commission and the Na- also known as road usage charges or VMT fees or simply tional Surface Transportation Infrastructure Financing Com- taxes (D’Artagnan Consulting 2012). The fundamental con- mission—have independently called for a transition from cept is that drivers would be assessed a cents-per-mile tax the current fuel-tax system to a mileage-based fee system to for every mile that is driven within the taxing jurisdiction fund the nation’s highway infrastructure (NSTPRSC 2007, (region, state, or nation), regardless of the vehicle type, fuel pp. 51-54; NSTIFC 2009, p. 7). A recent report by the Gov- source, or engine technology. ernment Accountability Office (GAO) examined the feasibility of mileage fees and recommended a federally sponsored Recommendation: In jurisdictions that do impose special pilot program to evaluate the viability, costs, and benefits of taxes, fees, or surcharges on PEVs as a means of requiring mileage fee systems, particularly for commercial trucks and contribution to roadway upkeep, governments should ensure PEVs (GAO 2012b). GAO (2012b, p. 45) found that two- that such taxes are proportionate to actual usage, just as cur- thirds of state DOTs (34 of 51, including the District of Co- rent motor fuel taxes are proportionate to usage. lumbia) reported that they would support federally led field

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 88 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 73

FIGURE 4-5 Historic and forecast gasoline- tax revenue for Washington state, FY 1990 to FY 2040. SOURCE: WSTC (2014, p. 5).

Intervening Policy Considerations in the PEVs and most in need of technological breakthrough. The Taxation of Plug-in Electric Vehicles tax credit is also limited in the amount available per taxpayer ($7,500) and limited in duration (credit is phased out after The last policy issue examined is whether other inter- the manufacturer reaches vehicle sales of 200,000). vening policy considerations might trump the general trans- In contrast, there is no intentional or targeted tax in- portation-tax paradigm of user pays, at least until PEVs have centive to encourage PEVs to drive on public roadways.6 reached some level of market penetration. U.S. tax policy Instead, the government’s pro-PEV scheme consists of tax has a long and successful track record of encouraging in- credits, rebates, fee reductions, and exemptions for the pur- novation (Reuters 2013). There are many examples in the chase and ownership of the PEV—but not for its use of pub- current U.S. tax code (and state tax codes) where taxes are lic roadways. The fact that PEVs do not pay the fuel tax or exempted, credited, or rebated to promote the development a similar road usage tax stands apart from the vast majority or proliferation of services, assets, or activities deemed to of tax policies that are transparent, legislatively granted, and provide a public benefit, such as dependent-care tax benefits targeted in scope, quantity, or duration. and research and development or manufacturing tax credits. To the extent policy makers wish to continue providing That tax forbearance acts as the public’s investment in the PEV drivers with the financial benefit of not paying the fuel societal good produced. tax (or alternative road user charge), serious consideration Most tax incentives are limited in scope, duration, or should be given to explicitly and intentionally adopting such amount, so as to target more carefully the specific activity a policy in the same manner as other tax incentives. Although to be encouraged and to limit the public’s subsidization (or it might initially seem odd to enact a law or regulation that investment). The current federal $7,500 tax credit for PEVs specifically exempts an activity (PEV driving) that is already is a good example of a narrowly targeted federal subsidy 6 (IRS 2009). As currently enacted, the amount of the credit One could argue that allowing PEVs to drive in the high-occupancy-vehicle lane is an incentive to drive, as opposed to an incen- increases on the basis of the capacity of the PEV battery be- tive to own, and that the resulting loss of occupancy in the lane for cause the battery is the most expensive component unique to other vehicles represents a public “investment.”

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 89 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

74 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 4-6 PEV-specific measures for transportation funding. NOTE: PEV, plug-in electric vehicle. SOURCE: Based on data from C2ES (2015). Courtesy of the Center for Climate and Energy Solutions.

untaxed, it could be an effective strategy for addressing the pend on the availability of charging infrastructure, whether perceived issues around fairness and more clearly elaborat- at home, at work, or in public locations (see Chapter 5 for an ing the government’s innovation policy by setting criteria like in-depth discussion of charging infrastructure needs). those for other tax incentives found in the U.S. and state tax Electrical permit requirements appear to vary widely codes. from jurisdiction to jurisdiction (see Table 4-4), as does the amount of time required to apply for and process permits Recommendation: Federal and state governments should and to obtain a final electrical inspection to certify compli- adopt a PEV innovation policy where PEVs remain free from ance with applicable electrical codes. Consumer interest in special roadway or registration surcharges for a limited time to PEVs could be seriously impeded if PEV buyers must bear encourage their adoption. high permit and installation costs and experience delay in the activation of their home chargers. STREAMLINING CODES, Some forward-looking jurisdictions are making adjust- PERMITS, AND REGULATIONS ments in their electrical codes and permit processes to expedite installation and activation of a home-based charger.7 Fur- Although there are some applicable federal and state thermore, many jurisdictions are proactively amending their permitting processes that affect PEV infrastructure deploy- building codes to require that new construction be “forward ment—such as federal environmental laws (for example, the compatible” with devices for charging at home (DOE 2014c). National Environmental Policy Act, NEPA) and state regu- In its interim report, the committee suggested that state lations—cities, counties, and regional governments are at and local governments ensure that their permit processes are ground-zero for consumer adoption and use of PEVs. Travel appropriate for the type of infrastructure project and poten- distances, trip patterns, and vehicle registration data show that most PEV registrations and travel will be within urban- 7 Portland, Oregon; Raleigh, North Carolina; and San Francisco, California are three municipalities that have instituted programs to ized areas. The usefulness of the vehicles will largely de- expedite electrical permit processes.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 90 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 75 tial impact to the site or broader environment (NRC 2013). Recommendation: Local governments should streamline There are instances where extensive permit processes and permitting and adopt building codes that require new con- environmental review have been undertaken that would have struction to be capable of supporting future charging instal- been appropriate for a highway expansion project but are ill lations. Governments could implement new approaches, per- suited for the simple installation of a DC fast-charging sta- haps on a trial basis, to learn more about their effectiveness tion (C2ES 2012). For example, Oregon DOT has reported while still ensuring personal and environmental safety. that even though the DC fast-charging stations installed in Oregon were provided under a master contract by a single ANCILLARY INSTITUTIONAL ISSUES vendor, the environmental permit process for each station RELATED TO SUPPORT FOR differed based on the source of funding used to pay the con- PLUG-IN ELECTRIC VEHICLES tractor for otherwise identical stations (A. Horvat, Oregon Department of Transportation, personal communication, Battery Recycling and Disposal June 2014). If the charging station was funded with U.S. DOT money through the federal Transportation Investment PEV battery recycling and disposal needs will affect the Generating Economic Recovery (TIGER) grant program, costs and acceptance of PEVs and the infrastructure require- each station was required to undergo heightened NEPA per- ments to support them. At the end of its useful life in the mitting, including an assessment of potential underground vehicle, the battery must be disposed of, either by applying it hazardous materials. However, if the station was to be fund- to a secondary use (for example, as a back-up power source ed through DOE, there were no permit requirements beyond in a stationary application) or by reusing materials and com- those for ordinary state and local permits. ponents that have value and disposing of the remainder as waste. The cost of disposal, less any value in secondary use Finding: Regulatory and environmental officials often do or of recycled parts and materials, ultimately must be paid by not understand the nature, uses, and potential site impacts of the vehicle owner. Actions that reduce this cost will lower a charging stations. As a result, unnecessary permit burdens and cost barrier to PEV use. costs have been introduced to the installation process for pub- PEV manufacturers, waste disposal firms, and others lic charging stations. are working to create PEV battery recycling and disposal systems. If their efforts lag expansion of the PEV market, it Recommendation: Federal officials should examine current is conceivable that when significant numbers of PEVs begin NEPA and other permitting requirements to determine the to reach the end of their lives, a battery-disposal bottleneck most appropriate requirements for the class of infrastructure could present an obstacle to PEV production and sales. PEV to be installed; the federal government should adopt uniform and battery manufacturers have stated that lithium batteries rules that would apply to all charging installations of a simi- contain no toxic substances that would preclude their dis- lar asset class, regardless of the capital funding source used posal in the ordinary waste stream (Kelty 2008; Panasonic to pay for them. 2014). However, because reducing the environmental effects of motor vehicle transportation motivates public support of Finding: The permitting and approval processes for home- the PEV market and is attractive to many PEV purchasers, based and public charging installations need more clarity, PEV producers have an incentive to develop recycling and predictability, and speed. reuse options for the batteries.

TABLE 4-4 Variation in Residential Electric Permit Fees by City or State Permit Fee ($) Region Number of Permits Average Minimum Maximum Arizona 66 96.11 26.25 280.80 Los Angeles 109 83.99 45.70 218.76 San Diego 496 213.30 12.00 409.23 San Francisco 401 147.57 29.00 500.00 Tennessee 322 47.15 7.50 108.00 Oregon 316 40.98 12.84 355.04 Washington 497 78.27 27.70 317.25 SOURCE: ECOtality (2013).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 91 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

76 Overcoming Barriers to Deployment of Plug-in Electric Vehicles In the longer term, recycling of high-value materials or of Automotive-type Rechargeable Energy Storage Systems components could be important for restraining PEV battery (J2950) (SAE International 2014). costs. Although projections indicate that material shortages No federal or state laws yet require recycling of the bat- are unlikely to seriously constrain PEV battery production, teries contained in PEVs. California and New York require large-scale conversion of the fleet to PEVs probably would recycling of small rechargeable batteries. In New York, sell- increase consumption of certain materials, including lithium ers are required to receive used batteries of that type, and and cobalt, enough to raise prices significantly. Efficient re- battery manufacturers are required to develop plans for col- cycling would moderate material price increases (Gaines and lection and recycling. The California law requires sellers Nelson 2010). to accept used batteries (Gaines 2014). Those laws could The sections below describe the status of recycling tech- provide a pattern for future laws applying to PEV batteries. nology; the regulations and standards affecting recycling; The federal government regulates the transportation of bat- prospects for secondary uses of batteries; present involve- teries as hazardous materials (PRBA 2014), but the transport ment of vehicle and battery manufacturers, recycling firms, regulations appear to be aimed mainly at the risk of fire from and others; and possible areas for federal action. sparks or short circuits. European Union regulations have established require- Finding: Reducing the environmental impact of motor ve- ments for collection and recycling of all batteries sold to hicle transportation attracts buyer interest and public support consumers in the European Union. The manufacturer or dis- for PEVs. Therefore, although the disposal of lithium-ion tributor of the consumer product is responsible for compli- PEV batteries does not appear to present adverse health risks ance (European Commission 2014). nor does it have substantial financial advantages, provision for environmentally sound battery disposal will facilitate de- Finding: Industry standards regarding design and labeling of velopment of the PEV market. PEV batteries are necessary for efficient and safe recycling.

Recycling Technology Secondary Uses

Technologies available today for lithium-ion battery re- PEV battery performance (energy storage capacity) de- cycling recover certain elementary materials from the bat- clines with use until it becomes unacceptable for powering a tery structure and the cathode, such as cobalt and nickel. The vehicle. A battery in this condition, however, might still be us- lithium in the cathode is not recovered (ANL 2013; Gaines able for other applications, such as energy storage by utilities 2014). Most of the materials obtainable from recycling lithi- to satisfy peak demand, storage of energy from an intermittent um-ion batteries are of little value compared with the cost of generator like a solar energy facility, or as backup power in recovery, and newer battery designs that use less expensive a residence. Developing the market for such secondary uses materials (in particular, cathodes that do not contain cobalt) would reduce the cost of the battery to its initial owner, the yield even less value in recycling. Therefore, recycling is not PEV purchaser. Reuse delays but does not eliminate the need economical (Kumar 2011; Gaines 2012). Processes under for eventual recycling or disposal of the battery. development seek to recover intact, reusable cathode materi- It is most helpful to view battery secondary use (B2U) als that have more value than their elemental components as an economic ecosystem—a collection of independent (ANL 2013). stakeholders that could co-evolve around a value chain to bring depleted batteries from the PEV into a secondary sys- Standards and Regulations tem. The maximum potential and limitations of the B2U ecosystem are set by the original design and architecture of Battery standards are essential for efficient and safe dis- the vehicle-battery system. Because the vehicle manufactur- posal and recycling. Designing batteries with recycling in ers specify the design for the vehicle-battery pack and the mind reduces the cost of recycling, and standardization of parameters for its production, they are currently the most designs simplifies the operation of recycling facilities. La- critical player in the development of such an ecosystem. To beling is necessary to ensure that batteries of different com- enable a B2U market to evolve, the vehicle manufacturers position can be properly sorted for recycling. Design stan- must find enough value from participating in the B2U eco- dards also could facilitate secondary uses. system to develop a strategy that complements their propri- The Society of Automotive Engineers (SAE) is actively etary PEV technologies. engaged in vehicle electrification standards. Standards under A B2U strategy must consider the design, development, development related to battery disposal include Vehicle Bat- and manufacture of a battery system with the intent to serve tery Labeling Guidelines (J2936), Identification of Transpor- two purposes: (1) the initial use in the vehicle and (2) another tation Battery Systems for Recycling Recommended Practice application, most likely stationary. An optimal B2U strategy (J2984), Standards for Battery Secondary Use (J3097), and requires the design and use of the battery to maximize the Recommended Practices for Transportation and Handling value of the system over its entire extended life cycle. Bowl-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 92 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 77 er (2014) developed a model to evaluate trade-offs along the increases among vehicle batteries, charging systems, and the secondary use value chain. The modeling showed that cir- national electric grid. cumstances can exist in which the economic incentives for secondary use become attractive, but this can only be accom- Finding: Vehicle manufacturers appear to recognize a prac- plished with the active participation of all the stakeholders in tical responsibility for disposal of batteries from their ve- the B2U value chain. hicles, although their willingness to bear this responsibility Each vehicle manufacturer could independently develop voluntarily as PEV sales grow and the fleet ages remains to and use such a model to integrate its own technical param- be seen. Unlike the European Union, the United States im- eters into the development of a proprietary B2U strategy. poses no legal requirements for battery disposal on manufac- Current evidence suggests that the market will begin with turers or sellers. such proprietary deployments. For example, Nissan was first to announce the use of an on-vehicle battery to supplement Finding: There is a potential market for secondary uses of electric energy to a demonstration home near its headquar- PEV batteries that are no longer suitable for automotive use ters (Pentland 2011). The removal of a depleted PEV bat- but retain a large share of their storage capacity. Whether led tery that had been optimized for stationary use would seem a by private companies or public agencies, an effective collabo- logical next step. Ford, Tesla, and Toyota have been reported ration among the entities that design and manufacture PEVs, as pursuing various strategies (Woody 2014). the vehicle owners, and the users and purveyors of stationary PEV manufacturers are engaged in developing technol- electric systems can materially assist the development of an ogy and exploring the market for stationary battery applica- economically efficient secondary-use marketplace. tions. Most such efforts are in early stages and include the following examples: Recycling Arrangements and Capabilities

• Nissan Motor Company and Sumitomo Corporation The principal participants in the PEV battery recycling have formed a joint venture (4R Energy Corporation) system will be the vehicle owner, the party that accepts or is to store energy from solar generators and other appli- required to accept the responsibility for battery disposal (most cations using PEV batteries (Srebnik 2012; 4R Energy likely the vehicle manufacturer), companies in the recycling 2013; Sumitomo 2014). Sumitomo announced installa- industry, and producers and purchasers of stationary storage tion of a prototype system assembled from 16 used PEV units that can reuse PEV batteries. At present, most PEV bat- batteries at a solar farm in Japan in February 2014. A teries that have gone out of use probably have passed through battery system has been installed in an apartment build- PEV dealerships, and manufacturers appear to recognize that ing in Tokyo (Nissan Motor Corporation 2013). The they will be expected to provide for battery disposal. venture is working on developing additional applica- Lead-acid battery recycling is well established in the tions for used batteries. United States and internationally and is sustained by the • Tesla Motors is supplying batteries to SolarCity, a com- value of the recycled lead (that is, recyclers pay for the used pany that leases and installs solar panels for residential batteries they process). Nearly all lead-acid batteries are re- and business customers. The battery is a component of cycled. The established firms with experience in recycling the solar panel system. Trial residential systems were technology and in the logistics of battery collection, trans- installed in 2013 (Woody 2013). The system is not report, and handling can provide the industrial base for PEV ported to be reusing PEV batteries but represents a po- battery recycling (Gaines 2014). The U.S. battery recycling tential market for reuse. firm Retriev Technologies (until 2013 known as Toxco In- • A Toyota subsidiary (Toyota Turbine) has begun reus- dustries) recycles lithium-ion PEV batteries (Retriev Tech- ing Toyota HEV NiMH batteries in solar panel energy nologies 2014). Retriev and the U.K. battery recycling firm management systems that have been sold to Toyota ve- Ecobat Technologies are reported to be developing processes hicle dealerships (Toyota Turbine 2013; Nikkei Asian for recovery of intact cathode materials from PEV batteries Review 2014). (ANL 2012), a process that has potential for reducing the net • General Motors and ABB in 2012 demonstrated a sys- cost of battery production and disposal. The Belgian materi- tem that packaged five used Chevrolet Volt batteries in als and recycling firm Umicore has established a facility in a stationary back-up power unit for residential or busi- North Carolina to dismantle PEV and HEV batteries before ness applications (General Motors 2012). shipment of components to its processing plant in Belgium (Umicore 2014). Alternatively, the federal government could develop a Vehicle manufacturers have arrangements with recy- common public framework that would disseminate informa- clers for battery disposal and have had some involvement tion on the actions and processes that create second-use val- in developing improved processes. For example, Tesla has ue to the potential participants in a national B2U value chain. arrangements with recycling companies in Europe and North That approach might become appropriate as standardization America for recycling and disposal of used battery packs

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 93 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

78 Overcoming Barriers to Deployment of Plug-in Electric Vehicles (Kelty 2011) and plans to recycle batteries in-house at what The training program is funded by a grant from DOE, as it calls the Gigafactory, a battery plant that it intends to build part of the department’s effort to promote PEV use (NFPA (Tesla 2014). 2014). The NFPA project has developed a variety of training materials and programs and information resources and has Finding: The solid waste disposal industry has developed conducted a series of courses to train instructors. The NFPA technologies for acceptable disposal of PEV batteries, and training program is supported by research, involving full- technological improvements might succeed in extracting scale testing, to determine best practices for response to in- greater net value from recycled materials. However, PEV cidents involving PEVs. The research has been supported by battery recycling will not pay for itself from the value of DOE, the National Highway Traffic Safety Administration recycled materials. (NHTSA), and the automotive industry (Long et al. 2013). At the federal level, NHTSA develops and distributes Finding: Battery disposal is not a near-term obstacle to PEV EMS training standards and curricula, organizes coopera- deployment; PEV batteries can be safely disposed of in the tive activities, maintains databases, and evaluates state EMS general waste stream, and regulating battery disposal at this systems (NHTSA 2014a). NHTSA has published guidance time could increase the cost of PEV ownership. Thus, federal on safety precautions for vehicle occupants, emergency regulatory action does not appear necessary at this time. responders, and towing and repair workers when a PEV is damaged by a collision (NHTSA 2012; NHTSA 2014b). The Finding: PEV manufacturers, the solid waste industry, and guidance is brief and general and does not contain detailed standards organizations are working to develop disposal, re- technical information or response instructions. cycling, and reuse technologies. Although federal action is not required, there appear to be opportunities for federal sup- Recommendation: DOE and NHTSA should cooperate in port of industry efforts. long-term monitoring of the implementation and effectiveness of the NFPA EV Safety Training program. The monitoring Recommendation: Although battery recycling does not should determine whether the program is reaching local emer- present a barrier to PEVs in the near term, the federal gov- gency responders, whether the skills it teaches prove useful in ernment should monitor the developments in this area and practice, and whether it is timely. be prepared to engage in research to establish the following: efficient recycling technologies, standards for battery design REFERENCES and labeling that will facilitate safe handling of used batteries and efficient recycling, and regulation to ensure safe ANL (Argonne National Laboratory). 2012. “Advanced transportation and environmentally acceptable disposal of Battery Research, Development, and Testing.” http:// batteries that promotes efficient recycling and avoids creat- www.transportation.anl.gov/batteries/index.html. ing unintended obstacles. Accessed March 14, 2013. ANL. 2013. “PHEV Batteries.” http://www.transportation. Emergency Response anl.gov/batteries/phev_batteries.html. Battaglia, S. 2013. Are You Willing to Pay an Extra Tax to Police, firefighters, and emergency medical services Drive Your Electric Vehicle? Your Energy Blog, Febru- (EMS) personnel responding to road crashes that involve ary 28. http://www.yourenergyblog.com/are-you-will- PEVs must be aware of the hazards associated with PEVs ing-to-pay-an-extra-tax-to-drive-your-electric-vehicle/. that differ from the hazards associated with gasoline-powered Accessed March 1, 2015. vehicles in wrecks, and they must be trained in procedures Bowler, M. 2014. “Battery Second Use: A Framework for for mitigating these hazards. The hazards are risks of electri- Evaluating the Combination of Two Value Chains.” Un- cal shock, fire, and exposure to toxic substances (NHTSA published Doctoral Dissertation, Clemson University. 2012, p. 2). Because highway emergency response in the CAL PEV (California Plug-in Electric Vehicle Collaborative). United States is the responsibility of thousands of indepen- 2012a. “A Toolkit for Community Plug-in Electric Ve- dent local police, fire, and EMS organizations, training and hicle Readiness.” August. http://www.pevcollaborative. communication of information are challenging activities. All org/sites/all/themes/pev/files/docs/toolkit_final_web- the emergency responders will require training and access to site.pdf. the necessary equipment to discharge batteries safely after an CAL PEV. 2012b. “Transportation Funding Consensus State- accident and on other safe handling procedures. ment.” September. http://www.evcollaborative.org/sites/ The most important nationwide PEV emergency re- all/themes/pev/files/TransportationFunding_PEVC_ sponse training activity is Electric Vehicle Safety Training, a ConsensusStatement_120924_Final.pdf. project of the National Fire Protection Association (NFPA). CBO (Congressional Budget Office). 2012. “How Would NFPA is a nonprofit membership organization engaged in Proposed Fuel Economy Standards Affect the Highway development of codes and standards, training, and research.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 94 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 79 Trust Fund?” May. http://www.cbo.gov/sites/default/fil EPA (U.S. Environmental Protection Agency). 2013. “Light- es/cbofiles/attachments/05-02-CAFE_brief.pdf. Duty Automotive Technology and Fuel Economy Trends: CBO. 2013. “Testimony on the Status of the Highway Trust 1975 Through 2013.” http://www.epa.gov/otaq/fetrends. Fund before the Subcommittee on Highways and Tran- htm. Accessed March 1, 2015. sit, Committee on Transportation and Infrastructure, European Commission. 2014. “Frequently Asked Questions and U.S. House of Representatives.” Washington, DC. on Directive 2006/66/EU on Batteries and Accumula- July 23. http://www.cbo.gov/sites/default/files/cbofiles/ tors and Waste Batteries and Accumulators.” http://ec. attachments/44434-HighwayTrustFund_Testimony.pdf. europa.eu/environment/waste/batteries/pdf/faq.pdf. C2ES (Center for Climate and Energy Solutions). 2012. FHWA (Federal Highway Administration). 1997. “Highway “Peer Exchange for State DOTs on Electric Vehicles— Finance.” https://www.fhwa.dot.gov/ohim/hs97/hfpage. Charging Infrastructure Workshop.” Raleigh, NC, June htm. 25-26. http://www.c2es.org/initiatives/pev/implement- FHWA. 2014. “Highway Statistics, 2012. Table HF-10.” Of- ing-action-plan/state-dots-electric-vehicles#ChargingIn fice of Highway Policy Information: Highway Statistics frastructureWorkshop. Series, March. http://www.fhwa.dot.gov/policyinforma- C2ES. 2014. “PEV Action Tool.” http://www.c2es.org/pev- tion/statistics/2012/hf10.cfm. action-tool. Accessed October 23, 2014. 4R Energy Corporation. 2013. “Business Model.” http://www. C2ES. 2015. “PEV-Specific Measures for Transportation 4r-energy.com/en/services/. Accessed December 3, 2013. Funding.” Accessed March 1, 2015. http://www.c2es. Gaines, L. 2012. To recycle, or not to recycle, that is the org/initiatives/pev/maps/pev-specific-measures-infra- question: Insights from life-cycle analysis. MRS Bul- structure-funding. letin 37(4): 333-338. D’Artagnan Consulting. 2012. “Washington State Road Us- Gaines, L. 2014. “The Future of Automobile Battery Recy- age Charge Assessment: Domestic and International Re- cling.” Presentation to the Committee on Overcoming view and Policy Context.” Steering Committee Brief- Barriers to Electric-Vehicle Deployment, Washington, ing, September 13. http://waroadusagecharge.files.word DC, February 25. press.com/2012/08/international-domestic-review_fi- Gaines, L., and P. Nelson. 2010. “Lithium-ion Batteries: Ex- nal.pdf. amining Material Demand and Recycling Issues.” http:// DOE (U.S. Department of Energy). 2008. “Linkages of www.transportation.anl.gov/pdfs/B/626.PDF. DOE’s Energy Storage R&D to Batteries and Ultracapac- GAO (U.S. Government Accountability Office). 2011. High- itors for Hybrid, Plug-in Hybrid, and Electric Vehicles.” way Trust Fund: All States Received More Funding http://www1.eere.energy.gov/ba/pba/pdfs/vehicle_ Than They Contributed in Highway Taxes from 2005 to energy_storage_r_and_d_linkages.pdf. 2009. GAO-11-918. Washington, DC, September. http:// DOE. 2013a. 2013 Annual Merit Review Results Report. www.gao.gov/assets/520/511454.pdf. http://energy.gov/sites/prod/files/2014/04/f14/2013_ GAO. 2012a. Batteries and Energy Storage: Federal Initia- amr_02.pdf. tives Supported Similar Technologies and Goals but DOE. 2013b. “Energy Innovation Hub: Batteries and Energy Had Key Differences. GAO-12-842. Washington, DC, Storage.” Joint Center for Energy Storage Research. August. http://www.gao.gov/assets/650/647742.pdf. http://science.energy.gov/~/media/bes/pdf/hubs/JCESR GAO. 2012b. Highway Trust Fund: Pilot Program Could _Fact_Sheet.pdf. Help Determine Viability of Mileage Fees for Certain Ve- DOE. 2013c. “Maps and Data.” Alternative Fuels Data Cen- hicles. GAO 13-77. Washington, DC, December. http:// ter. http://www.afdc.energy.gov/data/categories/fuel-co www.gao.gov/assets/660/650863.pdf. nsumption-and-efficiency#10562. General Motors Corporation. 2012. “GM, ABB Demonstrate DOE. 2014a. Fiscal Year 2013 Annual Progress Report for Chevrolet Volt Battery Reuse Unit.” November. http:// Energy Storage R&D. Publication No. DOE/EE-10-38. media.gm.com/media/us/en/gm/news.detail.html/con- Washington, DC. http://energy.gov/eere/vehicles/down- tent/Pages/news/us/en/2012/Nov/electrification/1114_ loads/vehicle-technologies-office-2013-energy-storage- reuse.html. rd-progress-report-sections. INL (Idaho National Laboratory). 2014a. “The EV Project.” DOE. 2014b. “Electric Vehicle Community Readiness Proj- http://avt.inel.gov/evproject.shtml. Accessed October 23, ects.” http://www1.eere.energy.gov/cleancities/electric_ 2014. vehicle_projects.html. INL. 2014b. Observations from The EV Project in Q4 2013. DOE. 2014c. “State Laws and Incentives.” Alternative Fu- http://avt.inel.gov/pdf/EVProj/EVProjectSummaryRe- els Data Center. http://www.afdc.energy.gov/laws/sear portQ42013FINAL.pdf. ch?utf8=%E2%9C%93&keyword=building+code&sea IRS (Internal Revenue Service). 2009. “Notice 2009-89, New rch_by=keyword. Qualified Plug-in Electric Drive Motor Vehicle Credit.” In- ECOtality. 2013. “The EV Project.” http://www.theevproj ternal Revenue Bulletin 2009-48. http://www.irs.gov/irb/ ect.com/overview.php. Accessed March 1, 2015. 2009-48_IRB/ar09.html.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 95 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

80 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Kelty, K. 2008. “Mythbusters Part 3: Recycling Our Non- NSTPRSC (National Surface Transportation Policy and Toxic Battery Packs.” Tesla Blog, March 11. http://www. Revenue Study Commission). 2007. Report of the Na- teslamotors.com/blog/mythbusters-part-3-recycling- tional Surface Transportation Policy and Revenue Study our-non-toxic-battery-packs. Commission: Transportation for Tomorrow. December. Kelty, K. 2011. Tesla’s Closed-Loop Battery Recycling Pro- http://transportationfortomorrow.com/final_report/. gram. Tesla Blog, January 26. http://www.teslamotors.com/ ODOT (Oregon Department of Transportation). 2007. “Fi- blog/teslas-closed-loop-battery-recycling-program. nancial Services Update: Fuels Tax Revenue.” ODOT Kumar, A. 2011. The lithium battery recycling challenge. Financial Services, February. http://www.oregon.gov/ Waste Management World 12(4). http://www.waste-man- ODOT/CS/FS/news/vo2-no2_fsupdate.pdf. agement-world.com/articles/print/volume-12/issue-4/ Panasonic. 2014. “Panasonic Batteries.” Product Information features/the-lithium-battery-recycling-challenge.html. Sheet. http://na.industrial.panasonic.com/products/batte Long, R., A. Blum, T. Bress, and B. Cotts. 2013. Best Prac- ries/rechargeable-batteries/lithium-ion. Accessed March tices for Emergency Response to Incidents Involving 1, 2015. Electric Vehicles Battery Hazards: A Report on Full- Pentland, W. 2011. “Nissan Leaf to Power Homes.” Forbes, Scale Testing Results. Fire Protection Research Founda- August 5. http://www.forbes.com/sites/williampentland/ tion, June. http://www.nfpa.org/research/fire-protection- 2011/08/05/nissan-leaf-to-power-homes/?partner research-foundation/reports-and-proceedings/electrical- =yahootix. safety/new-technologies-and-electrical-safety/emer- PRBA (Portable Rechargeable Battery Association). 2014. gency-response-to-incident-involving-electric-vehicle- Information on Batteries in Transport: Packaging and battery-hazards. Shipping Batteries Safely. http://www.prba.org/publica- Michigan Clean Energy Coalition. 2011. “Plug-in Ready Mich- tions/batteries-in-transport-shipping-batteries-safely/. igan: An Electric Vehicle Preparedness Plan.” http://cec- Rall, J. 2013. Sustainable State Transportation Funding and mi.org/wp-content/uploads/2011/11/Plug-In-Ready- Spending. National Conference of State Legislatures. Michigan.pdf. NCSL NRI Committee 2014 Spring Webinar Series, NFPA (National Fire Protection Association). 2014. “About April 24. http://www.ncsl.org/documents/transportation/ the Project.” Electric Vehicle Safety Training. http://www. Jamie.pdf. evsafetytraining.org/about-us.aspx. Retriev Technologies. 2014. “Lithium Ion.” http://www.re- NHTSA (National Highway Traffic Safety Administration). trievtech.com/recycling/lithium-ion. Accessed March 1, 2012. Interim Guidance for Electric and Hybrid-Electric 2015. Vehicles Equipped with High Voltage Batteries. DOT Reuters. 2013. “Thomson Reuters Names the World’s Top 100 HS 811 574. January. Most Innovative Organizations for 2013.” Press Release, NHTSA. 2014a. “About NHTSA EMS.” http://www.ems. October 7. http://thomsonreuters.com/press-releases/102 gov/mission.htm. Accessed October 23, 2014. 013/thomson-reuters-2013-top-100-global-innovators. NHTSA. 2014b. “Interim Guidance for Electric and Hybrid- Road Usage Fee Pilot Program. 2013. H.R. 3638, 113th Con- Electric Vehicles Equipped with High Voltage Batteries: gress. https://www.congress.gov/bill/113th-congress/hou Version 2.” DOT HS 811 576. Washington, DC. March. se-bill/3638. Nikkei Asian Review. 2014. “Toyota Spruces Up Green Rocky Mountain Institute. 2014. “Project Get Ready.” http:// Credentials with Battery Recycling.” January 28. http:// www.rmi.org/project_get_ready. Accessed October 23, asia.nikkei.com/Business/Companies/Toyota-raising- 2014. green-profile-by-spearheading-recycling-of-used-HV- SAE International. 2014. Vehicle Electrification Standards. batteries. http://www.sae.org/standardsdev/vehicleelectrification. Nissan Motor Corporation. 2013. “The Compact Lithium-ion htm. Battery.” Environmental Activities. http://www.nissan- Smart, J., and S. White. 2014. “Electric Vehicle Charging global.com/EN/ENVIRONMENT/CAR/FUEL_BAT- Infrastructure Usage Observed in Large-Scale Charg- TERY/DEVELOPMENT/LITHIUM_ION_BAT- ing Infrastructure Demonstrations.” Presentation to the TERY/. Accessed February 25, 2014. Committee on Overcoming Barriers to Electric-Vehicle NRC (National Research Council). 2013. Overcoming Bar- Deployment, Irvine, CA, February 25. riers to Electric-Vehicle Deployment: Interim Report. Srebnik, K. 2012. “Nissan Leaf Innovation and Strategy.” Washington, DC: The National Academies Press. http://chargedsv.org/wp-content/uploads/2012/08/Sreb- NSTIFC (National Surface Transportation Infrastructure Fi- nik_Charged2012SV_08232012.pdf. nancing Commission). 2009. Paying Our Way: A New Sumitomo Corporation. 2014. “World’s First Large-Scale Framework for Transportation Finance. February. http:// Power Storage System Made from Reused EV Batteries financecommission.dot.gov/Documents/NSTIF_Com- Completed in Japan.” February 7. http://www.sumitomo mission_Final_Report_Mar09FNL.pdf. corp.co.jp/english/news/detail/id=27673.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 96 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Government Support for Deployment of Plug-in Electric Vehicles 81 Tesla. 2014. “Gigafactory.” Tesla Blog, February 26. http:// Weissmann, J. 2012. “Why Your Prius Will Bankrupt Our www.teslamotors.com/blog/gigafactory. Highways.” Atlantic, February 2. http://www.theatlantic. Toyota Turbine and Systems Inc. 2013. “Enrich Lifestyles by com/business/archive/2012/02/why-your-prius-will- Using Energy Efficiently for the Better Future of the Envi- bankrupt-our-highways/252397/. ronment.” http://www.toyota-global.com/company/profi Woody, T. 2013. “How Tesla Batteries Are Powering an En- le/non_automotive_business/new_business_enterpris- ergy Revolution.” Atlantic, December 5. http://www.the es/pdf/toyota_turbine_and_systems_e.pdf. atlantic.com/technology/archive/2013/12/how-tesla- TRB (Transportation Research Board). 2006. The Fuel Tax batteries-are-powering-an-energy-revolution/282056/. and Alternatives for Transportation Funding: Special Woody, T. 2014. “Car Companies Take Expertise in Battery Report 285. Washington, DC: The National Academies Power Beyond the Garage.” New York Times, March 25. Press. WSDOT (Washington State Department of Transportation). TRB. 2011. Equity of Evolving Transportation Finance Mech- 2013. “EV Tax and Fee Comparison.” Presentation at anisms: Special Report 303. Washington, DC: The Na- EV Roadmap Six Conference, Portland, OR, July 30. tional Academies Press. WSTC (Washington State Transportation Commission). Umicore. 2014. “Supply: Hybrid Electric Vehicles/Electric 2014. Washington State Road Usage Charge Assessment: Vehicles.” http://www.batteryrecycling.umicore.com/su Business Case Evaluation Final Report. http://waroadus pply/HEV_EV/. Accessed October 23, 2014. agecharge.files.wordpress.com/2014/01/wa-ruc-busi- Vekshin, A. 2013. “Electric-Car Owners Get Taxed for Not ness-case-evaluation_01-07-14.pdf. Paying Gas Taxes.” Bloomberg BusinessWeek, June 6. http://www.businessweek.com/articles/2013-06-06/elec tric-car-owners-get-taxed-for-not-paying-gas-taxes.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 97 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles

The deployment of plug-in electric vehicles (PEVs) and have an incentive to build each category of charging infra- the fraction of vehicle miles traveled that are fueled by elec- structure, with particular attention to how infrastructure in- tricity (eVMT) depend critically on charging infrastructure. vestments would be recovered. The committee provides its PEV charging infrastructure (described in Chapter 2) is fun- findings and recommendations throughout this chapter. damentally different from the well-developed infrastructure In this chapter, the committee’s analysis of infrastructure for gasoline fueling. It can be found in a variety of loca- deployment assumes (1) no disruptive changes to current PEV tions, from a PEV owner’s home to a workplace to parking performance and only gradual improvements in battery capac- lots of restaurants, malls, and airports. A variety of charg- ity over time, (2) early majority buyers who do not plan to ing options are available, from AC level 1 chargers that use make changes to their lifestyles to acquire a PEV, (3) elec- 120 V ac electric circuits that are present in almost every tricity costs that are significantly less expensive than those building to DC fast chargers that do not yet have a technol- of gasoline per mile of travel, and (4) a cost for public and ogy standard. The charging rate also varies from slow (time- workplace charging that is at least as high as that for home insensitive) charging to fast (time-sensitive) charging. Each charging. The committee notes that the need for charging in- infrastructure category also has different upfront and ongo- frastructure could conceivably be mitigated by investments in ing investment costs and returns and different entities that battery swapping stations, which use robotic processes and al- would have an incentive to build such infrastructure, rang- low drivers to swap batteries in less than 3 minutes. The first ing from vehicle owners who might spend about $1,000 to major initiative for battery swapping services was launched upgrade their home outlet or electric panel to corporations by Better Place, which built networks of stations in Israel and and governments that could spend $100,000 to build a DC Denmark but declared bankruptcy in May 2013. Tesla has an- fast-charging station. The public charging stations might nounced a plan to add battery-swap technology at its network also require technology to monitor usage and bill customers. of fast-charging stations (Vance 2013). However, this model PEV deployment and eVMT will be constrained if charging is not widely available at this time and is not discussed further infrastructure is not conveniently located or if the available in this report. infrastructure does not facilitate charging within a convenient time frame. Thus, critical questions for vehicle manu- CHARGING INFRASTRUCTURE AND EFFECTS ON facturers and policy makers are how are vehicle deployment DEPLOYMENT OF PLUG-IN ELECTRIC VEHICLES and eVMT affected by the availability of various charging AND ON ELECTRIC VEHICLE MILES TRAVELED infrastructure types and what is the cost effectiveness of infrastructure investments relative to other investments that As discussed in Chapter 2, today’s charging infrastructure manufacturers and the government could make to overcome technology consists of AC level 1 and AC level 2 chargers, barriers to PEV deployment. which are typically used when charging time is not a prime This chapter considers scenarios for deploying PEV consideration, and DC fast chargers, which are typically used charging infrastructure and the potential effect of that infra- when charging time is an important consideration. All PEVs structure on PEV deployment and eVMT. The committee has can charge with AC level 1 and level 2 chargers, and most categorized infrastructure by location (home, workplace, in- battery electric vehicles (BEVs) can also charge at DC fast tracity, intercity, and interstate) and power (AC level 1, AC chargers. In the future, some plug-in hybrid electric vehicles level 2, and DC fast). The infrastructure categories are ranked (PHEVs) might be equipped to use DC fast chargers, but there in order of importance for increasing PEV deployment and is little motivation to make such a change because PHEVs can eVMT from the perspective of owners of the four PEV classes use their internal-combustion engines (ICEs) to circumvent as defined in Chapter 2. The experience and needs of current the need to charge. Charging infrastructure locations and in- early adopters were considered by the committee, but deploy- vestments range widely from an existing extensive network ment scenarios are focused on mainstream PEV deployment. of private chargers (or simply ordinary outlets) at homes and The chapter concludes by considering which entities might workplaces to an expanding infrastructure of public chargers, 82

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 98 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 83 such as those at retailers or shopping malls or along highways. ing infrastructure. That does not mean that electric-charging Workplace and public charging infrastructure might require infrastructure is not important for PHEV deployment, how- payment for electricity or time occupying the charger or be ever. PHEV drivers might still heavily use charging at pri- restricted to vehicles belonging to a subscription plan or to a vate and public locations to maximize their value proposi- certain vehicle manufacturer. tion in terms of cheaper charging, convenience, or personal In the mature market, the ideal number, location, and type values, such as environmental concerns. For example, data of charging infrastructure will depend on the demand for dif- from the EV Project on early adopters of the Chevrolet Volt ferent types of PEVs, their use, and their geographic distribu- show that 14 percent of charging events occurred away from tion. Conversely, although there has been little research on the home, which is similar to the percentage of charging away relationship between charging-station deployment and PEV from home (16 percent) for Nissan Leaf drivers (ECOtality deployment, the availability of charging infrastructure and the 2013; Smart and Schey 2012). Each charging-infrastructure rate of its deployment might itself influence PEV deployment category and the impact of each category on different PEV and use. Figure 5-1 shows six categories of charging-infra- classes are discussed in detail in the sections below. structure deployment, ranked in a pyramid that reflects their relative importance as assessed by the committee. As noted Home Charging above, the categories are defined by location and power. The term intercity refers to travel over distances less than twice the Home charging is a virtual necessity for mainstream range of limited-range BEVs, and interstate refers to travel PEV buyers of all four vehicle classes given that the vehicle over longer distances. is typically parked at a residence for the longest portion of Table 5-1 provides the committee’s assessment of the the day. As shown in Figure 5-2, the U.S. vehicle fleet spends effect of charging infrastructure on different PEV classes. about 80 percent of its time parked at home, and more than Evaluating infrastructure by type of PEV might help to ad- 50 percent of the U.S. vehicle fleet is parked at home even dress misconceptions about charging infrastructure needs. during weekday work hours. Most early adopters of PEVs For example, PHEVs do not require electric charging for have satisfied their charging needs primarily by plugging range extension because drivers have the option of fueling their vehicles into 120 V (AC level 1) or 240 V (AC level 2) with gasoline. BEVs, which have only electricity as a fuel receptacles at home during overnight hours or other periods option, are much more affected by the availability of charg- when it is convenient to leave their vehicles idle. Even the

FIGURE 5-1 PEV charging infrastructure categories, ranked by their likely importance to PEV deployment, with the most important, home charging, on the bottom, and the least important, interstate DC fast charging, at the top. NOTE: AC, alternating current; DC, direct current.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 99 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

84 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 5-2 Vehicle locations throughout the week on the basis of data from the 2001 National Household Travel Survey. SOURCE: Tate and Savagian (2009). Reprinted with permission from SAE paper 2009-01-1311 Copyright © 2009 SAE International.

large 85 kWh battery in a Model S can be fully charged over- from the U.S. Census and the U.S. Department of Energy night with the 10 kW AC level 2 charger recommended by (DOE) Residential Energy Consumption Survey to estimate Tesla for home use. A full battery charge will not usually be the potential for residential charging of PEVs using various needed each night because such charging will typically re- assumptions about missing data on, for example, the pres- place only the electricity used for the previous day’s driving. ence and size of driveways, the usability of electric outlets, For typical daily trip distances, only a few hours of charging and the number of parking spaces actually available for park- will be required for all types of PEVs. ing. Although 79 percent of U.S. households have dedicated Home charging is a paradigm shift in refueling behavior off-street parking, many households have multiple vehicles, for drivers accustomed to refueling quickly at gasoline sta- and under base-case assumptions, only 56 percent of vehi- tions. Many find home charging more convenient than refu- cles have dedicated off-street parking, and only 47 percent eling at public stations. For example, in the EV Project study, at an owned residence. Additionally, although 38 percent of about 85 percent of Volt charging events and 80 percent of all U.S. households are estimated to have charging access Leaf charging events occurred at home (Smart 2014a). for at least some vehicles, only an estimated 22 percent of all Home-charging infrastructure is not a barrier to PEV U.S. vehicles have a dedicated home parking space within deployment for households with a dedicated parking spot reach of an outlet sufficient to recharge a small PEV battery with an electric outlet nearby. According to the U.S. Cen- overnight. sus Bureau (2011a), nearly two-thirds of U.S. housing struc- Given the number of households with access to dedi- tures have garages or carports.1 Similarly, a representative cated parking with an outlet, PEVs could become a much telephone survey of 1,004 U.S. adults found that 84 percent larger share of the U.S. vehicle market while still relying on of respondents had dedicated off-street parking and 52 per- ubiquitous residential circuits to accommodate most charg- cent of respondents had a garage or dedicated parking spot ing needs. Given the large number of households that do not with access to an outlet (Consumers Union and the Union yet drive PEVs and could take advantage of the convenience of Concerned Scientists 2013). Traut et al. (2013) used data of charging at home, the scenario that seems most likely to emerge over the next decade is one in which the growth of 1 Some of the structures accommodate multiple households. demand for PEVs comes primarily from households who

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 100 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 85

TABLE 5-1 Effect of Charging-Infrastructure Categories on Mainstream PEV Owners by PEV Classa Infrastructure Category PEV Class Effect of Infrastructure on Mainstream PEV Owners Interstate Long-range BEV Range extension, expands market DC fast charge Limited-range BEV Not practical for long trips Range-extended PHEV NA – not equipped Minimal PHEV NA – not equipped Intercity Long-range BEV Range extension, expands market DC fast chargeb Limited-range BEV 2 × Range extension, increases confidence Range-extended PHEV NA – not equipped Minimal PHEV NA – not equipped Intracity Long-range BEV Not necessary DC fast chargeb Limited-range BEV Range extension, increases confidence Range-extended PHEV NA – not equipped Minimal PHEV NA – not equipped Intracity Long-range BEV Not necessary AC levels 1 and 2b Limited-range BEV Range extension, increases confidence Range-extended PHEV Increases eVMT and value proposition Minimal PHEV Increases eVMT and value proposition Workplace Long-range BEV Range extension, expands market Limited-range BEV Range extension, expands market Range-extended PHEV Increases eVMT and value proposition; expands market Minimal PHEV Increases eVMT and value proposition; expands market Home Long-range BEV Virtual necessity Limited-range BEV Virtual necessity Range-extended PHEV Virtual necessity Minimal PHEV Virtual necessity a Assumptions in this analysis are that electricity costs would be cheaper than gasoline costs, that away-from-home charging would generally cost as much as or more than home charging, that people would not plan to change their mobility needs to acquire a PEV, and that there would be no disruptive changes to current PEV performance and only gradual improvements in battery capacity over time. b It is possible that these infrastructure categories could expand the market for the various types of PEVs as appropriate, but that link is more tenuous than the cases noted in the table for other infrastructure categories. NOTE: AC, alternating current; BEV, battery electric vehicle; DC, direct current; eVMT, electric vehicle miles traveled; NA, not applicable; PEV, plug-in electric vehicle; PHEV, plug-in hybrid electric vehicle.

intend to meet their charging needs predominantly through charging, only 27 percent of multifamily dwellings had park- slow charging at home. ing spaces with access to charging (Consumers Union and Lack of access to charging infrastructure at home will the Union of Concerned Scientists 2013). Multifamily resi- constitute a significant barrier to PEV deployment for house- dential complexes can face many challenges in installing PEV holds without a dedicated parking spot or for whom the park- charging equipment; some are similar to a typical commer- ing location is far from access to electricity. Those demo- cial building, and others are unique to multifamily dwellings. graphic groups include many owners and renters of housing Similar to commercial buildings, the electrical panel might in multifamily dwellings and many households in large cities be far from the desired charging location, and installation can with on-street parking. About 25 percent of U.S. households therefore be costly. live in multifamily residential complexes (U.S. Census Bu- Unique to multifamily residential complexes are the own- reau 2011b), and the telephone survey noted above indicated ership, responsibility, liability, and control of each individual that although 61 percent of single-family houses had access to parking space. Multifamily residential complexes have many

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 101 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

86 Overcoming Barriers to Deployment of Plug-in Electric Vehicles ways to assign parking to their residents, including dedicated, interesting model for extending PEV driving to households shared, and leased parking. For residents who have dedicated without access to home charging is to deploy PEVs in car- spaces, the main challenges besides the installation costs are sharing fleets. That approach is particularly important for the questions within the governance structure for multifamily res- large portion of multifamily dwelling residents who are not in idential complexes concerning (1) who should bear the cost the new vehicle market as compared with single-family home of upgrading the main panel (if needed) and (2) who will pay residents. Car sharing is discussed from a consumer perspec- for the electricity for charging the PEV. Those costs can be tive in Chapter 3. prohibitive for an individual consumer if he or she is responsible for upgrading service to the main panel for the multi- Finding: Homes are and will likely remain the most impor- family dwelling. For residents who have shared spaces, ad- tant location for charging infrastructure. ditional questions need to be resolved within the governance structure of the multifamily residential complex concerning Finding: Lack of access to charging infrastructure for resi- installation costs, use of charge-enabled spaces, and payment dents of multifamily dwellings is a barrier that will need to be for the electricity. Because no charging space is dedicated to overcome to promote PEV deployment to that segment. a specific resident, an individual is discouraged from investing in the installation of a charging station because that would Workplace Charging not necessarily guarantee him or her the right to use it. In addition, the use of the charging station can no longer be tied Charging at workplaces provides an important opportu- to an individual and raises the question of who should pay nity to encourage the adoption of PEVs and increase eVMT. for the electricity. Lastly, for leased or rented spaces, there is BEV drivers could potentially double their daily range as long the question of ownership of the PEV charging equipment: as their vehicles could be fully charged both at work and at which entity should pay for the PEV charging equipment and home, and PHEV drivers could potentially double their all- how should liability be assigned? If tenants are liable for all electric miles. Extending the electric range of PHEVs with upgrades, they have a disincentive to perform the upgrades workplace charging improves the value proposition for PHEV because they might leave. If the owners are liable for all up- drivers because electric fueling is less expensive than gaso- grades, they have a disincentive to install them unless they line. For BEVs and PHEVs, workplace charging could expand can charge a premium for them or otherwise be compensated. the number of people whose needs could be served by a PEV, For residents who do not have any parking available and thereby expanding the market for PEVs. Workplace charging must rely on on-street parking, the same challenges exist ex- might also allow households that lack access to residential cept that the owner or deciding body is not the multifamily charging the opportunity to commute with a PEV. Further- residential complex. Instead, it is the local city government more, Peterson and Michalek (2013) estimated that installing that must make policy decisions surrounding installation and workplace charging was more cost-effective than installing operation of PEV charging equipment (Peterson 2011). public charging; however, it should be noted that installing Lack of home charging at multifamily complexes or in workplace or public charging was substantially less cost effec- neighborhoods with on-street parking is a barrier to deploy- tive than improving the all-electric range of a vehicle. ment for owners of all types of PEVs, but most importantly Data from early adopters in the EV Project shows that BEVs, particularly limited-range BEVs for which daily charg- workplace charging is used when it is available (Table 5-2). ing cannot, like PHEVs, be replaced with gasoline or, like Specifically, Nissan Leaf drivers who had access to work- long-range BEVs, postponed. It is also a barrier to increased place charging obtained 30 percent of their charging energy eVMT for all PEV owners. Overcoming lack of home charg- at work, and Chevrolet Volt drivers who had access to working at multifamily residential complexes and in neighborhoods place charging obtained 37 percent at work. Furthermore, with on-street parking requires providing such consumers there is some evidence that workplace charging enables lon- with designated parking spaces to charge their vehicles dur- ger routine commutes or more daily miles. Of Nissan Leafs ing prolonged times when their vehicles are not in use, such that had workplace charging, 14 percent routinely required as at workplaces. Although retrofits of multifamily housing workplace charging to complete their daily mileage (at least for PEV charging might be difficult, facilitating installation of 50 percent of days), but another 43 percent of the Leaf ve- home-charging infrastructure can be accomplished by prepar- hicles required workplace charging to complete their daily ing the sites for installation during initial construction. Cali- miles on some days (at least 5 percent of days). Moreover, fornia mandatory building codes will require new multifamily Nissan Leaf drivers extended their range by 15 miles or 26 dwellings to be capable of supporting future charging installa- percent on days when charging was needed to complete their tions (DOE 2014a).2 Additionally, multifamily dwelling own- trips (such days averaged 73 miles traveled) and by 12 per- ers might choose to contract with a charging provider to facili- cent on days when they charged even though a charge was tate installation and payment for charging services. Another not required to complete their trips (Smart 2014b). In considering whether to provide workplace charging, 2 For an explanation of these codes, see California Green Building employers confront a number of challenges. One set of chal- Code A4.106.8.2 and California AB 1092.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 102 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 87

TABLE 5-2 Charging Patterns for Nissan Leafs and Chevrolet Volts Percent Charging Energy Obtained at Various Locations Vehicle Home Work Other All Drivers Nissan Leaf 86 — 14 Chevrolet Volt 85 — 15 Drivers with Access to Workplace Charging Nissan Leaf (~12%)a 68 30 2 Chevrolet Volt (~5%)a 60 37 3 a Numbers in parentheses are percentage of drivers known to have access to workplace charging. SOURCE: Based on data from ECOtality (2014a,b).

lenges is to determine the rate of PEV adoption by employees, of increasing the number of workplace charging stations in what level of charging would be sufficient for their needs, and proportion to the number of employees who express an in- how access to chargers can be ensured as the number of PEVs terest in using them. This tends to have positive feedback increases. A worker who relies on workplace charging of a effects as increases in the number employees who use work- BEV might not be able to return home if no charger is avail- place charging stations stimulate other employees’ interests able. There is also the possibility that electricity provided to in acquiring PEVs (Jennings 2013), thereby contributing to a employees will have to be paid for by the employees or taxed continuing expansion in the number of workplace chargers. as income (IRS 2014).3 A requirement to assess the value of Other firms, however, have been reluctant to provide work- the charging or report the imputed income could be an impedi- place charging on grounds of equity, expressing concerns ment to workplace charging. Yet another potential impediment about providing a perk that would benefit only a relatively arises from the surcharges that utilities impose on companies small number of employees, at least initially (Musgrove that draw more than a threshold level of power. Such demand 2013). Recognizing workplace charging as an important op- charges (discussed in Chapter 6) can be substantial. portunity to expand PEV deployment and eVMT, DOE sup- Workplace charging is becoming available at a small ports the EV Everywhere Workplace Charging Challenge. but growing number of companies that offer it as a way of The Workplace Charging Challenge and the Clean Cities attracting and retaining employees and as a way of distin- program both provide several guides and resources for em- guishing themselves as green companies.4 It is an attractive ployers to simplify the process of adding workplace charg- perk if the employer provides charging for the same price ing (DOE 2014b; DOE 2013). or less than is available at home. In assessing the reasons for offering workplace charging, some employers anticipate Finding: Workplace charging could be an alternative to that concerns about carbon emissions from commuting will home charging for those who do not have access to charging eventually generate much stronger pressures for workplace infrastructure at home. charging and are attempting to move expeditiously by expanding their network of charging stations now (Ahmed Finding: Charging at workplaces provides an important op- 2013). Because of the costs involved and the fact that add- portunity to encourage PEV adoption and increase the frac- ing a charging station leaves fewer parking spaces available tion of miles that are fueled by electricity. for employees who do not drive PEVs, Cisco has a policy Finding: The administrative cost to assess the value of charg- 3 IRS Publication 15-B states that any fringe benefit is taxable ing or report the imputed income could be an impediment to and must be included in the recipient’s pay unless the law explic- workplaces to install charging. itly excludes it. Although exclusions currently apply to many fringe benefits, the issue of excluding electricity that employers provide at workplace chargers has apparently not yet been explicitly addressed. Recommendation: The federal government should explicitly The issue does not arise at workplaces that engage an outside entity address whether the provision of workplace charging at the (the installer of the charging infrastructure) to manage the charging expense of employers should be included in the recipient’s units and collect a monthly fee from workers who use them. pay or regarded as a benefit that is exempted from taxation. 4 To facilitate the process, the Department of Energy (DOE), under the Workplace Charging Challenge launched in January 2013, offers various resources to interested employers, building owners, Public Charging Infrastructure employees, and others. The resources include information about PEVs, their charging needs, and activities that DOE and communi- A critical question to answer is whether lack of public ties across the country are doing to support PEV deployment.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 103 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

88 Overcoming Barriers to Deployment of Plug-in Electric Vehicles charging infrastructure is a barrier to PEV deployment.5 As range BEV drivers, who have the greatest need for charging. shown in Figure 5-1, home charging infrastructure is and is One study looked at the locations where light-duty vehicles expected to remain more convenient and more critical to PEV parked and modeled optimal charging locations assuming deployment than public charging infrastructure. There is no similar trip needs for PEV drivers and ICE drivers (Chen et consensus in the research and policy communities, howev- al. 2013). Other studies have examined trip diary data from er, on the impact of public charging infrastructure on PEV such cities as Seattle and Chicago and such states as California deployment. Experience in Japan indicates that increased to see which trips were not likely to be completed with to- availability of public charging stations reduces range anxiety day’s BEVs and sought to place chargers to allow completion and leads to more miles driven by BEVs. For example, the of these “failed” trips. Models were optimized by minimiz- building of a single additional fast charger for a TEPCO fleet ing time or distance deviations from trips required to drive to of BEVs increased eVMT from 203 km/month to 1,472 km/ charging locations. The study of California drivers found that month. Interestingly, no additional energy consumption from with an 80-mile limited-range BEV, 71.2 percent of the total the public charger was observed after building the second miles driven and 95 percent of trips could be completed with charger, but drivers allowed their state of charge to go below no public charging required. Optimal placement of 200 DC 50 percent, a sign that their fear of running out of charge had fast chargers in the state would allow those drivers to complete been alleviated (Anegawa 2010). over 90 percent of miles with two or fewer charges (Nicholas DOE (2015) estimates that there were more than 9,300 et al. 2013). The data from Chicago and Seattle metro areas public charging stations in the United States as of April 2015; showed that no public charging was needed to complete 94 many stations, however, are only accessible to members of percent and 97 percent of trips, respectively, and optimally lo- associated subscription-based plans or to vehicles produced cating 100 or 50 stations with 10 AC level 2 chargers each in by individual manufacturers. Interactive maps of charg- Chicago or Seattle resulted in mean route deviations of only ing stations are updated frequently on the DOE Alternative 1.6 and 0.3 miles, respectively, to make the remaining trips Fuels Database and through the PlugShare website (DOE (Andrews et al. 2013). As noted, most studies have not in- 2015; Recargo 2014). Nearly 8,700 of the public charging vestigated the effect of charging infrastructure deployment on stations provide AC level 2 chargers, which can add about vehicle deployment. 10-20 miles of range to a vehicle for each hour of charging, The majority of public charging stations are not yet depending on the model and driving conditions. More than heavily used. For example, public DC fast chargers in the 800 public DC fast-charging stations had also been installed EV Project were occupied on average 2.3 percent of the time by April 2015 (DOE 2015). Networks of DC fast chargers from October-December 2013, and public AC level 2 char- have been installed in Washington, Oregon, and California; gers were occupied 5.5 percent of the time on average (INL along the East Coast I-95 corridor; and the “Tennessee Tri- 2014). Despite that low utilization, it is not unusual at some angle,” which connects Nashville, Chattanooga, and Knox- popular stations for drivers to have to wait for a charging ville. Clusters of DC fast chargers are also in Dallas-Fort plug to become available. In addressing the adequacy of the Worth, Houston, Phoenix, Atlanta, Chicago, and Southern existing network of public charging infrastructure, it is im- Florida. Tesla and Nissan Motors—manufacturers of the ve- portant to understand the factors that contribute to both over- hicles that have led BEV sales in the United States—have utilization and underutilization. The factors include the ratio been actively engaged in expanding their networks of fast of charging stations to PEVs in any given area, the location chargers. In fact, most of the chargers outside of the regions of charging stations, the cost of using the stations, the amount noted above are part of the proprietary Tesla network of Su- of time it takes to recharge, and restrictions on station use as- perchargers (see Chapter 2, Figure 2-10). Tesla had installed sociated with either subscription-membership requirements more than 190 charging stations in the continental United or incompatible hardware. Low utilization of the charging States and Canada by April 2015 and has plans to expand its stations in a given area does not necessarily imply that the network to several hundred stations by the end of 2015, with network of charging infrastructure is adequate and could in- the stated goal that 98 percent of U.S. drivers are within 100 stead reflect any combination of the factors noted. Similarly, miles of a Supercharger by 2015 (Tesla 2014). Nissan has queuing at charging stations does not necessarily imply that announced plans to add at least 500 fast-charging stations more charging stations should be built, but it is unlikely that by mid-2015 and has partnered with CarCharging to expand most potential customers would be willing to wait for multi- networks in California and on the East Coast and with NRG/ ple charges to be completed. To the extent that the demand to eVgo to develop a network in the Washington, D.C. area use charging stations is not uniformly distributed over time (CarCharging 2013; Nissan 2013). and that investments in charging stations are costly, a certain Several studies have modeled optimal numbers and loca- degree of queuing is inherent in a network of charging stations of PEV charging sites from the perspective of limited- tions that optimally balances the cost of waiting to charge against the cost of building more charging stations. In addi- 5 The term public charging infrastructure refers to charging infra- tion, at stations that do not impose usage fees or charges for structure that is located in public spaces but does not imply that the electricity consumed, queuing might partly reflect the fact services are offered for free.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 104 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 89 that using those stations is cheaper than charging at home. are often parked for just an hour or two, such as at shop- For some locations, such as retail establishments, medical ping malls, museums, libraries, and restaurants. Installation facilities, and commercial parking lots, for-pay AC level 2 of chargers at those locations is often seen as a way for busi- infrastructure is used more frequently than free public AC nesses to attract customers. Charging providers are also in- level 2 infrastructure; this might indicate better siting of or stalling AC level 1 and AC level 2 within cities as part of more chargers to reduce queueing at for-pay infrastructure their subscription-based business model. Some utilities are (Smart and White 2014). also installing infrastructure and are motivated to provide Over the course of its study, the committee heard con- public charging to encourage PEV deployment and hence cerns that public funding combined with pressures to install sell more electricity to residential customers with PEVs. In- public infrastructure quickly has led to some poor siting de- frastructure-deployment models are discussed in more detail cisions. So, the fundamental questions remain—how much at the conclusion of this chapter. public infrastructure is needed and where should it be located? Although the committee did not attempt to establish There are many complexities associated with installing pub- guidelines for locating public charging infrastructure, it lic charging infrastructure that need to be considered. It can seems reasonable to assume that to maximize the use of in- be located within cities, such as at malls or parking lots, or tracity charging infrastructure, chargers must be dispersed along interstate highways or other corridors. It can include AC around metropolitan areas and placed at convenient loca- level 1, AC level 2, and DC fast charging. It can be costly to tions. Siting of public charging stations is driven by a variety install and maintain, and its effect on deployment and eVMT of motivations, and the stations are operated by both pub- remains unclear, although it enables PEV drivers to extend the lic and for-profit entities. Charging providers might locate electric range of their vehicles beyond the mileage that can be public stations to maximize revenue from for-pay stations, driven on a single charge and might encourage the adoption of to establish their image as a green business or government, limited-range BEVs by mitigating concerns about becoming to induce customers to stop at their establishments, to take stranded. However, a substantial amount of public charging advantage of favorable conditions (such as no-cost land or infrastructure that is obviously unused could become a symbol easy access to electricity source), to increase deployment that PEVs are not as practical as had been hoped. The follow- of vehicles, to increase eVMT, or to relieve range anxiety. ing sections consider the location of public infrastructure and Data from intracity AC level 2 infrastructure associated with its effects on PEV deployment and eVMT. the EV Project indicate that chargers located at parking lots and garages, transportation hubs, workplaces, and public or Finding: Public charging infrastructure has the potential to municipal sites were used most frequently. Least frequently provide range confidence and extend the range for limited- used sites were at educational institutions, multifamily resi- range BEV drivers, to allow long-distance travel for long- dences, and medical facilities (Smart 2014c). range BEV drivers, and to increase eVMT and the value prop- The effects of intracity AC level 1 and level 2 charg- osition for PHEV drivers. ing infrastructure vary by PEV class as seen in Table 5-1. Long-range BEVs will have little use for slow charging Finding: More research and market experience are needed in public locations as there will be little value of charging to determine how much public infrastructure is needed and slowly given their sufficient all-electric range. However, where it should be sited to promote PEV deployment and to they might choose to top-off their charge when convenient encourage PEV owners to optimize vehicle usage. or if perks, such as free parking at an airport, are available. Limited-range BEVs are expected to experience the most Recommendation: The federal government through the De- utility from intracity AC level 1 and level 2 charging by as- partments of Energy or Transportation should sponsor research suring them that they will not be stranded if their charge is to study the impact of the public charging infrastructure, includ- depleted and by allowing them to extend their daily mileage ing the extent to which its availability affects PEV adoption. beyond a full battery charge. With limited battery ranges and no other choice for fuel, charging in public is an attractive Intracity AC Level 1 and Level 2 option for limited-range BEVs. Both minimal and extended- Charging Infrastructure range PHEVs are predicted to use intracity AC level 1 and level 2 charging for increased eVMT and hence to realize an Public AC level 1 and level 2 chargers are now avail- increased value proposition of their vehicles. However, they able in some cities, especially where PEV deployment has do not need intracity chargers for range extension or range been relatively strong. Because AC level 1 chargers provide confidence because they can also fuel on gasoline. Increased about 4-5 miles of operation per hour of charge, they could eVMT from charging in public might be particularly useful be used when charging time is not a primary concern, such for minimal PHEVs whose smaller batteries could be nearly as at airports and train stations, where people park their cars fully charged in a shorter time, thus extending their small for prolonged periods. They can also be installed easily us- ranges substantially if they are able to charge frequently ing accessible 120 V outlets. AC level 2 chargers are also throughout the day. becoming increasingly available at locations where vehicles

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 105 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

90 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Intracity DC Fast-Charging Infrastructure limited-range BEVs and enabled long-distance travel for long- range BEVs. An example of such a network is the corridor DC fast-charging technology was described in Chapter 2. of DC fast chargers installed at about 40-mile intervals along Although DC fast chargers are often considered for corridor Interstate 5 in Washington and Oregon. Such infrastructure travel, such as between cities or states, the majority of the fast- provides long-range BEVs with multiple places to acquire a charge infrastructure is installed within cities and their metro charge on an extended trip and enables limited-range BEVs areas. There are some data to indicate that BEV owners prefer to travel between two cities in the same region. For travel fast charging to complete a journey or otherwise to create op- between cities where stops to charge might be inconvenient, tions for using the vehicle beyond its routine range. EV Project DC fast chargers are expected to be used primarily for range data on the percent of DC fast charges that occurred on trips extension and are expected to receive less use than DC fast of a given length provide information on charging behavior of chargers within cities. Although data from the EV Project is Nissan Leaf drivers (Smart and White 2014; J. Smart, Idaho primarily from cities, a preliminary study of charging along National Laboratory, personal communication, November 6, the I-5 corridor shows that most charges do in fact occur with- 2014). In the fourth quarter of 2013, after the institution of in cities rather than between them (Smart 2014d). Although fees to charge at some DC fast-charging locations, 56 percent some early adopters of limited-range BEVs have chosen to of outings that included a fast charge were greater than 60 drive their vehicles long distances requiring multiple battery miles round trip, and 44 percent of outings that included a fast charges, the committee’s view is that the vast majority of lim- charge were less than 60 miles round trip. Some of the less ited-range BEV drivers will restrict themselves to a range that than 60 mile round-trips that included a DC fast charge might requires at most one full charge between neighboring cities. reflect the value a driver places on a DC fast charge even when As noted, PHEVs are not equipped to use DC fast-charging it is not required to complete the trip. However, many of the stations and can extend their range by refueling on gasoline. short trips (63 percent) started with a less than full battery, Thus, interstate DC fast chargers are projected to be the indicating that the charge might have been required to return least important type of infrastructure for PEVs because it will home. When an outing included a DC fast charge and began not (or cannot) be used by PHEVs and will be inconvenient with a full battery, average round trip distance was 87.5 miles. for limited-range BEVs. However, it should be noted that That observation again indicates that many trips that include there are alternative scenarios in which interstate DC fast a DC fast charge required a charge to complete, and DC fast chargers do become an important type of infrastructure. An charging might have been the most convenient way to acquire example of such a scenario is if the market becomes domi- the charge. nated by long-range BEVs that are used as primary vehicles. The impact of intracity DC fast-charging infrastructure If that is the case, home charging infrastructure will continue varies by PEV type, as noted in Table 5-1. Long-range BEVs to be most important for drivers’ everyday usage, and work- will have little use for fast charging in cities as their vehicle place and intracity infrastructure will be relatively unimport- range is unlikely to require range extension or range confi- ant. Intercity and interstate charging would, in that scenario, dence. However, charging at a DC fast-charging station would enable long-range BEVs to take longer trips with relative ease. allow them to acquire a full battery charge more quickly than Vehicle manufacturers, especially those focused on BEVs, are home charging; this option might be valuable to a long-range building intracity, intercity, and interstate DC fast-charging BEV owner, particularly one who does not have a place to infrastructure; this indicates that they think it is valuable. It is charge at home. The committee notes that Tesla—the only cur- not clear whether they are doing this for marketing or business rent producer of a long-range BEV—is implementing a model strategy reasons or to spur vehicle deployment in the near term in which charging at its DC fast charger stations is included in or whether they believe that this type of infrastructure will be the price of the vehicle. Limited-range BEVs are expected to necessary in the future. experience the most utility from intracity DC fast charging as it provides range confidence that they will not be stranded and MODELS FOR INFRASTRUCTURE DEPLOYMENT range extension in less time than that required for AC level 1 or level 2 charging. In April 2014, Nissan began offering new To understand how best to overcome any infrastructure Leaf buyers in several markets free public charging through barriers to PEV deployment, one must consider the installa- a special card that allows using several charging providers. tion and operating costs for the different categories of charg- Range-extended and minimal PHEVs are unable to use DC ing infrastructure, the possible deployment models, and who fast-charging infrastructure, so this segment of infrastructure might have an incentive to build such infrastructure. Several deployment does not apply to PHEV owners. different entities might have an incentive to build or operate charging infrastructure; these include vehicle owners, work- Intercity and Interstate DC places, retailers, charging providers, utilities, vehicle manu- Fast-Charging Infrastructure facturers, and the government. Their motivations might include generating revenue, improving air quality, selling more The availability of DC fast chargers along highways electricity, or selling more PEVs. On the basis of information connecting cities and states has facilitated regional travel for

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 106 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 91

TABLE 5-3 Entities That Might Have an Incentive to Install Each Charging Infrastructure Category Infrastructure Category Who Has an Incentive to Install? Location Type Interstate DC fast Vehicle manufacturer, government Intercity DC fast Vehicle manufacturer, government Intracity DC fast Vehicle manufacturer, government, charging provider, utility Intracity AC level 1 or level 2 Utility, retailer, charging provider, vehicle manufacturer Workplace AC level 1 or level 2 Business owner, utility Home or fleet base AC level 1 or level 2 Vehicle owner, utility NOTE: AC, alternating current; DC, direct current.

received during site visits and from presentations from vari- Workplace Charging ous infrastructure providers, the committee’s assessment of the possible builders of each infrastructure category is sum- Private charging infrastructure at workplaces is likely marized in Table 5-3. It should be noted that the most critical to be funded by the businesses or organizations. The instal- infrastructure (home charging) is also the least logistically lation and operating costs of workplace charging might be complicated and least expensive to build, and the costs and justified by the employer as a perk to attract and retain em- complications generally increase for faster charging and ployees or to brand the company with a green image. Be- more public locations. The following paragraphs discuss the cause vehicles are parked at work for long periods of time infrastructure-deployment models associated with each in- (see Figure 5-2), many workplaces do not find it necessary frastructure segment and the installation and operating costs. to upgrade even to AC level 2 charging. Some parking lots might already have AC level 1 outlets that can be repurposed Home Charging for vehicle charging; however, more convenient or upgraded infrastructure might also be installed. Another entity that Private charging infrastructure at home is likely to be might have an interest in installing workplace charging is a funded by the homeowner. Financing and logistics of install- utility, which could earn additional revenue from the sale of ing home charging infrastructure is not considered to be an electricity at worksites. important barrier for homeowners who have dedicated parking The cost of installing charging varies from workplace spots adjacent to their homes. Homeowners who own PEVs to workplace but is generally higher than that for installing have a clear incentive to install home charging. Many will also single-family home charging and lower than that for public find the expense of upgrading to AC level 2 infrastructure to charging infrastructure. The costs of labor and conduit for be a good investment, especially owners of long-range BEVs installing charging units in existing parking lots and garages who might want to charge their vehicle batteries more quickly. depend mainly on how much digging and resurfacing is in- Aside from vehicle owners paying to install charging infra- volved. There are also potential costs associated with elec- structure, other deployment models are being implemented. tric service upgrades for AC level 2 chargers, which might Some providers of subscription-based charging have expand- be the best choice for most currently available PEVs that ed into providing residential charging infrastructure as part of have large electric ranges. Cisco provided a set of ballpark their subscription service. Utilities might also have an interest estimates to the committee and indicated that the average in providing residential charging infrastructure as it would in- cost of installing an AC level 2 charging station has been crease electricity usage at the residence. $10,000-$15,000 (with economies of scale), that the ongoing As discussed previously, multifamily residential home costs of paying a vendor to manage the stations has aver- charging faces many more barriers, and it is not clear that aged about $25 per station per month, and that the electricity many owners of complexes, drivers of vehicles, or munici- costs have been low (Ahmed 2013). However, Bordon and palities will have incentive to install charging at multifamily Boske (2013) suggest that the cost of installing an AC level residences or at on-street charging locations in residential 2 charger in a commercial garage or on a public street ranges neighborhoods. However, owners of multifamily residences from $2,000 to $8,000 on the basis of estimates from three might be motivated to install chargers because they can earn separate sources. points toward Leadership in Energy and Environmental De- In addition to installation costs, operating costs of pro- sign certification (AeroVironment 2010). They might also be viding charging to employees must be considered. The com- able to market their property as green and offer charging as mittee received reports that the costs of electricity were not a an attractive amenity to prospective renters. barrier to deployment of workplace charging, but two logisti-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 107 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

92 Overcoming Barriers to Deployment of Plug-in Electric Vehicles cal concerns were raised. As mentioned above and discussed ing the auxiliary services and features needed for a publicly further in Chapter 6, demand charges that could increase the accessible unit, including warranty, maintenance, customer cost of electricity to the employer could be a cost barrier to authentication, and networking with point-of-sale capabili- workplaces installing charging for employees. Also, the po- ties to collect payment from customers. Installation costs can tential need to classify workplace charging as imputed income also vary because of other enhanced safety and security mea- has resulted in logistical barriers given the associated admin- sures that are often required by local permitting authorities, istrative requirements for monitoring charging time or energy such as lighting and revenue-grade meters. Those options and making associated payroll adjustments. In part to avoid can add up to $90,000 to the basic cost of the fast-charging that potential problem and also to outsource charger installa- equipment itself. Additional costs might also be incurred if tion and maintenance, some employers have chosen to con- multiple plugs are required for compatibility. tract with charging providers to install and operate charging infrastructure, including charging for the electricity provided. Retailers

Finding: Some workplaces appear to have incentives for in- A number of major retailers have shown interest in pro- stalling charging infrastructure, including fostering an envi- viding space for charging stations (Motavalli 2013),6 partic- ronmentally friendly image and providing the perk to retain ularly when the capital costs are subsidized. Such infrastruc- and recruit employees. ture can attract customers to park and spend time and money in the retail establishments and might also provide favorable Recommendation: Local governments should engage with branding for the retailers. Most of the charging units that re- and encourage workplaces to consider investments in charging tailers have provided to date have been AC level 1 or level 2 infrastructure and provide information about best practices. stations, which are used primarily for intracity charging. The costs of building charging infrastructure at retail establish- Public Charging Infrastructure ments range widely but are probably similar to workplaces and related to the amount of conduit required to provide As discussed above, charging infrastructure generally electric access at parking spots. It is not clear that the extra becomes more complicated and more costly to build and money spent in retail establishments by customers who use operate as it becomes more publicly accessible and delivers the charging stations is sufficient to provide retailers with faster charging. The potential owners and operators of public incentives to incur the capital costs of installing charging charging infrastructure are discussed in the sections below. stations, as distinct from simply covering electricity charges Generally, companies that install and operate public charg- and service costs. When capital costs are covered by others, ing stations have five sources from which they can seek to however, retailers have tended to contract with charging pro- cover their capital and operating costs: the government, utili- viders to build and maintain charging stations and possibly ties, vehicle manufacturers, charging-station hosts, and driv- charge customers for their use. ers. Most companies have depended on government grants to finance a large part of their investments to date, and it is Electric Utilities difficult to tell whether their business models will be sustainable in the absence of public funding. The electric utility companies could emerge as a willing The costs of DC fast-charging stations are generally source of capital for public charging stations. That conclusion much higher than the costs of AC level 2 stations. In general, reflects the prospect that a network of public charging stations the capital costs depend on several factors: whether the prop- would induce more utility customers to purchase PEVs, which erty must be purchased, leased, or rented; what distance must would lead not only to electricity consumption at the public be spanned to connect to higher voltage supply lines; wheth- chargers, but also to much greater consumption of electric- er upgrades are required, for example, because of insufficient ity at residences served by the utilities. If public charging in- transformer capacity; how much trenching and conduit are frastructure drives greater eVMT and greater deployment of needed to reach the charging station; and how much repav- vehicles, capital and variable costs for public infrastructure ing or restriping of the parking area is required to accommo- might be covered by the incremental revenue from additional date the charging station. In total, the costs can range from electricity that PEV drivers consume at home, where rough- $100,000 to $200,000. As an example, Table 5-4 shows the ly 80 percent of PEV charging takes place (Francfort 2011). average costs of installing charging stations in Washington Most such charging infrastructure is expected to be built intra- State with DC fast chargers and AC level 2 chargers as part city. Austin Energy (2012), with the help of a series of federal of the publicly funded West Coast Electric Highway proj- government grants, is an example of a utility that has chosen ect. The totals shown in the table—ranging from $109,500 to $122,000—exclude the costs of purchasing, renting, or leasing land. The basic cost of a DC fast-charging station 6 Major retail companies that have installed or plan to install charg- is about $10,000 to $15,000, but the total equipment cost ing stations for their customers include Best Buy, Chili’s, Cracker - Barrel, Kroger, Macy’s, 7-Eleven, Tim Hortons, Walgreens, and of the Washington state stations averaged $58,000, reflect Whole Foods.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 108 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 93

TABLE 5-4 Costs of Installing Public DC Fast-Charging Stations for the West Coast Electric Highway Projecta Component Cost DC fast-charging equipment $58,000 • 50 kW DC public fast-charging station (480 V ac input) per unit • 3-year warranty and point-of-sale capabilitiesb • Payment of all electricity dispensed (including utility demand charges) • Overhead lighting and required safety equipment

Level 2 charger colocated next to DC fast-charging station $2,500 • 240 V/30 A AC level 2 public charger per unit • Same terms and conditions as listed above

Equipment installation (labor and electric-panel upgrade) $26,000 • Separate power drop or meter for the charging station per location • Electric panel upgrade (if required) • Construction and environmental and electricity permits • Trenching, backfill, and site restoration • Installation of conduit and power lines to charging station • Installation of concrete pad and electric stub-out • Installation of curb or wheel stop and overhead lighting • Installation and testing of equipment

Utility interconnection $12,500 to $25,000 • Costs are highly variable and depend on cost-recovery policies of the electric-power provider and condition of per location existing power distribution componentsc • Generally includes utility costs for preliminary engineering and design, transformer upgrades, and labor for connection to the grid

Host-site identification, analysis, and screening $5,000 • Identification of potential sites per location • Consultation with electric-power providers

Negotiation, legal review, and execution of lease $6,000 • Making contact with several property owners per location • Exchanging and negotiating lease documents • Executing and recording documents Total for DC fast charger and 3-year service $109,500 to $122,000 a Land costs are not included here. b Point-of-sale capabilities might include radiofrequency identification authentication and networking to back-office functions (such as account management and customer billing), equipment status signals, and credit card transactions. c Additional costs could be incurred if addition of multiple chargers increases demand charges or requires additional electricity service upgrades. NOTE: A, amperes; AC, alternating current; DC, direct current; kW, kilowatt; V, volt. SOURCE: Based on data from PB (2009).

to install a network of AC level 2 charging stations in its ser- infrastructure. Such a mechanism would not have to rely vice area, where it is the only electricity provider, and to offer on government subsidies or cross subsidization from house- its residential customers unlimited use of the chargers for less holds that did not own a PEV. That said, whether utilities that than $5 a month. In addition, Austin Energy provides incen- invested their own capital in charging stations could earn a tives for a range of additional infrastructure charging catego- respectable rate of return over time would depend on state- ries as part of its strategic objectives in demand management level regulatory policies that are used to encourage utility and ancillary services (K. Popham, Austin Energy, personal investment. communication, December 18, 2014). The committee notes that theoretically all utilities ser- Commercial Charging Providers vicing a given geographical area would collectively have a viable business model if there were a mechanism to (1) Mostly in response to government grants, several pri- separate out and pool the electricity sales to all households vate companies have entered the business of installing and that owned PEVs within that area and (2) share the revenues managing public charging stations. These charging stations from that pool in proportion to the amounts that the differ- are a mix of AC level 2 and DC fast chargers and are located ent utilities contributed to investments in public charging both between and within cities. The companies have been

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 109 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

94 Overcoming Barriers to Deployment of Plug-in Electric Vehicles experimenting with different models in their efforts to recov- in the market. They might be one of the only private sector er their capital costs and the costs of electricity. For example, entities with a motive to install fast charging along intercity ChargePoint (2014) is pricing on a per-charge-event basis, and interstate highways, as this type of infrastructure is the while NRG/eVgo (2014) relies on both a monthly subscrip- most expensive to build and is unlikely to generate high tion fee and a fee per minute of plug-in time. Depending on returns from for-pay charging. As noted earlier, Tesla has state legislative and regulatory rulings, charging providers launched a program to install several hundred supercharging might avoid being regulated as utilities by not charging in stations along major long-distance transportation corridors proportion to the amount of electricity consumed (see fur- throughout the United States, while Nissan has launched ther discussion in Chapter 6). NRG/eVgo relies on its fee several joint ventures to increase substantially the number of per minute of plug-in time as a mechanism for encouraging fast chargers available in key market areas (DeMorro 2014). drivers to limit the amount of time that their vehicles occupy In the absence of government subsidies, it seems unlikely the parking spaces adjacent to the chargers. that any companies other than BEV manufacturers could have Although it might be easy to cover the variable costs a business case for covering the installation and maintenance of their operations from the various fees paid by customers costs of DC fast-charging infrastructure deployed in intercity (for example, monthly subscription fees or fees per charging and interstate highway corridors. Whether the infrastructure event or per minute of charging time), generating an attrac- would be publicly accessible is uncertain as a vehicle man- tive rate of return on invested capital is much more challeng- ufacturer would have little incentive for providing charging ing. One of the early providers of charging infrastructure, infrastructure for PEVs that it did not produce. For example, ECOtality, encountered financial difficulties and filed for only Tesla customers can use Tesla-built chargers because of a bankruptcy in October 2013; its Blink assets, including the Tesla-specific plug. In the case of Nissan, which is also build- network of Blink charging stations, have been purchased by ing and subsidizing chargers, their chargers can be used by CarCharging (Wald 2013). many types of PEVs but might require payment from those The infrastructure-deployment model adopted by NRG/ not covered under Nissan’s No-Charge-to-Charge plan. eVgo provides a unique approach. It is oriented toward providing a simple and complete set of services to residential Federal Government customers. NRG/eVgo (2014) offers its Houston customers a 1-year contract for a $15 monthly fee that covers the installa- If a category of charging infrastructure is deemed to be tion of charging equipment at home and provides unlimited particularly effective at inducing PEV deployment but no access to its network of public stations at 10 cents per minute private sector entity has a strong case for building such infra- of plug-in time. And unlike most other public stations, its structure, the federal government might consider funding it Freedom Chargers include DC fast chargers and AC level 2 as a worthwhile investment. The committee heard concerns chargers and are located mainly along major transportation that government money was likely to crowd out private in- corridors within the metropolitan areas it serves.7 vestments in infrastructure and to lead to poor siting deci- Box 5-1 provides a hypothetical calculation for the eco- sions in some cases. To ensure that charging infrastructure nomics of providing public charging stations using a business developers have an incentive to site chargers so that they model that collects monthly subscription fees and also charges will be well used, government infrastructure funding should customers for charging time. The calculation suggests that it comprise only a portion of the funding for a charging station might be difficult for charging providers to survive unless and should not go toward stations that would be deployed their capital costs are at least partially subsidized by public without government funding. Also, more research should be funding or by others, such as vehicle manufacturers. That said, done to ascertain what categories of charging infrastructure the committee heard concerns from private charging develop- lead to increases in deployment and eVMT. ers that subsidizing infrastructure investments tended to un- dermine the business models of firms that were prepared to Finding: Utilities that can capture the entire residential elec- finance infrastructure with their own capital. tricity consumption of PEV owners appear to have a viable business model for investing in public charging infrastructure. Vehicle Manufacturers Finding: Initiatives undertaken by Tesla and Nissan suggest Vehicle manufacturers might deploy public charging in- that vehicle manufacturers that wish to penetrate the market frastructure to drive sales of PEVs or to position themselves for BEVs perceive a business case for investing in extensive networks of DC fast-charging stations. 7 eVgo areas include Houston, Dallas-Fort Worth, Los Angeles, San Finding: Apart from BEV manufacturers and utilities (or Francisco, San Diego, the San Joaquin Valley, and Washington, D.C. To the extent that the provision of a network of fast-charging stations groups of utilities), the committee has not been able to iden- helps catalyze PEV sales, total electricity consumption will increase tify any private sector entities that have an attractive busi- by much more than electricity consumption at the eVgo charging sta- ness case for absorbing the full capital costs of investments tions and provide additional profits for NRG, which generates elec- in public charging infrastructure. tricity.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 110 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 95

BOX 5-1 Some Hypothetical Economics for Providers of Public Charging

This box considers the economics of providing a network of K public AC level 2 charging stations to serve N customers who rely primarily on residential charging but, on average, add 30 minutes of charge four times a month (or 1 hour of charge twice a month) at public chargers.

Assume that each charging station involves a capital outlay of $10,000 and that the investor requires a payback in 3 years, which in round terms amounts to about $3,600 per year per station, or $300 per month per station.

Assume that customers are charged 10 cents for each minute of plug-in time and that each hour of charging generates $2 of revenue over and above electricity costs plus maintenance costs. Thus, use of the charging network generates net revenues of $4N per month.

Assume that customers are willing to pay a subscription fee of $F per month for the assurance of access to the network of stations, implying subscription revenue of $NF per month.

Then the break-even value of N, calculated as a function of F, must satisfy NF = 300K - 4N, or

N = 300K/(F + 4)

And the break-even value of F as a function of K/N can be expressed as

F = 300(K/N) - 4

This suggests that a firm with 200 subscribers for every 10 charging stations could break even by charging a subscription fee of $11 per month.

Note, however, that the economics becomes much more difficult for networks of DC fast chargers, which require much larger capital outlays, or for AC level 2 networks that have to compete with networks of fast chargers.

Finding: The federal government might decide that provid- Austin Energy. 2012. “For Less Than $5 Per Month, Austin ing public charging infrastructure serves a public good when Energy Offers Unlimited Electric Vehicle Charging at others do not have a business case or other incentive to do so. More Than 100 EVerywhere™ Stations.” Press release, March 2. Recommendation: The federal government should refrain Borden, E.J., and L.B. Boske. 2013. Electric Vehicles and from additional direct investment in the installation of pub- Public Charging Infrastructure: Impediments and Oppor- lic charging infrastructure pending an evaluation of the rela- tunities for Success in the United States. Center for Trans- tionship between the availability of public charging and PEV portation Research, University of Texas at Austin, July. adoption or use. CarCharging. 2013. “Nissan and CarCharging to Expand Electric Vehicle Quick Charger Network.” News release, REFERENCES May 14. http://www.carcharging.com/about/news/all/nis san-and-carcharging-to-expand-electric-vehicle-quick- AeroVironment. 2010. “EVs and LEED Certification.” AV charger-network/. Connect, October. http://www.avinc.com/plugin/nl002/ev ChargePoint. 2014. “Start Charging Today.” https://www.char s_and_leed_certification. gepoint.com/get-started/. Accessed October 23, 2014. Ahmed, A. 2013. “Workplace Electric Vehicle Charging.” Pre- Chen, T.D., K.M. Kockelman, and M. Khan. 2013. “The sentation to the Committee on Overcoming Barriers to Electric Vehicle Charging Station Location Problem: A Electric-Vehicle Deployment, Washington, DC, August 13. Parking-Based Assignment Method for Seattle.” Pre- Andrews, M., M.K. Dogru, J.D. Hobby, Y. Jin, and G.H. Tucci. sented at the 92nd Annual Meeting of the Transportation 2013. “Modeling and Optimization for Electric Vehicle Research Board, Washington, DC, January 13-17. http:// Charging Infrastructure.” Innovative Smart Grid Tech- www.caee.utexas.edu/prof/kockelman/public_html/ nologies Conference, Washington, DC, February 24-27. TRB13EVparking.pdf. Anegawa, T. 2010. “Development of Quick Charging System Consumers Union and the Union of Concerned Scientists. for Electric Vehicle.” World Energy Congress, Montre- 2013. “Electric Vehicle Survey Methodology and As- al, Canada, September 12-16. http://89.206.150.89/doc sumptions: American Driving Habits, Vehicle Needs, and uments/congresspapers/322.pdf.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 111 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

96 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Attitudes Towards Electric Vehicles.” December. http:// Motavalli, J. 2013. “Shop While You Charge: Consumers and www.ucsusa.org/assets/documents/clean_vehicles/UCS- Retailers Are Plugging In.” New York Times, July 9. and-CU-Electric-Vehicle-Survey-Methodology.pdf. Musgrove, R. 2013. “Global Vehicles.” Presentation to the DeMorro, C. 2014. “Nissan Launches “No Charge to Charge” Committee on Overcoming Barriers to Electric-Vehicle Program for LEAF Buyers.” Clean Technica, July 10. Deployment, Washington, DC, December 3. http://cleantechnica.com/2014/07/10/nissan-launches- Nicholas, M., G. Tal, and J. Woodjack. 2013. “California State- charge-charge-program-leaf-buyers. wide Charging Assessment Model for Plug-in Electric Ve- DOE (U.S. Department of Energy). 2013. Plug-In Electric hicles: Learning from Statewide Travel Surveys.” Working Vehicle Handbook for Workplace Charging Hosts. Clean Paper UCD-ITS-WP-13-01. Institute of Transportation Cities, August. http://www.afdc.energy.gov/uploads/pub- Studies, University of California, Davis. January. lication/pev_workplace_charging_hosts.pdf. Nissan. 2013. “Nissan to Triple Electric Vehicle Fast-Chargers; DOE. 2014a. “California Laws and Incentives.” Alternative Partners with NRG Energy to Build First Fast-Charge Fuels Data Center. Last updated November 2014. http:// Network in Washington D.C. Area.” Press release, Janu- www.afdc.energy.gov/laws/state_summary?state=CA. ary 31. http://nissannews.com/en-US/nissan/usa/releases/ DOE. 2014b. “EV Everywhere Workplace Charging Chal- nissan-to-triple-electric-vehicle-fast-chargers-partners- lenge.” http://energy.gov/eere/vehicles/ev-everywhere- with-nrg-energy-to-build-first-fast-charge-network-in- workplace-charging-challenge. Accessed December 30, washington-d-c-area. 2014. Nissan. 2014. “Nissan launches programs to make LEAF DOE. 2015. “Alternative Fuels Data Center: Alternative Fu- charging free and ‘EZ.’ Press release, April 16. http:// els Station Locator.” http://www.afdc.energy.gov/loca- nissannews.com/en-US/nissan/usa/releases/nissan- tor/stations. Accessed April 8, 2015. launches-programs-to-make-leaf-charging-free-and-ez. ECOtality. 2013. “What Kind of Charging Infrastructure Do NRG/eVgo. 2014. “Electric Car Charging Plans.” http:// Chevrolet Volt Drivers in The EV Project Use?” The EV www.nrgevgo.com. Project. Lessons Learned White Paper. September. http:// PB (Parsons Brinckerhoff). 2009. “Alternative Fuels Corri- avt.inl.gov/pdf/EVProj/VoltChargingInfrastructureUs- dor Economic Feasibility Study.” Prepared for Washing- ageSep2013.pdf. ton State Department of Transportation, Office of Pub- ECOtality. 2014a. “Where Do Nissan Leaf Drivers in The lic/Private Partnerships, January 23. http://www.wsdot. EV Project Charge When They Have the Opportunity wa.gov/NR/rdonlyres/5C14E610-713A-4600-A88D- to Charge at Work?” The EV Project. Lessons Learned C567AF49D096/0/AltFuelsFinalReport.pdf. White Paper. March. http://avt.inel.gov/pdf/EVProj/Char Peterson, D. 2011. Addressing Challenges to Electric Vehi- gingLocation-WorkplaceLeafsMar2014.pdf. cle Charging in Multifamily Residential Buildings. The ECOtality. 2014b. “Where Do Chevrolet Volt Drivers in The Luskin Center for Innovation. Los Angeles, CA. June. EV Project Charge When They Have the Opportunity to http://www.luskin.ucla.edu/sites/default/files/EV_Mul- Charge at Work?” The EV Project. Lessons Learned White tifamily_Report_10_2011.pdf. Paper. March. http://avt.inel.gov/pdf/EVProj/ChargingL Peterson, S., and J. Michalek. 2013. Cost effectiveness of ocation-WorkplaceVoltsMar2014.pdf. plug-in hybrid electric vehicle battery capacity and charg- Francfort, J. 2011. “U.S. Department of Energy’s Vehicle ing infrastructure investment for reducing U.S. gasoline Technology Program: Plug-in Electric Vehicle Real World consumption. Energy Policy 52: 429-438. Data from DOE’s AVTA. Idaho National Laboratory, Recargo. 2014. “Charging Station Finder.” http://www.plug- EPRI Infrastructure Working Council, Tempe, AZ.:” De- share.com/. Accessed October 23, 2014. cember. http://mydocs.epri.com/docs/publicmeetingma Smart, J. 2014a. “PEV Infrastructure Deployment Costs and terials/1112/PDNNSH5Q3Z2/Day%201%20Plu- Drivers’ Charging Preferences in the EV Project.” SAE gIn%20Electric%20Vehicle%20Codes%20&%20Stan- 2014 Hybrid and Electric Vehicle Technologies Sympo- dards.pdf. sium, La Jolla, CA, February 11. http://avt.inl.gov/pdf/ INL (Idaho National Laboratory). 2014. EV Project Electric EVProj/SAEHybridEVSympFeb2014.pdf. Vehicle Charging Infrastructure Summary Report: Octo- Smart, J. 2014b. “Workplace Lessons Learned through the Na- ber 2013 through December 2013. INL/MIS-10-19479. tion’s Largest PEV Charging Projects.” DOE Workplace http://avt.inel.gov/pdf/EVProj/EVProjectInfrastruc- Charging Challenge Summit, Alexandria, VA, November tureQ42013.pdf. 18. http://avt.inl.gov/pdf/EVProj/WorkplaceChargingDa- IRS (Internal Revenue Service). 2014. Publication 15-B: Em- taInsights.pdf. ployer’s Tax Guide to Fringe Benefits. Washington, DC: Smart, J. 2014c. “EV Charging Infrastructure Usage in Large- Government Printing Office. Scale Charging Infrastructure Demonstrations: Public Jennings, D. 2013. “EV Charging at Lynda.com.” Presenta- Charging Station Case Studies for ARB.” Plug-in Electric tion to the Committee on Overcoming Barriers to Elec- Vehicle Infrastructure Information Gathering Meeting, tric-Vehicle Deployment, Washington, DC, August 13.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 112 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Charging Infrastructure for Plug-in Electric Vehicles 97 July 15. http://avt.inel.gov/pdf/EVProj/EVInfrastructure- Tesla. 2014. “Supercharger: The Fastest Charging Station on UsageARBJul2014.pdf. the Planet.” http://www.teslamotors.com/supercharger. Smart, J. 2014d. “Latest Insights from The EV Project and Accessed May 18, 2014. ChargePoint America PEV Infrastructure Demos.” Pre- Traut, E.J., C.C. TsuWei, C. Hendrickson, and J.J. Michalek. sentation at the GITT meeting at INL, August 12. http:// 2013. US residential charging potential for electric ve- avt.inel.gov/pdf/EVProj/GITTEVProjectAug2014.pdf. hicles. Transportation Research Part D 25: 139-145. Smart, J., and S. Schey. 2012. Battery electric vehicle driv- U.S. Census Bureau. 2011a. “2011 American Housing Sur- ing and charging behavior observed early in the EV vey for the United States, Table C-02-AH.” http://www. Project. SAE International Journal of Alternative Pow- census.gov/housing/ahs/data/national.html. ertrains 1(1): 37-33. U.S. Census Bureau. 2011b. “2011 American Housing Sur- Smart, J., and S. White. 2014. “Electric Vehicle Charging vey for the United States, Table C-01-AH.” http://www. Infrastructure Usage Observed in Large-scale Charg- census.gov/housing/ahs/data/national.html. ing Infrastructure Demonstrations.” Presentation to the Vance, A. 2013. “Gone in 90 Seconds: Tesla’s Battery-Swap- Committee on Overcoming Barriers to Electric-Vehicle ping Magic.” Bloomberg Businessweek, June 21. http:// Deployment, Irvine, CA, February 25. www.businessweek.com/articles/2013-06-21/gone-in-

Tate, E., and P. Savagian. 2009. “The CO2 Benefits of Elec- 90-seconds-teslas-battery-swapping-magic. trification E-REVs, PHEVs and Charging Scenarios.” Wald, M. 2013. “A Recharging Industry Rises.” New York SAE Technical Paper 2009-01-1311. SAE International, Times, November 12. http://www.nytimes.com/2013/11/ Warrendale, PA. 13/business/energy-environment/electric-cars-give-rise- to-a-recharging-industry.html.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 113 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Implications of Plug-in Electric Vehicles for the Electricity Sector

An important component of the ecosystem of the plug-in why electric utilities have not been allowed to take a more electric vehicle (PEV) is the electric utility, which provides proactive role in facilitating the deployment of PEVs and the the electricity that powers the vehicle.1 Electric utilities in associated charging infrastructure (Anegawa 2010). There- the twenty-first century have experienced eroding demand fore, it is important to examine the current electricity sector (see Figure 6-1) and view PEVs as a potential source of in- and consider what impediments might exist. creased demand (Kind 2013; EEI 2014). The Edison Electric Accordingly, this chapter examines potential impedi- Institute, the largest trade association for electric utilities, ments from the perspective of the individual components of contends that the industry needs increased electrification of electric utilities (the distribution, transmission, and generation the transportation sector for the electricity sector to remain components) and overall system control. To put the discus- viable and sustainable in the long term (EEI 2014). sion in context, the committee first describes the physical and An important concern raised by the public and policy economic structure of electric utilities. Physical constraints makers, however, is the ability of electric utilities to accom- in the distribution infrastructure for PEV charging are iden- modate PEV charging, a concern that impacts not only PEV tified next, followed by a discussion of potential economic owners but also the public more broadly. At the current time, constraints and impediments within the delivery system. One PEV charging requirements account for about 0.02 percent of scenario for a hypothetical utility of the future is described at the energy produced and consumed in the continental United the conclusion of the chapter. The committee’s findings and States (EIA 2012).2 Were the share of the PEV fleet to reach recommendations are provided throughout the chapter. as high as 20 percent of private vehicles, the estimated impact One important point that should be noted before begin- would still account for only 5 percent of today’s electricity ning the discussion of the electricity sector is that the federal production (DOT 2014; EIA 2012).3 Accordingly, the electric- government has only limited powers in directly influencing or ity sector does not perceive PEVs as posing any near-term or modifying the policies and behavior of the owners or opera- mid-term challenges. However, some have assumed that elec- tors of the retail electricity sector. Although the Federal En- tric utilities cannot accommodate transportation electrification ergy Regulatory Commission (FERC) maintains authority to with the current grid infrastructure. That mistaken belief is regulate transmission and wholesale sales of energy in inter- also held in other countries and has been cited as a key reason state commerce, the retail electricity sector is regulated heavily and almost entirely by individual state regulatory com- 1 An electric utility is a publicly or privately owned company that missions. Thus, the ability of private-investor-owned electric generates, transmits, and distributes electricity for sale to the public utilities to foster or impede the development of PEVs will and includes vertically integrated utilities that own their generation vary significantly based on the actions of the individual state plants, transmission components, and distribution wires and un- utility commissions. Furthermore, different regulatory bodies bundled utilities that separate the generation, transmission, distri- oversee municipal-owned utilities, federally owned utilities, bution, and retail into different businesses. Although the majority of electric utilities in the United States are privately owned, there are cooperative utilities, and, as indicated, the wholesale markets. a substantial number of generally smaller utilities that are owned These jurisdictional and regional regulatory differences limit and operated by regional organizations or municipal governments, the federal government’s ability to affect the practices of the often referred to as munis. The largest muni in the United States is U.S. electricity sector (see, for example, U.S. Court of Ap- the Los Angeles Department of Water and Power. peals 2014 decision on FERC Order 745). 2 This estimate assumes that each PEV consumes about 10 kWh/ day. Finding: State jurisdiction over retail electric rates constrains 3 This estimate assumes the aforementioned consumption for vehicle charging and that there would be 192.5 million light-duty ve- the federal role in directing the electricity sector to foster PEV hicles on the road, which is equivalent to the number in 2011 in the growth. United States (DOT 2014). 98

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 114 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Implications of Plug-in Electric Vehicles for the Electricity Sector 99

History 2011 Projections 12

10 FIGURE 6-1 U.S. electricity demand growth,

te 1950-2040. From the 7 percent annual growth 8 rates from the 1950s through the 1970s to the Ra 3-year moving average declines of the 1980s and 1990s when aver- 6 age growth in demand was about 3 percent per Trendline year, the first decade of this century has been 4 nearly flat with an average growth rate of only 0.7 percent. SOURCE: EIA (2013). 2 Annual Growth 0

-2 1950 1960 1970 1980 1990 2000 2010 2020 2030 2040

Year

FIGURE 6-2 Schematic of U.S. electric power delivery system. SOURCE: U.S.-Canada Power System Outage Taskforce (2004).

THE PHYSICAL AND ECONOMIC Today’s structure of the electricity sector and the busi- STRUCTURE OF THE ELECTRICITY SECTOR ness entities within it have been in a state of constant flux and evolution since April 1996, when the FERC issued its Figure 6-2 is a schematic of the U.S. electricity sector. Orders No. 888 and 889, which formally separated genera- Generation companies produce electricity from fossil or non- tion, transmission, and distribution from each other, thereby fossil (nuclear and renewable) sources. Transmission entities providing open access to transmission in the United States are responsible for high-voltage transmission and frequently to any generating entity and allowing for the operation of for overall system control. Distribution companies are pub- highly fluid wholesale electric markets (FERC 1996). In the licly or privately owned companies that sell, state by state, Northeast, Midwest, Southwest, Texas (ERCOT), and Cali- price-regulated electric energy to retail customers, residen- fornia, Order 888 has resulted in the creation of Independent tial, commercial, and industrial. They may be independent System Operators (ISOs); in the remainder of the country, or part of a vertically integrated electric utility. Regional Transmission Operators (RTOs) act as wide area

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 115 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

100 Overcoming Barriers to Deployment of Plug-in Electric Vehicles system operators. The ISOs and RTOs operate and control PHYSICAL CONSTRAINTS IN THE the transmission system and manage the organized whole- DISTRIBUTION INFRASTRUCTURE sale markets between generators and retail suppliers and large industrial customers. Independently owned electricity Although the introduction of PEVs is not constrained by generators operate by selling wholesale electricity into or- the transmission system or the generation capacity, the electric ganized or bilateral markets; that electricity is transmitted sector distribution infrastructure, which is a lower voltage and by separate corporate and operational entities to distribution lower capacity segment of the electric power system, could companies, which serve retail consumers. face operational constraints. PEVs are not, nor are they anticipated to be, uniformly distributed within the country or any GENERATION AND TRANSMISSION region but are instead generally expected to be locally concentrated (see Chapter 3). PEVs have typically been concentrated For roughly 60 percent of the United States, the electric- in specific geographic areas that have higher median incomes, ity sector operates through organized markets coordinated place higher values on environmental issues and energy se- by ISOs (EIA 2011). Most electric consumers in the United curity, and have higher average educational levels. Those de- States get their energy from generators within large, central- mographics suggest that PEV acquisition will be concentrated ly controlled regional networks. Their energy is transmitted in particular residential areas of the distribution system. As a over high-voltage wires that are regulated by the FERC. That result, any of the potential problems for the distribution sys- energy is finally delivered through a distribution system reg- tem noted above will most likely be localized (Maitra 2011). ulated by state public utility commissions (PUCs) that are re- Several scenarios in which problems could arise are discussed sponsible for setting the price paid per kilowatt hour. Where below. states have opted for retail competition, such as in Ohio and The first scenario in which PEVs could pose an operation- Texas, the state commissions oversee and approve the man- al constraint on the distribution infrastructure is when several ner in which the sellers of retail energy structure their ser- PEVs are simultaneously being charged on one transformer vices rather than set the price per kilowatt hour for electricity or one branch circuit that was designed to serve the traditional delivered to consumers. loads of a few residences. In that scenario, PEV charging could Understanding the electric power delivery chain is criti- affect power system stability; for example, charging could cal for understanding the current and future interactions because a voltage drop in the local distribution system or cause tween electric utilities and PEV charging systems and for voltage and current phase imbalances. Thus, the introduction identifying any impediments that might be introduced by of several PEVs could necessitate upgrades to the distribution electric utilities. As with virtually all end uses of electricity, system, such as a new transformer or a larger branch circuit that the point of contact between the electricity sector and the end would not otherwise have been needed. user is the distribution company, regardless of whether it is The charging of an individual PEV could be a challenge residential charging, public charging, or fleet charging. It is to the distribution company if that charging is coincident with at the local electricity distribution level that concentrations peak electricity consumption on any individual distribution of PEVs might stress the delivery infrastructure (Maitra system element operating at full capacity. It would be ex- 2011). However, even with high adoption rates for PEVs and tremely rare for PEV charging to coincide in time with the therefore for vehicle charging, the impact on the electricity distribution company’s peak, which typically occurs between system at large is insignificant. noon and 6 p.m. It is more likely that a PEV would be charg- Although both the generation and the transmission sec- ing at a time that coincides with the peak electricity usage of tors are critical to the ultimate delivery of electricity for ve- a residential circuit, which is typically between 5 p.m. and 11 hicle charging, they are not an impediment to PEV accep- p.m. That scenario at the residential circuit level could over- tance because meeting the demand created by PEV charging load four components of the distribution infrastructure: the is well within the planning and operational capability of the service drop (the wire from local transformer to the home or electricity sector. From the perspective of the largely competi- other point of charge), the local distribution transformer, feed- tive wholesale electricity market, any increase in demand is ers (wires from local distribution transformer to distribution welcome, particularly demand that has the potential to smooth substation), or a substation transformer. Figure 6-3 provides daily variability (a characteristic of vehicle charging). an example of hourly demand for electricity at a substation within a residential distribution system and illustrates the pat- Finding: There is no anticipated impact on either the gen- tern of residential consumption for several cases. Case 1 illus- eration or the transmission sector of the U.S. electric power trates what might happen without any incentives for off-peak system from the introduction of PEVs. Thus, the existing charging. It shows a measurable impact on the peak and indi- capability to generate and transmit power within the United cates that without incentives to reduce charging on peak, there States is not now nor is it anticipated to be a deterrent to the could be specific locations where additional capital -invest adoption of PEVs. ments might be needed to accommodate the added demand from PEV charging.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 116 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Implications of Plug-in Electric Vehicles for the Electricity Sector 101

FIGURE 6-3 Hourly demand for electricity at a substation in a residential distribution system. NOTE: A, amperes; kW, kilowatt; PHEV, plug-in hybrid electric vehicle; V, volt. SOURCE: Maitra et al. (2009). Image courtesy of Electric Power Research Institute.

From the perspective of the distribution company, PEV and scale of PEV charging that is dictated by the type of charging represents an added uncertainty for the planning trips that are taken in the vehicles. Currently, PEV charging process. There are multiple dimensions to the issue, includ- behavior exhibits a gradual load curve that peaks at about 7 ing how many PEVs will be purchased, where PEVs will be p.m., when most PEV owners arrive at home from work and charged, and whether the pattern of charging will be coinci- plug in to charge at the same time. Even then, the number of dent with local peak electricity consumption. Finally, there is upgrades at the distribution level has been minor—less than the question of whether there are state-regulator-approved ac- 0.75 percent of PEVs have required a local distribution sys- tions that the distribution company can take to alter the pattern tem to upgrade a component—and has cost ratepayers only of charging demand to minimize or potentially eliminate any $36,029 overall (CPUC 2014). negative effects, such as strong pricing incentives, timing re- Another study, by the largest California utilities (E3 strictions, or indirect or direct charging control. Research done 2014), demonstrates that even at high PEV adoption levels, in California on different pricing incentives shows that PEV the impacts on the distribution grid are minimal. The E3 study owners are price responsive, that larger price differentials used the distribution data and load patterns for the Califor- encourage customers to charge off-peak, and that customers nia utilities, analyzed the distribution of PEV adoption at the tend to remain on these time-of-day, price differentiated tariffs 9-digit zip code level, and forecast the incremental cost from (CPUC 2012a). PEVs on each individual distribution line and transformer Research also indicates that even without time-differen- until 2030 for two scenarios: a normal case that meets the tiated rates, PEV charging patterns tend to follow a pattern California zero-vehicle-emission mandate and a case that that has only moderate effects on distribution system peaks has adoption levels three times higher than the normal case. (CPUC 2014). With the near-term adoption levels antici- The study found that even for the highest adoption levels, the pated for PEVs, there is still a natural diversity in the time cost would be less than 1 percent of the annual distribution-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 117 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

102 Overcoming Barriers to Deployment of Plug-in Electric Vehicles upgrade costs of the California utilities. The study also found system upgrades that are incurred solely on their behalf or that time-of-use charging would reduce costs to customers by whether those costs can be spread over all electric customers. 60 percent compared with charging at any time during the day. The distribution company rate tariffs that are offered to Some concern has been expressed about future patterns end-use retail customers could raise obstacles to PEV adop- of charging and the resulting impact on the reliability of the tion, including (1) inconsistency between rate tariffs, (2) lack distribution system with the introduction of DC fast charging of price incentives, (3) high average costs for electricity usage (see Table 5-1). However, because the typical driving distance for residential customers, and (4) high costs for commercial for a PEV is not likely to change because of fast charging, the and industrial customers due to demand charges (see Table higher charging levels simply mean that PEVs will charge in a 6-1 for descriptions of various rate structures). These potential shorter period of time while requiring the same overall quan- obstacles can confuse retail customers about the best available tity of energy. The higher power, shorter duration charging is electricity rate and the price advantage that they might receive unlikely to have a substantial effect on the distribution infra- by using electricity as a transportation fuel. Commercial con- structure. Furthermore, data from the EV Project indicate that sumers might have the added disincentive of a demand charge DC fast charging represents only a small proportion of charg- that is triggered by increased peak load. ing for vehicles (less than 1 percent of the energy demand for The price paid by the end user for energy varies substan- the Nissan Leafs in the study) (INL 2014). tially between customer classes—industrial, commercial, and residential—and varies even more substantially from Finding: PEV charging has had a negligible effect on the region to region, state to state, and distribution company to distribution-system components to date and is expected to distribution company. State-regulated rate structures are de- have a negligible future effect at the anticipated rates of PEV signed to allow a regulated retailer to recover its fixed and adoption. variable costs and earn a fair rate of return. The costs include the variable cost of generated or purchased energy and POTENTIAL ECONOMIC CONSTRAINTS OR a return on capital invested in generation, transmission, and IMPEDIMENTS WITHIN THE DELIVERY SYSTEM distribution along with the operating costs of the company. The task of the PUCs is to allocate the full and reasonable With its existing capabilities, the generation and trans- costs of providing reliable energy across time, geography, mission elements of the U.S. electric power system are suf- and customer class. State jurisdictional authority in setting ficiently robust to provide the infrastructure and deliver the retail electricity rates has resulted in little or no consistency energy required for PEV charging. As indicated above, any in the final price of electricity in terms of both the absolute physical constraints or impediments to the distribution sys- price per kilowatt-hour of electricity and the rate structure tem will be highly localized and most likely will be only itself. Uniform change appears to be nearly impossible given within individual distribution branches in the near to mid- the fact that electric tariffs seen by all consumers (residen- term. Thus, any constraints on PEV adoption that could arise tial, commercial, and industrial) vary widely as a function of from the electricity sector are more likely to be economic the underlying energy generation structure, the tax structures rather than physical or technical. that the distribution companies face, and the vagaries of be- The economic constraints are primarily associated with ing regulated by 50 different state regulators and the local two factors: high underlying electricity costs and ineffective- regulatory bodies that oversee more than 2,000 municipal ly aligned rate structures. High underlying electricity costs and cooperative utilities. On the other hand, that same vari- reduce the financial benefit of owning a PEV by making the ability has allowed for multiple experiments in how to de- costs to drive the PEV closer to those of an ICE vehicle. sign rate structures for PEV charging. The electricity cost is most often a function of the underly- The substantial differences in electric rates from one ing characteristics of generation on a regional basis, with the utility to another and between states are impediments to hydroelectric generation of the Northwest producing much PEV adoption because it prevents a sales campaign from less expensive electricity than fossil-fuel generation of the communicating easily or simply the economic benefits and Northeast. The regional differences in electricity costs add costs of PEVs to potential buyers. Consumers have become confusion to uniform explanations of the economic operat- accustomed to translating mpg values in national advertising benefits of PEV ownership, as noted in Chapter 3. ing for ICE vehicles, recognizing that the price of gasoline A minor economic concern is the small possibility that varies by at most 10 to 20 percent across the country. Com- system upgrades could in some cases be charged directly to pare that with the variability in the residential cost of elec- the PEV-owning customers who necessitate the upgrade. If tricity between Connecticut (18.22 cents per kilowatt-hour) that cost were charged to an individual or small set of cus- and Washington State (8.7 cents per kilowatt-hour), with the tomers, it would substantially raise their costs of owning and former slightly more than double (EIA 2014). That spread operating a PEV. The handling of any cost allocation would does not account for any differentials in peak and off-peak depend on distribution company tariffs that govern wheth- rates, if they exist, or any demand charges that might be ap- er individual customers are responsible for any electricity plied. Also, it does not consider the variety of types of PEVs,

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 118 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Implications of Plug-in Electric Vehicles for the Electricity Sector 103

s ’

ared family t cross cross t -

causation for -

peak hours of peak hours of - - peak).

- pond to event

period usage. use customers comp customers use compared customers use - - - bills to bills increase, especially

’ priced, off priced, off ls ls for monitoring response. and fficiency, conservation measures, measures, conservation fficiency, - - peak off to

- .

erage peak erage av -

residences

s consumption on the total cost of the system).

’ ltifamily Unlikely to reflect cost causation fully (the impact a of customer Does not encourage behavioral changes in energy usage (e.g., switching from on Likely to increase bills low for structure. tiered or rate flat with Fixed charges might not fully reflect cost single vs multifamily (e.g., customers of classes residences). conserve. to incentives decrease Might reflectDoes different not consumption single of needs vs mu Multiple tiers are difficultmore customers for to understand. Increasedactuallyutility not usageraise monthly might cost kWh. per received. is bill after only apparent are signals Price Poorly differentiated tiers causemight significan subsidies between customer groups. Might be not simple and understandable for residential sophisticated more larger, for used (typically customers commercial and industrial customers). Likely to increase bills low for structure. tiered with Could discourage energy e customer demand response, and customer generation. customers some cause Could above those with the marginal coal is if emissions carbon increase Might generating fuelduring low charging. Enrolled customers that do res not notifications are likely to see bill increases. Customer response to RTP is enhanced by and generally requires technology contro the marginal coal is if emissions carbon increase Might generating fuelduring low charging.

...... Disadvantages

causation. - peak charging PEVs of and -

g on their consumption pattern,

ine promotes affordability for basic basic for affordability promotes ine complishes several goals: reflects economic economic reflects goals: several complishes Simple and Simple understandable. cost reflect better May costs required fixed customer per the Reflects basis. per month on customer each serve to Basel needs. encourage to perceived are rates tier Higher conservation. tiers are relatively simpleTwo understand to structure. rate complex a with more compared causation. cost reflect better Might economic reflects goals: several Accomplishes value (marginal cost) of energy, encourages leads and use, peak reduces and conservation to economically efficient decision making. Dependin customers couldor see bills.higher lower Encourages off solar commercial or residential of use greater photovoltaics. utility. the for reduction peak direct Provides Enrolled customers who respond to event notifications will see bills. lower Ac value (marginal cost) of energy, encourages leads and use, peak reduces and conservation to economically efficient decision making. Simple and Simple understandable.

...... Advantages

period CPP, - maximum power

’

vice. providing electricity; peak demand periods. -

xample, Tier 1, electricity usage s monthly maximum power usage ’ t customers to meet peak and peak and partial

term price increase to a predetermined a predetermined to price increase term - s off of cumulativeof customer electricity usage a in ’

time system conditions. In a fixed a In conditions. system time

- time system conditions warrant a price adjustment. adjustment. a price warrant conditions system time -

time pricingtime (RTP) critical and peak (CPP). pricing -

kW basis for a customerkW basis for - $0.18/kWh). wn until real away from peak hours on a CPP event day. CPP event days are time system conditions. - n, transmission,n, and public purpose costs. program sts are highest, they send more accurate price signals to customers. to signals price accurate send more they highest, sts are Lower rates during a utility Higher rates during seasonal and daily peak periods.demand incremental when period, peak the during higher are rates TOU Because co

. . . Relevant Definition apply would kWh; it per cents in volumetrically charged rate average An to all usage (e.g., customers all to applicable (e.g., $5/month) charges are charges monthly Fixed regardless of usage; they are intended to reflect costs that do not ser change of availability constant ensure to with necessary are and usage the during usage electricity based charges on kWh per are charges Volumetric cyclebilling (e.g., $0.15/kWh); they are reflect to intended costschange that with usage (e.g., variable generation charges) and typically include generation,distributio as a function changes that rate A monthly bill cycle. The tiers generally are defined from a baseline quantitymonthly orminimum. Prices an “inverted in tier” “inclining or block” rate increase e For increases. usage electricity as cumulative the to baseline amount;up 2, electricity Tier 101 usage 130 percent to of from 4, Tier baseline; of percent 200 131 to from usage 3, electricity Tier baseline; baseline. of percent 200 than greater usage electricity a per on Calculated of cost the reflect to calculated generally are charges Demand $5/kW). (e.g., transmission and distribution facilities buil demands. Demand charges are in addition to volumetric energy charges schedules rate on those than are lower charges (per energy volumetric the but kWh), demandwithout charges. used. is that it day of time or the season to according electricity prices that rate A of cost the actual reflects closely more design TOU rate A by characterized is it dynamicA rate that allows a short real reflect to levels) (or level are days the but predetermined, are increase price the of duration and time the predetermined.not programs CPP provide customers participating an incentive to shift usage generally called 24 hours in advance. or an hour (typically notice short at adjusted be to prices allows rate dynamic A or timing the or price the Either conditions. system of as a function ahead) a day are both unkno Examples include real RTP allows prices to be adjusted frequently, typically on an hourly basis, toreflect real

Definitions, Advantages, and Disadvantages of of Rates Electric Various Types

1 -

Use

- of - TABLE 6 TABLE Type Rate of Flat rate Fixed charge and volumetric charge rates Tiered charges Demand Time (TOU) rate Critical Peak Pricing (CPP) Dynamic rate CPUC (2012b). from data on Based SOURCE:

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 119 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

104 Overcoming Barriers to Deployment of Plug-in Electric Vehicles which include BEVs that run only on electricity and PHEVs generally midday, time periods. Figure 6-4 shows an example that can run on a gasoline or electricity, and whose mix of of the impact of TOU rates on charging behavior as reported those fuels will vary by battery capacity and driving needs. in the EV Project (ECOtality 2013). The time during which Assembling a broad message for consumers on costs and a vehicle was connected to a residential charger and the time benefits is practically impossible given that fuel costs vary, during which the vehicle was actually drawing power were on average, by a factor of at least two and can vary by a examined in the service territories of the Nashville Electric factor of 4 or more.4,5 The difficulty in generalizing fueling Service (NES) and Pacific Gas and Electric (PG&E). NES costs is discussed further in Chapter 3. does not offer TOU rates, but PG&E does. Figure 6-4 shows Residential electric rate structures for vehicle charg- that while the vehicles were connected to residential chargers ing can also be an impediment to PEV adoption. Flat rates for similar times in the two service areas, demand for charging provide no incentive for the owner to charge the vehicle at energy was very different in the service area with TOU pric- the optimal time for the utility. Given that flat rates represent ing, PG&E. That finding indicates that user behavior in plug- averages over a broad customer base, if the PEV is used for ging in the vehicle is the same for both regions but that TOU commuting and thus is charged at night, for example, the flat pricing motivates customers to use the timers integrated with rate is likely to be high relative to the distribution company’s the vehicle or charger to control their charging time and mini- actual marginal cost of supplying electricity at that time and mize their cost. Given that PEV charging at residential sites at that location within the distribution system. The incentives most often is discretionary, in that it can occur any time after provided by time-of-use (TOU) rates are substantially better the vehicle returns home and before it is needed the next day, aligned with the true costs of serving electric customers but PEV owners can take advantage of time-differentiated rates to add to the distribution company’s cost if digital, multiregister charge their vehicles during the least costly period, benefiting meters are not already installed at the home. Although time- both the owner and the utility. differentiated rates generally benefit PEV owners, they can be TOU pricing has been in place at Tokyo Electric Power a disincentive if owners need to charge during high-priced, Company (TEPCO) for over 20 years (TEPCO, personal communication, December 10, 2013). TEPCO says that the 4 The estimates conservatively assume that TOU or RTP rates company has not needed to add any new generating capacity have only twice the variability seen in average rates. in over 20 years in large part because its rate structures send 5 Assuming that an ICE vehicle gets 30 mpg on $3.50 per gallon gasoline and travels an average of 11,500 miles per year, the net the appropriate price signals to customers, who in turn have savings per year for a PEV owner are $1,169 if electric costs are responded by conserving electricity during peak periods. $0.05/kWh, $997 if electric costs are $0.10 /kWh, and only $824 if Distribution rates for commercial and industrial cus- electric costs are $0.15/kWh at 300 Wh per mile. tomers typically contain demand charges. The economic ef-

(c) Energy Demand, NES, no TOU rates (d) Energy Demand, PG&E, TOU rates FIGURE 6-4 Residential charging behavior in NES and PG&E service territories, as measured in the EV Project. Panels (a) and (b) show average percent of vehicles plugged into residential chargers by time of day in the NES and PG&E service territories, and panels (c) and (d) show average charging energy demand by time of day in those territories. NOTE: NES, Nashville Electric Service; PG&E, Pacific Gas and Electric; TOU, time of use. SOURCE: ECOtality (2013).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 120 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Implications of Plug-in Electric Vehicles for the Electricity Sector 105 fect on commercial or industrial customers that are provid- allow these potential revenues to be claimed by PEVs. The ex- ing charging could be substantial and strongly negative if a isting operating and accounting logics implemented by PUCs single hour with unusually high charging demand were to and ISOs that allow customers to provide these services will cause an increase in the demand charge. Although it might need to be modified to accommodate PEVs, which are mobile be argued that one or more charging stations would represent loads that will be connecting at multiple and diverse locations only marginal increases in energy consumption for relatively as opposed to most (if not all) other distributed technologies large commercial entities, to the extent that a charging sta- that operate at a fixed location. Among those actions, the two tion is being used during the peak power consumption time most important are deciding which entity in the PEV ecosys- of day, it will have an impact on the maximum demand of tem should be compensated for the service provided and how the commercial or industrial entity. Exceeding the demand to measure compliance with any dispatch instruction given by threshold by any amount will increase the total cost of en- the electric utility or ISO. ergy to the facility and, in some cases, will hold the demand- rate charges at the higher level for many months to more Finding: The confusion caused by the substantial differenc- than a year. A study done by the EV Project demonstrates the es in electric rates offered to customers by different utilities importance of this issue; it found that demand charges could or states can be an impediment to PEV adoption. account for over 90 percent of the utility bill in some areas (ECOtality 2012). Thus, it is critical to note that although the Finding: TOU rate charging could provide a win-win situa- peak occurs only once and only for a brief period, the effect tion as the PEV owner pays for charging at a lower rate and the on the customer’s bill could be felt for far longer, and more utility benefits from moving the load from peak to off-peak. important, the increased cost could outweigh any potential Recommendation: To ensure that adopters of PEVs have in- benefits gained by providing PEV charging infrastructure. centives to charge vehicles at times when the cost of supply- There exists one additional impediment to PEVs that is ing energy is low, the federal government should propose that directly related to the rate structure but difficult to quantify. state regulatory commissions offer PEV owners the option of PEVs individually and in combination with other technolo- purchasing electricity under TOU or real-time pricing. gies likely to be implemented in the distribution system (such as distributed storage, distributed generation, and advanced ELECTRICITY SECTOR REGULATORY ISSUES controls) might be able to provide a benefit to the utility in FOR OPERATING A PUBLIC CHARGING STATION terms of ancillary services, such as regulation or reserves. The supply of those necessary services to the utility has a positive As noted in Chapter 5, utilities might have a viable busi- value in terms of cost savings—costs that the utility would ness case for deploying charging infrastructure. Provision of have had to expend but for the fact that the PEV or other dis- PEV charging services can benefit electric utilities as it can tributed device exists and is able to operate so as to benefit the increase utilization of fixed assets of the distribution - infra utility. The ancillary service benefits are real, even if difficult structure, potentially lowering rates, increasing revenues, or to separate from the benefits of other technologies in the dis- both. As noted, the provision of public charging might also tribution system. The fact that PEVs and other technologies encourage the adoption of PEVs, which could provide broad in the system can and do provide those services provides a customer or societal benefits from reduced greenhouse gas positive benefit to the operations of the utility and could rep- emissions or improved local air quality. However, one caveat resent a financial benefit.6 Although there is some difficulty in that needs to be recognized is that not all utilities are allowed precisely quantifying the potential benefit, using the regula- to provide charging services. Some states have granted partial tory framework that exists in California would provide about or full permission for electric utilities to provide PEV charg- $100 per kW per year of capability and could be an important ing services. That action has allowed Austin Energy (Texas), incentive for PEVs if passed on to PEV customers (E3 2014).7 Duke Energy (North Carolina), and Portland General Electric Regulatory structures implemented by PUCs and ISOs could (Oregon) to fill the need for PEV charging services. In Japan, TEPCO is allowed to support the deployment of public charg- 6 It has been suggested that benefits should be (and within most of the ISOs are) paid for based on the “avoided cost” of the utility. The ing infrastructure by providing necessary interconnection to difficulty is in calculating the avoided cost and therefore the size of the grid and internalizing the costs to the shareholders or rate- any benefits. Avoided costs represent a calculation of what it would payers, and this approach has meaningfully reduced the cost have cost the utility to acquire the service provided by the distributed to install DC fast chargers in TEPCO’s service territory (An- technology if the utility had to provide it. A further difficulty in esti- egawa 2010). mating this benefit is that the service is likely to be provided by multiple technologies within the distribution system, requiring sharing of Thus, many in the PEV and utility industries have called any avoided-cost benefits, and that far more of the service could be for greater latitude to provide charging services, particularly delivered than the utility requires at any point in time. in underserved markets where demand for PEV charging ex- 7 This estimate is the net present value over 10 years for a resource ists (C2ES 2012). Independent public charging providers, that would be available for the 100 peak hours of a year, assuming however, have concerns about policies that would allow elec- that the cost of a new entry has a weighted average cost of capital of tric utilities to provide charging services and believe that the an independent power producer.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 121 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

106 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 6-5 States that have regulations regarding who can own or operate a PEV charging station. NOTE: PEV, plug-in electric vehicle. SOURCE: Based on data from C2ES (2015). Courtesy of the Center for Climate and Energy Solutions.

utilities would have an unfair competitive advantage. Such activity,8 and in a few states, the issue has been addressed by policies could put independent public charging providers at the legislature rather than by regulatory interpretation (see a competitive disadvantage because utilities have substantial Figure 6-5).9 existing infrastructure and would be able to spread some of the cost of providing charging services across their customer Finding: Electric utilities that provide PEV charging services base independent of whether any individual customer owned have multiple reasons for doing so that can positively affect a PEV or used the public charging infrastructure. utility ratepayers and the utilities themselves. Another regulatory issue is the extent to which PEV- charging providers are considered to be offering electricity for Recommendation: As a means of encouraging consistency resale and thus would be regulated as a utility. As discussed between jurisdictions, the federal government should propose above, in most states, the retail sale of electricity is a com- that state regulatory commissions decide that public charg- mercial activity heavily regulated as a monopoly business. ing stations are not utilities and therefore not subject to utility Considering an independent public charging company to be regulatory oversight, specifically in setting rates for charging. a public utility and subject to public utility regulation would dramatically alter the company’s cost structure and its po- 8 For example, see California PUC Rulemaking 09-08-009, Code tential competitive position. In addition, it would affect the Section 740.2.14, July 2011; Arizona Corporation Commission company’s ability to raise capital. Many states have not yet Docket No. E-01345A-10-0123, Decision No. 72582, September made a distinction between the retail sale of power and the 15, 2011; and PUC of Oregon Guidelines Adopted; Utilities Or- provision of PEV charging services (Council of State Govern- dered To Make Revised Tariff Filings, January 19, 2012. 9 For example, see Washington Substitute House Bill 1571, 62nd ments 2013). However, a few state PUCs have taken up the Legislature, 2011 Regular Session, July 22, 2011; California As- issue of whether PEV charging services should be a regulated sembly Bill 631; Colorado General Assembly House Bill 12-1258; and New York Bill S5110-2013.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 122 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Implications of Plug-in Electric Vehicles for the Electricity Sector 107 Recommendation: Given that electric utilities and their rate- REFERENCES payers could benefit from increased PEV adoption, electric utility regulators should encourage their electric utilities to Anegawa, T. 2010. Needs of Public Charging Infrastructure provide PEV charging services to their customers when con- and Strategy of Deployment. TEPCO. Kanagawa, Japan. ditions indicate that all customers benefit. CPUC (California Public Utilities Commission). 2012a. Load Research Report and Pricing Study. Alternative Fueled THE UTILITY OF THE FUTURE Vehicles Order Instituting Rulemaking R.09-08-009. San Francisco, CA. December 28. The interactions of the electricity sector and PEV charg- CPUC. 2012b. “Workshop Presentation: Rate Design Elements, ing are not static or unidirectional. Internationally and now Concepts and Definitions.” Residential Rate Structure Or- increasingly in the United States, the most significant changes der Instituting Rulemaking R.12-06-013. December 5. in delivery of electricity are taking place on the customer’s CPUC. 2014. Load Research Report. Alternative Fueled Ve- premises or inside the meter, where the customer has more hicles Order Instituting Rulemaking R.09-08-009. San control than the utility.10 Such changes include programmable Francisco, CA. January 31. thermostats and smart appliances. There are also many chang- Council of State Governments. 2013. State Utilities Law es occurring within the distribution system, including the in- and Electric Vehicle Charging Stations. http://knowl- troduction of micro-grids; the increased deployment of dis- edgecenter.csg.org/kc/sites/default/files/Electric%20 tributed electricity generation in the form of small-scale solar Vehicle%20Charging%20Stations_0.pdf. photovoltaics (PV) and wind; the development of distributed C2ES (Center for Climate and Energy Solutions). 2012. An storage, including second-life PEV batteries; and advanced in- Action Plan to Integrate Plug-in Electric Vehicles with formation technology and control.11 the U.S. Electrical Grid. Report of the PEV Dialogue Consideration of PEVs in the future distribution system Group. March. http://www.c2es.org/initiatives/pev/action is an integral part of the ongoing planning that is focused on -plan-report. the utility of the future. PEV demand for charging energy will C2ES. 2015. Who Can Own/Operate a Charging Station? affect the total demand for energy at the distribution level. http://www.c2es.org/initiatives/pev/maps/who-can-own- The increase in demand might well be offset by an increase operate-a-charging-station. Accessed March 24, 2015. in supply from distributed generation. Combining residential DOT (U.S. Department of Transportation). 2014. “National PV with the multiple possible functions of a PEV as a distrib- Transportation Statistics.” http://www.rita.dot.gov/bts/ uted storage device and means of transportation is also seen sites/rita.dot.gov.bts/files/publications/national_trans- as a means of localized load balancing for the utility and cost portation_statistics/index.html. savings for the customer.12 In the future, increased informa- E3 (Energy and Environmental Economics). 2014. Califor- tion and communication technology combined with real-time nia Transportation Electrification Assessment Phase 2: economic price signals are anticipated to allow PEV battery Grid Impacts, October 23. http://www.caletc.com/wp- systems to become distributed storage systems capable of pro- content/uploads/2014/10/CalETC_TEA_Phase_2_Fi- viding energy and ancillary services to the distribution utility. nal_10-23-14.pdf. Termed, variously, smart charging, vehicle-to-home, and ve- ECOtality. 2012. Lessons Learned—The EV Project White hicle-to-grid, this capability will give the distribution-system Paper Syllabus. Prepared for the U.S. Department of operator added flexibility and control to manage the overall Energy Award #DE-EE0002194. http://www.theevpro- load on the system. ject.com/downloads/documents/Syllabus%20Rev%20 4.0.pdf. Finding: PEVs might be a large part of the utility of the fu- ECOtality. 2013. “How Do PEV Owners Respond to Time- ture and could help perform functions that the electric sector of-Use Rates while Charging Vehicles?” http://www.the deems valuable. However, issues associated with customer ac- evproject.com/cms-assets/documents/125348-714937. cess to their vehicles and effects on battery life will need to be pev-driver.pdf. resolved before vehicles can be fully integrated into the utility EEI (Edison Electric Institute). 2014. Transportation Electri- of the future. fication: Utility Fleets Leading the Charge. Washington, DC. June. EIA (U.S. Energy Information Administration). 2011. 10 The customer’s electric meter is where the “fence” is typically “About 60% of the U.S. Electric Power Supply Is Man- drawn, with the distribution company unable to see what happens aged by RTOs.” Today in Energy, April 4. http://www. on the customer premise beyond its meter. eia.gov/todayinenergy/detail.cfm?id=790. 11 The MIT Energy Initiative study The Utility of the Future rep- EIA. 2012. “Table A1. Total Energy Supply and Disposition resents one research effort under way to understand the impact of Demand” and “Table A2. Energy Consumption by Sector disruptive technologies on the utility distribution system. 12 It should be noted that PEV batteries beyond their useful life for and Source.” Annual Energy Outlook 2013 Early Release transportation might be useable as stationary storage devices within Overview. December 5. http://www.eia.gov/forecasts/aeo/ the distribution system (see Chapter 4). er/index.cfm. Accessed March 14, 2013.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 123 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

108 Overcoming Barriers to Deployment of Plug-in Electric Vehicles EIA. 2013. Annual Energy Outlook 2013 Market Trends— Maitra, A. 2011 Potential Impacts of Vehicles on the Grid. Electricity, May. Figure 75. http://www.eia.gov/forecasts/ Electric Power Research Institute. Palo Alto, CA. August. aeo/pdf/0383%282013%29.pdf. Maitra, A., J. Taylor, D. Brooks, M. Alexander, and M. Du- EIA. 2014. Electric Power Monthly with Data for December. vall. 2009. Integrating Plug-in Electric Vehicles with February. http://www.eia.gov/electricity/monthly/current the Distribution System. 20th International Conference _year/february2014.pdf. on Electricity Distribution, Prague, June 8-11. Electric FERC (Federal Energy Regulatory Commission). 1996. Pro- Power Research Institute. Palo Alto, CA. moting Wholesale Competition Through Open Access U.S.-Canada Power System Outage Taskforce. 2004. Final Non-discriminatory Transmission Services by Public Report on the August 14, 2003 Blackout in the United Utilities; Recovery of Stranded Costs by Public Utili- States and Canada: Causes and Recommendations. April. ties and Transmitting Utilities. Order No. 888. Issued http://energy.gov/sites/prod/files/oeprod/Documentsand- April 24. Media/BlackoutFinal-Web.pdf. Accessed April 23, 2013. INL (Idaho National Laboratory). 2014. EV Project Elec- U.S. Court of Appeals. 2014. Electric Power Supply Associ- tric Vehicle Charging Infrastructure Summary Report: ation, Petitioner v. Federal Energy Regulatory Commis- October 2013 through December 2013, INL/MIS-10- sion, Respondent and Madison Gas and Electric Compa- 19479. http://avt.inel.gov/pdf/EVProj/EVProjectInfra- ny et al., Intervenors. No. 11-1486. May 23. http://www. structureQ42013.pdf. cadc.uscourts.gov/internet/opinions.nsf/DE531DBFA7 Kind, P. 2013. Disruptive Challenges: Financial Implications DE1ABE85257CE1004F4C53/$file/11-1486-1494281. and Strategic Responses to a Changing Retail Electric pdf. Business. Edison Electric Institute. January.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 124 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles

One of the most important issues concerning the deploy- effective purchase price (price adjusted for any financial in- ment of plug-in electric vehicles (PEVs) is determining what, centives) plus the costs of fueling, maintaining, and insuring 1 if any, incentives are needed to promote this deployment. the vehicle, minus the resale or trade-in value. Although the Determining appropriate incentives is difficult because little distinction between price and total cost of ownership is im- is yet known about the effectiveness of PEV incentive pro- portant, most consumers find it difficult to estimate the latter grams. Therefore, the committee first considered the price or with much confidence, partly because they are not certain at cost competiveness of PEVs and the possibilities for reduc- the time of purchase how long they will keep their vehicles ing production costs. It next considered manufacturer and or how many miles they will drive in them and partly be- consumer incentives for purchasing or owning PEVs and cause fuel costs, maintenance costs, and resale values are then past incentive programs for other alternative-vehicle uncertain, particularly for newer technology vehicles like and fuel technologies. The chapter concludes with recom- PEVs. That said, tools to help prospective buyers estimate mendations on what the committee sees as the most compel- and compare the total ownership costs of different vehicles ling approaches to promoting PEV deployment. are now available at Edmunds.com and elsewhere. The U.S. Department of Energy (DOE) has created a VEHICLE PRICE AND COST OF OWNERSHIP website calculator where prospective buyers provide various driver-specific inputs (such as net vehicle price, normal daily A major consideration when purchasing or leasing a driving distance, annual mileage, and breakdown of mileage vehicle is the financial consequence. Thus, the two factors between city and highway) and can calculate the cumulative to consider are the vehicle price and the cost of ownership. cost of ownership over different time horizons under repre- Most people compare prices or monthly leasing payments sentative assumptions about other such factors as mainte- 2 and take into account any financial incentive that would re- nance and insurance costs (DOE 2014a). Cumulative cost duce the vehicle price. Because vehicle manufacturers and of ownership is distinct from total cost of ownership in that dealers are profit-oriented businesses, vehicle prices are it is a calculation for a given time horizon and typically does generally related to production costs. However, the relation- not include the trade-in or resale value. Thus, by focusing on ship between a manufacturer’s suggested retail price and its the cumulative costs of ownership over different time hori- production cost typically reflects a number of considerations zons, the DOE calculations avoid assumptions about resale and might differ across vehicles and over time. For example, values, which are highly uncertain for PEVs. on newly developed vehicles, such as PEVs, vehicle manu- Affordability—as reflected in price and total cost of facturers might be motivated to incur losses or relatively low ownership—is not the only consideration that influences the profit margins in the short run to promote sales and strength- types of vehicles that consumers choose. As emphasized in en their business positions and profit margins over the long Chapter 3, PEV adoption also depends importantly on con- run. That type of marketing strategy contributed to the even- sumer awareness, the variety of models available, uncertain- tual success of the Toyota Prius (Tellis 2013). Regardless of ties about new technologies and resale values, and various how the vehicle price is set, price is an important consider- vehicle attributes that determine its utility to the customer. ation for most consumers when shopping for a new vehicle. Because there is still much uncertainty about PEV technolo- Some prospective buyers also consider more broadly the costs of owning a vehicle—in particular, the costs of fu- 1 More sophisticated definitions of the total cost of ownership are eling, maintaining, and insuring the vehicle and its resale or based on the present discounted values of the various components trade-in value. The total cost of ownership (or overall cost to of cost (that is, they discount future costs relative to current costs). the consumer) of any specific vehicle can be viewed as the 2 The website calculator provides zip-code-specific assumptions about fuel prices that the user can override. 109

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 125 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

110 Overcoming Barriers to Deployment of Plug-in Electric Vehicles gies and the battery lifetimes, and because there are not as ciated with their sources of energy and drive trains.4 The table yet well-developed markets for used PEVs or their batteries, also includes the prices of the Toyota Prius, the Volkswagen consumers are likely to perceive more uncertainty about the Passat, and the Nissan Leaf, along with the average transac- total costs of owning PEVs than about the total costs of own- tion prices for small and midsize vehicle segments, including ing conventional vehicles. Those uncertainties make risk- the prices for the specialty segment. The committee empha- averse consumers less likely to purchase a PEV, other things sizes that the MSRPs shown in the table simply represent price being equal. They also strengthen the incentive to lease a points that manufacturers target, as distinct from the prices at PEV rather than purchase one. which vehicles are actually sold, which are typically less than Thus, leasing is a more frequent choice for PEVs than the MSRPs. Nevertheless, the MSRP comparisons provide for conventional vehicles (see Table 3-3) because it can make some useful perspectives. The average transaction price data monthly payments for the vehicle appear more affordable and reflect the actual prices paid by the consumer. reduce the risk of owning one. In a typical leasing arrange- In addition to providing information on vehicle prices, ment, ownership of the vehicle is transferred at a negotiated Table 7-1 includes ranges for annual fuel costs and 5-year cu- sales price from the dealer to a bank or some other finance mulative costs of ownership. As reflected in the table notes, company (often the financial arm of the vehicle manufactur- the estimates are based on a combination of the assump- er). In general, the sales price is influenced importantly by the tions included in the DOE calculator and the committee’s amount that the vehicle manufacturer charges the dealer for specific assumptions about annual vehicle miles traveled, the car and by any incentive payments or guarantees that the electricity costs, and gasoline prices ranging from $2.50 to manufacturer offers the finance company.3 The finance com- $4.00 per gallon. The estimates of 5-year cumulative costs of pany collects the monthly leasing payments and generally also ownership assume that purchasers pay MSRPs and receive receives (1) incentives or other subsidies provided by federal maximum tax credits, and the estimated costs in years two or state governments or (2) benefits from lower negotiated through five have not been discounted.5 sales prices resulting from government incentives provided to As indicated in Table 7-1, the MSRPs before consider- manufacturers or dealers. As such, potential PEV drivers who ation of the federal tax credits for the PEVs are all substan- do not have enough income to qualify for the federal income tially higher than the MSRPs for the HEVs and ICE vehicles tax credit could still benefit from any credit available with ve- listed in the table. After consideration of the $7,500 federal hicle leasing. tax credit, the adjusted MSRP for the Chevrolet Volt still substantially exceeds the MSRPs for the Chevrolet Cruze LS PRICE AND COST COMPETITIVENESS Automatic, the Toyota Prius, and the Volkswagen Passat and OF PLUG-IN ELECTRIC VEHICLES the average transaction price for the specialty small vehicle segment. It somewhat exceeds the MSRP of the Chevrolet The committee used three approaches to assess the ex- Cruze Diesel Automatic. The MSRP for the Ford Fusion En- tent to which prices or costs of ownership might have af- ergi, after adjusting for the $4,007 federal tax credit, exceeds fected PEV deployment to date. First, the committee com- MSRPs of the two other Fusion models, the Toyota Prius, and pared the manufacturers’ suggested retail prices (MSRPs, the Volkswagen Passat but is similar to the average transac- which are essentially the target prices) of various PEV mod- tion price of the specialty midsize vehicle segment. The 5-year els with those of comparative vehicles. Second, it evaluated cumulative cost of owning a Chevrolet Volt—as estimated by sales data, and, third, it considered consumer surveys. No the DOE calculator using representative assumptions—is, re- approach provided conclusive results. spectively, about $2,300 and $3,400 higher than the compa- Table 7-1 lists MSRPs for three relatively best-selling rable 5-year costs for the Toyota Prius and the Volkswagen PEV models, for several hybrid electric vehicles (HEVs), and Passat at a gasoline price of $2.50 per gallon and higher by for internal-combustion engine (ICE) vehicles. It excludes the Tesla Model S, which tends to be bought by relatively wealthy 4 It is common practice for a vehicle manufacturer to build mul- individuals whose purchase decisions might not be highly sen- tiple vehicle models on the same platform but that are in different sitive to price. The table allows one to compare the prices of market segments. Although the Chevrolet Cruze and Volt are in the the Chevrolet Volt and two Chevrolet Cruze models and to same size segment, they are not in the same market segment (stan- compare the prices of the Ford Fusion Energi and the Ford dard compact vs premium compact). 5 Fusion Automatic and Hybrid models; the comparisons are In theory, car payments that are spread over 5 years or less have present discounted values that equal, or closely approximate, the informative because the PEVs and comparative ICE and HEV vehicle MSRP when the payments are discounted at the rate of in- models are built on the same platforms and therefore have terest charged in financing the car payments. And while discount similar production costs for components other than those asso- rates between 0 and 6 percent would imply that the present discounted value of the 5-year stream of other costs (for fuel, tires, maintenance, insurance, inspection, and registration) was up to 3 An important influence on the negotiated sales price is the fi- several thousand dollars less than five times the annual average of nance company’s estimate of the value of the car at the end of the those costs—with relatively smaller differences for vehicles that lease, taking into account any guarantees by the vehicle manufac- have relatively lower annual fuel costs (such as PEVs)—the quali- turer. tative comparisons and the finding would not be affected.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 126 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 111

d,e

Year Cumulative Cumulative Year - Ownership 41,038 38,025 44,410 46,353 41,598 43,895 40,033 40,210 ------

Range for 5 Cost of 40,564 35,280 41,935 45,153 38,589 41,855 34,860 38,284 37,195 Excludes any Excludes cost of financing

ies, the credit is set at $2,500 plus $417 times the times $417 plus $2,500 at set is credit the ies,

1,605 1,608 - - engine;MSRP, manufacturers’ suggested retail price; 636 1,464 1,485 897 1,088 932 ------

al Revenue Code Section 30D (IRS 2009). (IRS 30D Section Code Revenue al Range Annual of Fuel Cost 540 915 990 657 1,002 680 435 581 1,005

c Passat, which is as of August 31, 2014. 31, August of as is which Passat, combustion combustion - ser pays MSRP and receives maximum tax credit.

MSRP Less Federal Tax Credit Tax Federal Less MSRP 26,685 19,530 25,810 30,693 22,400 27,280 21,510 24,200 20,995 in Vehicles Electric and Vehicles Comparative (dollars) - $4.50 per gallon for diesel fuel, $0.125 per kWh for electricity, and DOE assumptions about about assumptions DOE and electricity, for kWh per $0.125 fuel, diesel for gallon per $4.50 -

MSRP 34,185 19,530 25,810 34,700 22,400 27,280 29,010 24,200 20,995

f cylinder engines. - wheel drive; HEV, hybrid electric vehicle; ICE, internal -

$4.00 per gallon for gasoline, $3.00 gasoline, for gallon per $4.00 -

kWh, kWh, up to a of maximum $7,500. The subsidy extends to the first 200,000 PEVs sold by each Those who purchasers manufacturer.

5 Vehicle Type PHEV ICE vehicle ICE vehicle PHEV ICE vehicle HEV BEV HEV ICE vehicle Average Transactions Price 20,374 23,129 25,677 29,759 eeds charger installation and permitting requirements. Assumes purcha -

Year Year Cumulative Cost of Ownership for Selected Plug -

miles. 11,500 driven is vehicle when registration and licensing, insurance, maintenance, tires, for costs annual $2,235 of te

MSRPs MSRPs and5

inhybrid electric vehicle; S, speed.

- 1 -

Prius Hybrid

a Based on 11,500 vehicle miles traveled, $2.50 traveled, miles vehicle on 11,500 Based The federal tax credit is $2,500 for PEVs that have battery capacities below 5 kWh. For PEVs that have larger battery capacit Includes DOE estima All HEV and ICE vehicle models are ones with four with ones are models vehicle ICE HEV and All Volkswagen for except 2014, 31, July of as websites manufacturers’ from models 2014 for data MSRP on Based Data provided by Baum & Associates (A. Baum, personal communication, April 22, 2014). 22, April communication, personal Baum, (A. Associates & Baum by provided Data

TABLE 7 TABLE Model Chevrolet Volt Chevrolet Cruze LS Automatic (6S) Chevrolet Cruze Diesel Automatic (6S) Energi Fusion Ford Ford Fusion Automatic Ford Fusion Hybrid (FWD) Leaf Nissan Toyota Volkswagen Passat (6S) Alternative Comparators Small vehicles (average) Small vehicles (specialty) Midsize vehicles (average) Midsize vehicles (specialty) a b c amount that the battery Intern capacitySee credit. $7,500 exc full the of advantage take cannot taxes income federal in $7,500 least at pay otherwise not would d use. daily normal e and costs any associated with home f front FWD, vehicle; electric BEV, battery NOTE: PHEV, plug

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 127 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

112 Overcoming Barriers to Deployment of Plug-in Electric Vehicles about $1,000 and $800 at a gasoline price of $4.00 per gallon. Finding: Under the current program of federal tax credits, The 5-year cumulative cost of owning a Ford Fusion Energi the comparisons of MSRPs and cumulative ownership costs is higher than those of all the HEV and ICE vehicles listed in provide mixed evidence on whether price is currently an ob- Table 7-1, even at a gasoline price of $4.00 per gallon. stacle to the deployment of PEVs. However, in the absence In contrast, the MSRP of the Nissan Leaf, after consider- of the tax credits or other subsidies, analogous comparisons ing the $7,500 federal tax credit, is less than the MSRP of at prevailing MSRPs would be unfavorable to the PEVs. the Toyota Prius and is comparable to the average transaction price of the small vehicle segment. The 5-year cumula- Finding: Sales data and consumer survey data are difficult tive ownership cost of the Nissan Leaf, as estimated using the to interpret. They are consistent, however, with the view that DOE calculator, is, respectively, about $3,400 and $2,300 less price is a barrier to some buyers, but that others might be than the analogous costs of the Toyota Prius and the Volkswa- rejecting PEVs for other reasons. gen Passat at a gasoline price of $2.50 per gallon and about $5,200 and $5,400 less at a gasoline price of $4.00 per gallon. POSSIBILITIES FOR DECLINES IN PRODUCTION A second approach to assessing the price-competitive- COSTS FOR PLUG-IN ELECTRIC VEHICLES ness of PEVs is to evaluate data on sales volumes. The idea is that low sales volume could indicate that the vehicles are The extent to which PEVs are adopted over time will de- not price- or cost-competitive, although the analysis above pend on reductions in their production costs, on the policies indicates that the Nissan Leaf is highly competitive given that governments implement to promote PEV deployment, the federal tax credit. As noted in Chapter 1, about 290,000 and on the extent to which vehicle manufacturers decide to highway-capable PEVs were sold in the United States by price PEVs more attractively by relying on relatively low the close of 2014. Although there is no generally accepted markups in pricing and perhaps compensating for the loss wisdom on how rapidly sales of a new product line should of revenue by raising mark-ups on their portfolios of other expand during the early introductory stage, the number of vehicles.7 The three factors are not completely independent. PEVs sold in the United States to date has fallen short of Government policies toward research and development can aspirational goals,6 despite substantial incentives. affect battery costs, and policies, such as zero-emission re- Consumer survey data provide yet another approach quirements, can induce vehicle manufacturers to change for evaluating whether price is currently an obstacle to PEV their pricing strategies. This section focuses on likely reduc- deployment. The annual New Vehicle Experience Studies tions over time in the production costs of PEVs. conducted by Strategic Vision have surveyed large samples In general, the costs of producing PEVs will be driven of new-vehicle buyers and distinguished between those who down over time by a number of factors. The technologies be- actively looked at PEVs but chose not to purchase one, and ing used for PEVs are relatively new compared with technolo- those who did not even consider purchasing a PEV. In the gies used to produce ICE vehicles, which have been evolving former group, 41 percent indicated that current prices or re- and improving for more than a century. Thus, as discussed in bates were appealing and 27 percent said that current interest Chapter 2, it is expected that research and development will or lease rates were appealing; in the latter group, 25 percent lead to reductions in the costs of PEV batteries over time considered them appealing and 17 percent said that current through technological improvements, such as higher energy interest or lease rates were appealing (Edwards 2013). The densities, improved designs, and longer battery lives.8 survey data also showed that 43 percent and 45 percent of BEV and PHEV buyers, respectively, considered their pur- 7 Manufacturers typically estimate their direct labor and material chases “value for the money.” The data suggest that current costs of production and markup prices above direct production costs pricing might not be a primary barrier to greater PEV sales by amounts sufficient to cover the fixed costs of plant and equip- among early adopters, who tend to be less price sensitive ment, various indirect costs (including the costs of research and development, corporate operations, dealer support, and marketing), than the mainstream market, and that buyers are rejecting the and an allowance for profits. A study based on data from the 2007 vehicles for other reasons. Strategic Vision also found that annual reports of eight major vehicle manufacturers found that the 10 percent of new-vehicle buyers are actively shopping or average markup factor for the automobile industry was about 1.5 plan to shop for a PEV and that about 33 percent are open to (RTI/UMTRI 2009). However, manufacturers will sell vehicles at hearing more about what PEVs can do for them. Given that prices that the market will bear; thus, the markup on one vehicle model can be much greater than that on another vehicle model. 15 million new vehicles are sold each year, the data suggests 8 Innovations in other elements of vehicle technology are likely to an active potential market of about 1.5 million PEVs with lead to improved vehicle performance without necessarily gener- about two-thirds finding the pricing or lease rates appealing. ating substantial reductions in cost. Improving the aerodynamics, Despite these results, Strategic Vision stated that it still be- reducing friction, reducing the rolling resistance of tires, and reduc- lieves that “price is a critical barrier—even for more affluent ing weight could lead to a vehicle design with, for example, better customers” (Edwards 2013, p. 45). performance or more driving range, depending on what trade-offs were made in the overall vehicle design. That could lead to the need for a smaller battery, and the reduced cost of the battery would have 6 For example, in early 2011 DOE projected cumulative U.S. sales to be weighed against the increased cost of the above-mentioned of 1.22 million PEVs by 2015 (DOE 2011). improvements because such improvements generally come at

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 128 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 113 Cost reduction might also be realized as PEV production to fuel and perhaps less expensive to maintain. Others might volume increases and both the capital costs of investments in value certain of their attributes that are not present (or not production facilities and the indirect costs of research and de- present to the same degree) on conventional vehicles, such as velopment, corporate operations, dealer support, and market- the smooth and quiet ride, better acceleration, the convenience ing are spread over more vehicles. Such scale economies will of home fueling, less maintenance (fewer or no oil changes), also be realized in the supplier base with increases in the de- and the potential for reducing vehicle emissions and petro- mand for components, such as batteries, motors, and inverters, leum usage. enabling suppliers to reduce the prices that vehicle manufacturers are charged for those components. In addition, increases Finding: Although battery costs could decline by 50 or per- in the number of vehicles and the demand for components haps even 75 percent over the next decade, it is not clear will likely lead to greater competition among suppliers, which whether such a decline would be sufficient—by itself—to en- could intensify the downward pressure on component prices sure widespread adoption of PEVs once the current quotas for as suppliers innovate and generate better designs. And once federal tax credits are exhausted. a new vehicle technology becomes fairly firmly established, components tend to become more standardized, leading to ad- Finding: The decline over time in PEV production costs is ditional reductions in production costs. likely to occur gradually, and existing quotas for federal tax As emphasized in Chapter 2, the difference between credits might be exhausted for manufacturers of relatively the costs of producing PEVs and comparative conventional popular PEVs before costs can be substantially reduced. vehicles can be largely attributed to the high cost of high- energy batteries. Accordingly, the prospect for large-scale INCENTIVES deployment of PEVs depends importantly on how much battery costs decline over time. Other things equal, if bat- The production and purchase of PEVs is a classic chick- tery pack costs declined by as much as 50 percent over the en-and-egg problem. Manufacturers do not want to produce next 5 to 10 years, consistent with optimistic projections (see PEVs if no customers exist, and consumers cannot buy PEVs discussion in Chapter 2), the cost of producing a BEV with if vehicles are not available that meet their expectations. 24 kWh nominal battery capacity (analogous to the Nissan Therefore, regulatory requirements and incentives for manu- Leaf) would decline by roughly $6,000. Similarly, the costs facturers and consumers have been provided over the past few of producing PHEVs with 16.5 kWh and 7.6 kWh nominal years by states and the federal government to encourage PEV battery capacities (analogous to the Chevrolet Volt and Ford production and deployment. Most manufacturer incentives Energi) would decline by about $4,100 and $1,900. And a and mandates are contained in federal or state regulatory pro- 75 percent decline in battery pack costs (a highly optimistic grams discussed below. Most consumer incentive programs forecast)—to as low as $125 per kWh of nominal battery described below have involved purchase incentives, although capacity—would reduce the costs of producing the Nissan some have included ownership and use incentives. There have Leaf, the Chevrolet Volt, and the Ford Energi by an addi- also been incentives to install charging stations, the availabil- tional $3,000, $2,050, and $950, respectively. Such optimis- ity of which might also influence people’s willingness to pur- tic reductions in production costs would provide opportuni- chase PEVs. ties for the vehicle manufacturers to reduce the MSRPs for PEVs by amounts that largely offset, or more than offset, the Manufacturer Incentives and Regulatory Requirements effects of the pending expiration of the current program of federal tax credits. Incentives for manufacturers to produce PEVs are A detailed analysis of how the nonbattery costs of PEVs contained in the federal Corporate Average Fuel Economy are likely to evolve relative to those of comparative vehicles (CAFE) Standards and the Greenhouse Gas (GHG) Emis- is beyond the scope of this report. Recent NRC reports on sion Standards for light-duty vehicles. California and other the costs of different vehicle types, however, conclude that states have Zero-Emission-Vehicle (ZEV) programs that re- the costs of producing PHEVs and BEVs will likely remain quire the sale of PEVs in those states because PEVs are the greater than the costs of producing ICE vehicles and HEVs only qualifying technology that are currently mass produced. for at least the next two decades (NAS/NAE/NRC 2009; These manufacturer incentives and regulatory requirements NRC 2013b). can have the effect of reducing the vehicle price of PEVs It should be noted that PEV adoption does not require relative to other vehicles; they are reviewed in detail below. PEVs to be priced at or below the prices of conventional vehicles. Some might buy PEVs because they are less expensive Federal Regulatory Incentives for Plug-in Electric Vehicles some incremental cost (NRC 2011a, 2013a). These innovations in technology will likely also be applied to conventional vehicles as Fuel economy and GHG emissions from light-duty ve- manufacturers strive to meet fuel-economy requirements, but also hicles are regulated under the federal CAFE-GHG national at some incremental cost.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 129 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

114 Overcoming Barriers to Deployment of Plug-in Electric Vehicles program. Under the recently updated rule, vehicle manufac- PHEVs are treated as dual-fuel vehicles that use both turers must comply with fuel economy and GHG standards electricity and gasoline. Federal regulations stipulate how the that are equivalent to about 54.5 mpg and 163 grams of gasoline and electric energy consumption is measured in cer-

carbon dioxide (CO2) per mile for the fleet average of new tification test cycles for PHEVs. The measured electric energy vehicles by model year (MY) 2025 (EPA/NHTSA 2012a). consumption is converted, as in the BEV case, to an MPGe Although GHG emissions from light-duty vehicles are regu- by using the petroleum equivalency factor method. The elec- lated by the U.S. Environmental Protection Agency (EPA) tric MPGe and gasoline mpg must be weighted to obtain the under the Clean Air Act and fuel economy is regulated by fuel economy value used in CAFE compliance. Until 2019, the National Highway Traffic and Safety Administration PHEVs are assumed to use electric fuel 50 percent of the time (NHTSA) under the CAFE program, the federal agencies and gasoline 50 percent of the time (EPA/NHTSA 2012b). Be- have developed a single national program.9 The standards ginning in MY 2020, the weighting will be determined on the are fleet-based standards, meaning that a manufacturer can basis of the SAE J1711 fuel economy test method that uses build vehicles that are certified above and below the stan- a utility factor to estimate the fraction of driving on electric- dards as long as the fleet-wide average meets the standards. ity and assumes that the vehicle owner charges once per day The standards also offer an array of regulatory flexibilities, and drives in much the same way as today’s typical light-duty including the ability to bank or buy compliance credits and vehicle drivers. Given that method, a PHEV with 20-mile all- incentives for various types of technologies. Although the electric range (PHEV20) would be treated as a 90 mpg gaso- analyses done by EPA and NHTSA for their most recent reg- line-powered car, and a PHEV with 60-mile all-electric range ulation for 2017-2025 developed a cost-effective compliance (PHEV60) would be treated as a 226 mpg gasoline-powered demonstration pathway that shows how the standards for ve- car (Al-Alawi and Bradley 2014). hicles in 2025 can be achieved almost exclusively through The EPA GHG standards provide two temporary incen- conventional gasoline-powered vehicles, the regulations de- tives to vehicle manufacturers to produce PEVs. The first in- veloped by EPA and NHTSA have generous credits for PEVs centive is temporary treatment of PEVs as zero emissions (that that make it attractive for vehicle manufacturers to produce is, upstream emissions of power plants are ignored) for the PEVs. However, the very nature of the separate legislative portion of operation assumed to be powered by electricity. For authorities under which EPA and NHTSA operate to regulate BEVs and fuel-cell vehicles (FCVs), the MY 2017-2021 GHG light-duty vehicles means that the manner of crediting manu- standards set a value of 0 g/mile for the tailpipe CO2 emissions facturers of alternative-fuel vehicles, such as PEVs, diverges compliance value (EPA/NHTSA 2012b). PHEVs also receive between the CAFE and GHG standards. a value of 0 g/mile based on a formula to estimate the fraction The CAFE standard focuses on reducing petroleum us- of electricity usage. For MY 2022-2025, the program allows age in the United States. Federal law requires the CAFE pro- the 0 g/mile treatment up to a cumulative sales cap for each gram to evaluate PEVs and all other alternative-fuel vehicles manufacturer.11 After that cap is reached, the compliance val- by using a petroleum equivalency factor (PEF) to calculate ues for BEVs and the electric portion of PHEVs are based on a fuel economy compliance number for such vehicles.10 an estimate of the national average emissions associated with PEFs are used to convert the electric energy consumption producing the electricity needed to charge PEVs. However, measured in the certification test cycle of alternative-fuel the cumulative sales caps appear to be generous, so it is pos- vehicles, including BEVs and PHEVs, to an equivalent fuel sible that most PEVs will be treated as ZEVs. economy number. Manufacturers use the miles per gallon The second manufacturer incentive under the EPA GHG equivalent (MPGe) to calculate their fleet average mpg for standards is sales multipliers that effectively treat a single compliance purposes. In compliance with the law, only 15 PEV sold as more than one vehicle for compliance purposes. percent of the alternative fuel (such as electricity) that is con- The PEV sales multipliers start at 2.0 in MY 2017 for BEVs sumed during the test is counted toward the fuel economy and FCVs and 1.6 for PHEVs and then gradually decline to rating of an alternative-fuel vehicle. That treatment provides 1.0 by MY 2022, when they are proposed to be completely a strong incentive for manufacturers to produce alternative- phased out. The larger multiplier in the earlier years rewards fuel vehicles to comply with CAFE program requirements. manufacturers that are early market leaders. Allowing each For example, a BEV that is rated on the certification test PEV to count as more than one vehicle lowers the average cycle at 230 Wh/mile (roughly equivalent to the certification GHG per mile for a manufacturer. test cycle value for a Nissan Leaf) is treated as equivalent to By increasing the MPGe and decreasing the grams CO2 a 357 mpg gasoline-powered car (see Box 7-1). per mile of PEVs, the federal incentives from the PEF, zero

11 Manufacturers that sell 300,000 PHEVs, BEVs, and FCVs be- 9 The largest source of GHG emissions from light-duty vehicles tween 2019 and 2021 can use the 0 g/mile value for a maximum is CO2 that results from the combustion of gasoline or diesel fuel, of 600,000 vehicles starting in 2022. For all other manufacturers and this implies that fuel economy and GHG emissions are directly (those who sell less than the 300,000), the 0 g/mile value can be correlated, necessitating the development of a common set of stan- used only up to 200,000 vehicles. After the sales cap is reached, dards. emissions will be calculated using an upstream emission standard 10 49 U.S.C. 32904(a)(2)(B) and 49 U.S.C. 32905(a). calculated by EPA.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 130 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 115

BOX 7-1 Derivation of Petroleum Equivalent for a Battery Electric Vehicle

The PEF is derived by first calculating a full fuel cycle, gasoline-equivalent energy content of electricity and then dividing it by a 0.15 fuel-content factor. The PEF was developed to motivate the production of vehicles fueled with 85 percent ethanol (E85), and the 0.15

factor reflects the petroleum consumption of E85 vehicles. The gasoline-equivalent energy content of electricity (Eg) is calculated as follows:

Eg = gasoline-equivalent energy content of electricity = (Tg × Tt × C) / Tp where

Tg = U.S. average fossil-fuel electricity generation efficiency = 0.328

Tt = U.S. average electricity transmission and distribution efficiency = 0.924

Tp = petroleum refining and distribution efficiency = 0.830 C = watt-hours of energy per gallon of gasoline conversion factor = 33,705 Wh/gal Therefore,

Eg = (0.328 × 0.924 × 33,705)/0.830 =12,307 Wh/gal

The Nissan Leaf, which requires 230 Wh/mile, exhibits a range of 53.5 miles on the electric-energy equivalent of 1 gallon of gasoline (12,307/230 = 53.5). That is the certification test-cycle result. To provide an incentive for alternative-fuel vehicles, only 15 percent of the fuel consumed in the test cycle is counted, and the resulting MPGe for the Nissan Leaf is 12,307/(230 × 0.15) = 357 mpg for CAFE purposes.

emissions treatment, and sales multipliers allow the manu- other states to opt into the California standard. Nine states— facturers to produce higher emitting and less fuel-efficient Connecticut, Maine, Maryland, Massachusetts, New Jersey, gasoline-vehicle fleets and still meet their fleet average stan- New York, Rhode Island, Vermont, and Oregon—have adopt- dards. The incentives, therefore, create an internal cross sub- ed the California ZEV program as authorized under Section sidy that allows a manufacturer to reduce the cost of com- 177 of the Clean Air Act as part of their plans to meet federal pliance for their gasoline-vehicle fleet by producing PEVs. ambient air quality standards. The nine states and California Furthermore, because credits can be traded between manu- account for 28 percent of total U.S. light-duty vehicle sales. facturers, such companies as Tesla that produce excess CAFE Recently, eight states that have the ZEV program signed a and GHG credits can sell their credits to other manufacturers joint memorandum of agreement to cooperate on developing (Energy Independence and Security Act 2007). The value of policies to accelerate PEV deployment in their states (State the PEV credits under EPA and NHTSA regulations is dif- ZEV Programs 2013). The agreement includes the develop- ficult to estimate but might be about a few thousand dollars ment of a Multistate ZEV Action Plan that describes state ac- per vehicle based on the costs of regulatory compliance in tions to promote PEV deployment and recommends research the absence of PEVs (EPA/NHTSA 2012b). The California and stakeholder partnerships to support long-term develop- and state ZEV programs (see below) also generate credits for ment of the PEV market (ZEV Program Implementation Task various attributes, and the value has been estimated at up to Force 2014). Additionally, many of the ZEV states participate $35,000 per vehicle for the Tesla Model S, which generates in the Northeast Electric Vehicle Network, which works to up to seven credits per vehicle sold (Ohnsman 2013). promote PEV deployment. For a manufacturer to receive credit for vehicle sales State Zero-Emission-Vehicle Programs under the ZEV program, a vehicle must be categorized under the program as either a ZEV (a BEV or an FCV) as defined The California ZEV program provides an important by the program or a PHEV with an all-electric range of great- manufacturer requirement for PEVs. The Clean Air Act er than or equal to 10 miles (CARB 2013). CARB estimates Amendments of 1969 authorized California to develop more that the total number of ZEVs (BEVs and FCVs) and PHEVs stringent tailpipe standards than the rest of the country, and needed to comply for MYs 2018 through 2025 for Califor- the California Air Resources Board (CARB) used the author- nia and the nine other states is about 228,000 in 2018 and ity provided in Section 209b to adopt the original ZEV pro- 725,000 by 2025 (Keddie 2013). The ZEV Program Imple- gram in 1990. The ZEV program is a part of the state’s com- mentation Task Force (2014) predicts that by 2025, a little prehensive plan to meet federal and state ambient air quality more than 15 percent of new vehicles sold in participating standards. Later amendments to the Clean Air Act allowed states will be either ZEVs or PHEVs.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 131 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

116 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Finding: By creating an internal cross subsidy, existing consumers who lease, the leasing company typically claims federal and state regulatory programs for fuel economy and the credit and reflects the credit in the monthly lease rate, so emissions (CAFE, GHG, and ZEV) have been effective at leasers essentially see the benefit of the tax credit at the point stimulating manufacturers to produce some PEVs. The sale of sale. Although the PEV tax credits are analogous to the of credits from these programs between manufacturers has HEV and diesel-vehicle tax credits in the 1990s and 2000s, also provided an important incentive for PEV manufacturers which have since expired, the notable differences are that the to price PEVs more attractively. tax credit for most PEVs is much higher than the HEV and diesel credits. Because more people lease PEVs than pur- Finding: Because the ZEV program mandates sales of a cer- chase them, a higher fraction of PEV drivers see the benefits tain percentage of PEVs, its impact could be larger than the of the credit sooner; therefore, the effect of the PEV tax cred- incentives under the federal CAFE-GHG national program. its could be greater than the effect of the now expired HEV and diesel credits. However, a recent study found that 94.5 Consumer Incentives percent of survey respondents (adult drivers from the general public in 21 major U.S. cities) were not aware of PEV incen- The U.S. federal, state, and local governments have all tives and suggests that the effectiveness of the PEV credits experimented with consumer incentives to encourage PEV de- could be enhanced through greater consumer awareness and ployment. Many other countries have also used various policy education (Krause et al. 2013). tools to encourage consumer adoption. Box 7-2 defines the The U.S. state governments have offered a variety of fi- different types of financial incentives that have been used, and nancial incentives (see Table 7-2). The DOE Alternative Fuels Table 7-2 summarizes the various financial and nonfinancial Data Center maintains a database that provides a comprehen- incentives and the entities that have used them. The incen- sive listing of state incentives.13 Several states have offered tives that have been used to promote PEV deployment are purchase incentives in the form of tax credits in addition to the discussed below. one offered by the federal government. The monetary amount The committee defined four categories of financial -in varies from state to state; for example, as of August 2014, centives. Purchase incentives are one-time financial benefits available tax credits ranged from $605 in Utah to up to $6,000 earned by purchase of a PEV and include tax credits, tax de- in Colorado. The tax credits have also varied over time; many ductions, tax exemptions, and rebates. Ownership incentives have been reduced or recently expired. The method for cal- are recurring annual or periodic financial benefits that accrue culating the credit varies from state to state; some states sim- to PEV owners regardless of use and include exemptions ply calculate it on the basis of purchase price, and others use from (or reductions in) registration taxes or fees, weight sur- battery capacity and purchase price to determine the amount. charges, environmental taxes, or vehicle inspections. Use in- Several states have also used sales-tax exemptions or rebates centives are ongoing financial benefits realized by driving a to make the effect of the purchase incentive more immediate PEV and include exemptions from motor fuel taxes, reduced for those who choose to buy rather than lease. Some of these roadway taxes or tolls, and discounted or free PEV charg- purchase incentives are restricted to certain types of PEVs. ing or parking. PEV infrastructure incentives are one-time For example, the sales-tax exemptions in Washington and financial benefits for deploying PEV charging stations and New Jersey are restricted to BEVs, and the rebate in Illinois include tax credits, rebates, or other subsidies. A variety of is restricted to BEVs and range-extended PHEVs. California, these incentives have been used throughout the United States however, provides rebates to BEVs and PHEVs, although the and in other countries. Educating consumers on all the in- amount differs ($2,500 for BEVs and $1,500 for PHEVs in centives is challenging, and some confusion results because 2014). Using rebates and sales-tax exemptions is consistent incentives vary by location and often come and go without with recent research that compares the effectiveness of HEV much warning. tax credits, sales-tax exemptions, and rebates and finds that The primary consumer incentive offered by the U.S. the sales-tax exemptions and rebates appear to be more ef- federal government is a purchase incentive in the form of fective than tax credits possibly because of their immediacy, a tax credit. The tax credit amount varies depending on the transparency, and simplicity (Chandra et al. 2010; Gallagher capacity of the battery in the vehicles and will be phased and Muehlegger 2011). out at the beginning of the second calendar quarter after the State governments have also used ownership and use manufacturer produces 200,000 eligible PEVs as counted incentives to promote PEV deployment. The most common from January 1, 2010.12 To claim the credit, consumers who have been exemptions from registration fees or vehicle in- purchase a PEV must have sufficient tax liability and will spections and reduced roadway taxes or tolls. Local govern- not see the benefit until they file an annual tax return. For ments have also offered discounted or free PEV charging or parking. States have also provided financial incentives for 12 The federal tax credit is $2,500 for PEVs that have battery ca- installing PEV charging stations so that consumers will be pacities below 5 kWh. For PEVs that have larger battery capacities, the credit is set at $2,500 plus $417 times the amount that the bat- 13 See U.S. Department of Energy, “State Laws and Incentives,” tery capacity exceeds 5 kWh, up to a maximum of $7,500. http://www.afdc.energy.gov/laws/state.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 132 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 117

BOX 7-2 Financial Incentives

Tax credits and tax deductions are taken at the end of the tax reporting period. They act to lower the final year-end taxes owed to federal or state governments. Tax credits are considered more desirable because they directly offset a taxpayer’s liability in the exact amount of the credit. For example, if the end-of-year tax liability for a person was $18,000, a tax credit of $5,000 would directly lower the total taxes owed by that same amount, to $13,000.

In contrast, tax deductions reduce the amount of reported income that is subject to taxation, rather than directly offsetting taxes owed. If persons had taxable income of $60,000 (taxed at 25 percent), they would owe $15,000. If they took a $5,000 tax deduction, their taxable income would be reduced by that amount, to $55,000, which in turn would lower their tax liability by $1,250 to $13,750. Tax deductions are often subject to rules that limit the amounts that can be deducted or that restrict higher-income taxpayers from taking the full deduction.

As financial incentives, manytax credits are available to all persons who file a tax return whereastax deductions are available only to those persons who file a tax return that itemizes deductions. Studies show that in the United States fewer than 50 percent of all federal tax returns claim itemized deductions (Prante 2007).

Tax exemptions are recognized at the time of a transaction (for example, at the point of sale) or during a regular tax reporting period (for example, vehicle registration renewal process). By exempting an entire asset or activity from taxation, the financial benefits are often realized immediately, such as with a sales tax exemption on a vehicle purchase. Tax exemptions are not usually subject to income-based qualifications or limitations, as is the case with many tax deductions.

Rebates provided by the government can take several forms, depending on their structuring. The key distinguishing feature of a rebate is that it is earned (and often processed) at the time of a qualifying purchase. Some rebate programs require an individual to submit proof of a qualifying purchase directly to the government to receive a rebate check; other rebates are provided to the seller of qualifying goods or services so that the total purchase price to the consumer can be reduced in an equal amount. However structured, both consumers and sellers tend to prefer rebates over tax credits, deductions, or exemptions because the financial benefits are immediately realized at the time of the purchase transaction, regardless of tax rates and method of tax filing.

A fee-bate is a method of taxing or applying a surcharge or fee on certain activities or classes of assets that are deemed to have un- desirable social attributes to generate sufficient revenue to provide direct rebates for other activities or assets that are deemed to be more desirable. Because this section is more narrowly focused on the types of financial incentives that can be provided rather than the method of funding those incentives, a fee-bate system and rebates are treated in the same way because they both result in a rebate.

A subsidy is a more general term used to describe methods for government-provided financial assistance. A subsidy can take the common form of tax credits, deductions, exemptions, or rebates; or, a subsidy can include direct government grants, lower than market rate loans, loan guarantees, or myriad other ways for government to provide financial support.

sure that they will be able to charge their vehicle away from nually, and receive free parking, which is worth $5,000. They home. The most common and popular nonfinancial incentive are also permitted to drive in bus lanes and have access to offered by the states has been access to restricted lanes, such free public charging at over 450 locations in Oslo (Doyle and as bus-only, high-occupancy-vehicle, and high-occupancy- Adomaitis 2013). Another example is the Netherlands, which toll lanes. That incentive has been used by several states over had financial incentives that equaled as much as 85 percent the years to promote adoption of PEVs (and HEVs). of the vehicle price, although these have been reduced. It is Other countries have used incentives similar to those important to note that the financial incentives in the Nether- used by the United States, as summarized in Table 7-2. The lands are particularly important because electricity prices are most popular have been purchase incentives in the form of tax so high that the consumer’s incentive to use electricity as a exemptions or rebates, ownership incentives in the form of fuel is small. exemptions or reductions in registration or ownership taxes or One interesting purchase rebate program is the one of- fees, and use incentives in the form of reduced roadway taxes fered by the Clean Energy Vehicle Promotion Program in or tolls. Some of the financial incentives have been substan- Japan. It is notable because it has a clear sunset, the rebate tial. For example, Norway offers substantial tax breaks (no level declines every year on the basis of a preset formula, purchase tax, no annual registration tax, and no value-added and the rebate amount financed by the government depends tax) that amount to about $11,000 over the vehicle lifetime, on whether vehicle manufacturers meet a preset annual price or about $1,400 per year (Doyle and Adomaitis 2013). Com- target (see Figure 7-1). The administering agency, the Minis- muters also do not pay road tolls, which are worth $1,400 an- try of Economy, Trade and Industry (METI 2013) calculates

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 133 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

118 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

TABLE 7-2 Incentives for Plug-in Electric Vehicles (PEVs) by Country and State Type of Incentive Location Financial Incentives Purchase Incentives—one-time financial benefit earned by purchase of PEV Tax credits or deductions (realized only on U.S. federal government filing tax return) United States: Colorado, Georgia, Louisiana, South Carolina, Utah Other countries: Australia, Austria, Belgium, Israel Tax exemptions or rebates (realized at the United States: California, District of Columbia, Illinois, Massachusetts, point of sale) New Jersey, Pennsylvania, Texas, Washington Other countries: Canada (Ontario [for leased vehicles], British Columbia [purchased or leased], and Quebec [leased]), China, Estonia, France, Iceland, Ireland, Israel, Japan, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, U.K. Ownership Incentives—recurring annual or periodic financial benefit that accrues to PEV owners, regardless of use Exemption from or reduction in registration or United States: Arizona, Connecticut, District of Columbia, Illinois, Maryland ownership taxes or fees Other countries: Australia (Victoria), Austria, Belgium (Flanders), Denmark, Finland, Germany, Greece, Ireland, Italy, Japan Latvia, the Netherlands, Norway, Romania, Sweden, Switzerland (varies by region), U.K. Exemption from or reduction in weight surcharges United States: Colorado (collected annually at time of registration or renewal) Other countries: Japan Exemption from environmental taxes Other countries: Denmark Exemption from vehicle inspection United States: Arizona, California, Colorado, Connecticut, Delaware, District of Columbia, Georgia, Idaho, Illinois, Maryland, Massachusetts, Missouri, Nevada, New Jersey, New Mexico, New York, Oregon, Pennsylvania, Tennessee, Utah, Virginia, Washington, Wisconsin Use Incentives—on-going financial benefits realized by driving a PEV Exemption from motor fuel taxes United States: North Carolina, Utah, Wisconsin Other countries: European Union, Japan, Norway Reduced roadway taxes or tolls United States: California, New Jersey, New York Other countries: Austria, Czech Republic, Japan, the Netherlands, Norway, Switzerland (varies by region), U.K. Discounted or free PEV charging United States: Arizona, California, Delaware, Georgia, Illinois, Indiana, Kentucky, Maryland, Michigan, Minnesota, Nevada, Virginia Other countries: Japan, Norway Discounted or free PEV parking United States: Free parking available at some airports, such as Long Beach Airport; at parking garages in some states and localities, such as Nevada, Sacramento, and Santa Monica; and other locations, often with free charging Other countries: Denmark, Iceland, Norway PEV Infrastructure Incentives—one-time financial benefit for deploying PEV charging stations Tax credit or rebate for installing PEV charging station United States (individual and business): Arizona, California, Florida, Georgia, Illinois, Indiana, Louisiana, Maryland, Michigan, New York, Oklahoma, Oregon PEV charging infrastructure deployment subsidies United States (individual and business): California, Colorado, Connecticut, District of Columbia, Florida, Illinois, Maryland, Massachusetts, Nebraska, New Mexico, Ohio, Texas, Utah, Washington Other countries: Canada, European Union, Israel, Japan, Korea, Norway Nonfinancial Incentives

Use Incentives—on-going special privileges granted to PEV drivers Access to restricted lanes, such as bus-only, high- United States: Arizona, California, District of Columbia, Florida, Georgia, Hawaii, occupancy-vehicle, and high-occupancy-toll lanes Maryland, New York, North Carolina, Tennessee, Utah, Virginia Other countries: the Netherlands, Norway Reserved parking for PEVs United States: Arizona, California, Hawaii, Washington SOURCES: Based on data from Gallagher and Steenblick (2013); Brand et al. (2013); Beltramello (2012); Morrow et al. (2010); Tesla (2013); DOE (2014b); Doyle and Adomaitis (2013); EV Norway (2014); Mock and Yang (2014); IEA (2013); Jin et al. (2014).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 134 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 119 an annual price target by assuming a linear decline between to their immediacy and ease of transaction. The federal cash- a base price in 2012 and a long-term target price in 2016. To for-clunkers program, for example, offered a purchase rebate encourage vehicle manufacturers to reduce their sales prices and resulted in strong consumer response to the immediate every year, the government provides 100 percent of the re- subsidy (Huang 2010). bate if the manufacturer meets the annual target price but subsidizes only about 67 percent of the rebate if the manu- Finding: Given the research on fiscal incentives and HEVs, facturer exceeds the annual price target. the effectiveness of the federal income tax credit to motivate It is difficult to draw firm conclusions about the expe- consumers to purchase PEVs would be enhanced by convert- riences with incentive programs in other countries given the ing it into a rebate at the point of sale. cultural, political, and geographical differences. However, countries with substantial financial incentives for PEVs, Finding: The U.S. state and local governments offer a va- such as Norway and the Netherlands, have seen a high rate riety of financial and nonfinancial incentives; there appears, of PEV adoption. Those with little or no financial incentives however, to be a lack of research to indicate which incentives for PEVs—most notably Germany, which has not offered con- might be the most effective at encouraging PEV deployment. sumer incentives and has relied on demonstration programs in four major regions—have experienced minimal sales. Finan- Finding: The many state incentives that differ in monetary cial incentives, however, are not working everywhere, most value, restrictions, and calculation methods make it challeng- notably in China, where there has been tepid consumer uptake ing to educate consumers on the incentives that are available despite the substantial financial incentives offered. One early to them and emphasize the need for a clear, up-to-date source analysis of that puzzling situation concludes that Chinese con- of information for consumers. sumers are more concerned about vehicle performance than cost at this stage (Zhang et al. 2013). Further information on Finding: Overall, the experience worldwide demonstrates the international experience is provided in Appendix C. that substantial financial incentives are effective at motivating There has been little academic research about the ef- consumers to adopt PEVs. fectiveness of fiscal incentives in stimulating the adoption of PEVs. However, a greater body of evidence now exists re- PRICE OF CONVENTIONAL garding fiscal incentives and HEVs. Overall, that literature TRANSPORTATION FUELS AS AN suggests that financial incentives do motivate consumers INCENTIVE OR A DISINCENTIVE FOR THE to purchase more fuel-efficient vehicles (see, for example, ADOPTION OF PLUG-IN ELECTRIC VEHICLES Huang 2010; Sallee 2011; Ozaki and Sevastyanova 2011). In general, it also seems that the more immediate the incentive, High gasoline prices motivate consumers to drive less the more effective it is at persuading consumers to purchase and to purchase a more fuel-efficient vehicle, at least for the more fuel-efficient vehicle. Sales-tax exemptions -or re some time after prices rise noticeably. As noted by Diamond ductions and rebates at the state level have been associated (2009, p. 982), much more strongly with consumer adoption, presumably due

FIGURE 7-1 Japan’s clean energy vehicles promotion program. If a PEV’s price exceeds the dashed black line, the government subsidizes two-thirds of the difference. If a PEV’s price is below the dashed black line, the government subsidizes 100 percent of the difference. SOURCE: Based on data from METI (2013).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 135 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

120 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

FIGURE 7-2 U.S. HEV and PEV sales overlaid with U.S. gasoline prices. NOTE: HEV, hybrid electric vehicle; PEV, plug-in electric vehicle. SOURCE: ANL (2014).

Gasoline prices serve as the most visible signal for con- of PEVs. It is important to note, however, that the carbon taxes sumers to think about fuel savings and fuel economy, so applied by a few countries (Denmark, Norway, Sweden, and it is reasonable that relatively minor variations in gasoline Switzerland) and by the Canadian province of British Colum- prices could lead to significant changes in adoption pat- bia on transportation fuels do not strongly affect the prices terns, particularly for people in the market for a new car of petroleum fuels because the carbon content of gasoline as gas prices rise or fall. and diesel fuels is much less than that of coal. California’s low-carbon fuel standard, which imposed a compliance cost Rapid increases in gasoline prices that increased adop- of $13 per ton CO2 emissions, was assessed by Yeh and Wit- tion rates for HEVs provide some support for that assertion cover (2012) and was found to add one-tenth of a penny per (see Figure 7-2). By the same token, low gasoline prices create gallon to the cost of gasoline in 2012. Thus, although carbon a disincentive for PEV adoption. They reduce the savings in taxes and cap-and-trade regimes might be the most effective fuel costs that a consumer would realize by owning a PEV and methods for reducing GHG emissions, they might not provide could make the cumulative cost of PEV ownership appear less a meaningful incentive to purchase a PEV. attractive than the same cost of a conventional ICE vehicle. As of January 2015, U.S. gasoline prices were less than PAST INCENTIVES ON OTHER half of those in most European countries, including Bel- ALTERNATIVE VEHICLES AND FUELS gium, France, Germany, Italy, the Netherlands, and the U.K. (EIA 2015). The higher gasoline prices in Europe and Asia Over the past few decades, a number of federal and state are mostly due to considerably higher gasoline taxes, which policy initiatives have been implemented to stimulate the de- more than double the price of gasoline per gallon. Accord- ployment of alternative vehicles and fuels. Air quality, cli- ingly, numerous studies in the United States and elsewhere mate change, and energy security concerns have motivated have concluded that taxes on conventional transportation fuels the initiatives. The primary alternative vehicles and fuels that that substantially raise the gasoline price create an incentive have been considered in the light-duty fleet include HEVs, for consumers to purchase more fuel-efficient vehicles and to PEVs, and hydrogen FCVs, and methanol, ethanol, natural drive fewer conventional-vehicle miles (Diamond 2009; Mor- gas, propane, and biodiesel for use as fuels in conventional row et al. 2010; Small 2012; Burke and Nishitateno 2013). ICE vehicles.14 Key laws and regulations that are aimed di- Broader market-based policies like carbon taxes and cap- and-trade regimes theoretically could create a disincentive for 14 For the purposes of this report, the focus will be primarily on the use of conventional vehicles and an incentive for the use the lessons learned from alternative vehicles and fuels in light-duty vehicle applications.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 136 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 121 rectly at alternative fuels or that provide incentives for alter- ertson and Beard 2004; Hwang 2009; NRC 2010b, 2013a; native-fuel vehicles include the Alternative Motor Fuels Act Greene 2012). of 1988, the Clean Air Act of 1990, the Energy Policy Act of 1992, the California ZEV program (originally adopted in Methanol 1990), the Renewable Fuel Standard (part of the Energy Pol- icy Act of 2005 and the Energy Independence and Security There was great interest in methanol (MeOH) as a poten- Act of 2007),15 and the GHG and CAFE standards. tial motor vehicle fuel in the late 1980s and early 1990s for a The public sector approach to developing, promoting, number of reasons (API/WRI 1990). Its high octane content or deploying alternative-fuel vehicles includes (1) research could be used in some ICEs with a higher compression ratio and development, (2) demonstration projects, (3) fleet -de to improve efficiency; it could also result in lower emissions ployment, (4) niche market development, (5) public-private than conventional gasoline-fueled vehicles and thus lead to partnerships, and (6) various policy and financial incentives. improved air quality and better public health. From a national The general approach has been based on the supposition that security point of view, it could displace imported petroleum if alternative-fuel vehicles would need to be subsidized until the it was produced from biomass or from natural gas. The con- point where the life-cycle costs of the vehicles and fuel would version of global sources of remote natural gas that were of become competitive with those of gasoline-fueled vehicles; low economic value was envisioned as an approach to diver- market forces would thereafter operate without subsidies, sify the U.S. global supply chain for light-duty vehicle fuels leading to broad deployment (NRC 2008, 2010a, 2013b). and replace petroleum. In some cases in which advancements in technology were Given extensive experience with the use of MeOH and needed, government and private-sector funding of research ICEs in the world of competitive racing, the development of and development led to a technology push. DOE partnered MeOH-powered vehicles did not necessitate fundamental with the private sector on vehicle technologies and fuels technology breakthroughs and such vehicles were developed through such activities as the Partnership for a New Genera- for the market that could operate on either high levels of tion of Vehicles (PNGV), the FreedomCAR and Fuel Partner- methanol, such as 85 percent MeOH and 15 percent gaso- ship, and, currently, the U.S. Driving Research and Innova- line (referred to as M85), or on any combination between tion for Vehicle Efficiency and Energy Sustainability (U.S. M0 and M85. These flexible-fuel vehicles (FFVs) were sold DRIVE) Partnership (NAS/NAE/NRC 2009; NRC 2008, in the marketplace and represented a strategy for overcom- 2010b, 2013a). Extensive work on bioenergy crops and tech- ing the chicken-and-egg problem of attracting investment nologies for their conversion into ethanol and, now, “drop-in” in a methanol-fueling infrastructure before there are enough biofuels16 has also been funded (NRC 2011b). MeOH-fueled vehicles on the road and the lack of demand To create a market pull for some of the technologies un- for such vehicles until an extensive fueling infrastructure is in der development, various tax or policy mandates aimed at place. It was anticipated that with MeOH FFVs, the vehicle stimulating market demand were initiated. Financial incen- owner could use the existing gasoline infrastructure while a tives were especially needed for new technologies that were MeOH-fueling infrastructure was built in response to vehi- projected to be more expensive than the incumbent conven- cles deployed for fleets, incentives were implemented, and a tional technologies, at least in the initial and transitional business case for fuel investors became viable. phases. For example, tax credits for certain types of vehicles Despite subsidies, a broad M85 infrastructure never mate- and tax breaks for certain fuels were put into the tax code to rialized. Furthermore, the continued improvement of gasoline- stimulate the adoption of the new, more costly technologies, powered vehicles along with the development of reformulat- at least for a period of time until increased production and ed gasoline resulted in gasoline-powered vehicles that could economies of scale drove costs down to the point where the achieve the same or better emission performance as promised new technologies would be competitive in the marketplace. by MeOH and essentially eliminated the need for MeOH-fu- It was also thought that adoption by fleets and promotion eled vehicles. in niche markets would, in many cases, help with this transition period by increasing sales and production and thus Ethanol driving down costs, but studies evaluating the programs do not provide a clear picture of whether that strategy is useful Similar to MeOH, ethanol (EtOH) is a fuel with a high oc- (Leiby and Rubin 2004; McNutt and Rodgers 2004; Rob- tane content and one that could be produced from a variety of domestic resources, although it has an energy density per unit 15 Pub. L.110-140, 121 Stat. 1758. USC § 17001. volume only two-thirds that of gasoline.17 It continues to be of 16 The term drop-in biofuel refers to the conversion of biomass interest with a focus on, as with methanol, the development of into fuels, such as gasoline and diesel, that are compatible with, FFVs that can operate on mixtures from 0 percent EtOH and and can be “dropped into,” the current fueling infrastructure. This approach would avoid overcoming the barriers to developing and investing in the infrastructure necessary for a separate fueling sys- 17 The lower volumetric energy density results in a lower miles- tem, such as would be required for fuels, such as ethanol, methanol, per-gallon fuel efficiency compared with gasoline and, all else be- natural gas, and hydrogen. ing equal, will require more frequent refueling.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 137 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

122 Overcoming Barriers to Deployment of Plug-in Electric Vehicles 100 percent gasoline to 85 percent EtOH and 15 percent gaso- cheap in the 1990s as it has again in the past few years. Most line (E85) to address the chicken-and-egg problem. EtOH was of the vehicles developed were dedicated CNG vehicles that originally used as a gasoline additive during summer months avoided the extra vehicle cost and complexity that would be to reduce air pollutant emissions. Later, the EtOH program needed for a dual-fuel vehicle, although some dual-fuel natu- was viewed as a means of displacing petroleum and using do- ral gas vehicles were offered in the market in 2012-2013, mestic resources for EtOH production. Vehicle manufactur- stimulated by low natural gas prices and projections of future ers were given credits in CAFE regulations, and this led to a plentiful reserves and associated low prices. significant production of ethanol-capable FFVs. Although the CNG is typically stored onboard the vehicle at 3,600 psi program might have been a successful transition strategy for and requires high-pressure fueling stations. Incentives and ultimately replacing gasoline with ethanol, the reality was that mandates were provided in the 1990s to bring the vehicles to most of the ethanol-capable FFVs used little, if any, ethanol market, but because of the need for high-pressure fueling sta- (NRC 2002). tions and bulky storage tanks on the vehicle and the shorter In addition, the federal government invested a great deal driving range compared with comparable gasoline-powered of research and development in the development of nonfood vehicles, they tended to be used in fleets where the vehicles crops (such as species of trees and grasses) that could be returned to a central station at the end of each day and could be grown on energy plantations and whose cellulose could serve refueled. They were somewhat more expensive than compa- as a feedstock for conversion technologies to produce EtOH rable gasoline-powered vehicles, trunk space was somewhat (NAS/NAE/NRC 2009; NRC 2011b). It was envisioned that, compromised, driving range was shorter, and an extensive if successful, renewable fuel production system would have refueling infrastructure was not, and still is not, available. low net GHG emissions and enhance energy security and Consumers did not embrace CNG vehicles, and these vehicles would not compete with land for the production of food. The have not moved beyond the niche fleet markets. development of cost-effective cellulosic-based EtOH technologies that can compete with gasoline has proven more Fuel-Cell Vehicles difficult to achieve than anticipated. But there is ongoing demonstration and development of such biomass conversion Another major alternative vehicle and fuel technology technologies, and it remains to be seen how much this alterna- that has been promoted and developed to varying degrees tive fuel will contribute to the U.S. transportation fuel supply. by the public and private sector is the hydrogen-fueled FCV. As with MeOH, an extensive system of fueling sta- It uses on-board hydrogen in a fuel cell to produce electric tions supplying E85 has yet to emerge, even though there power to drive the vehicle. Because its only emission is wa- are millions of EtOH FFVs on the road (most of which use ter vapor, it is classified by California as a ZEV. The federal gasoline) and the U.S. Congress mandated the use of a cer- government and the private sector have provided substantial tain amount of EtOH through the Renewable Fuel Standard funding for research and development, for vehicle demon- (NRC 2011b). To date, most EtOH is produced from corn or strations, and for parts of the needed hydrogen infrastructure sugar cane and is used as a renewable-fuel replacement for (NRC 2010b, 2013a). There has been significant technical petroleum with up to 10 percent EtOH blended into gasoline. progress and promise of driving ranges and fueling times In some places, mostly the Midwest, E15 can be sold and comparable with those of conventional vehicles, but they are used in light-duty vehicles with a model year 2001 or later. still a work in progress. Cost-effective production of hydro- Increasing the percentage for conventional vehicles has re- gen, deploying the necessary hydrogen infrastructure, and ceived opposition from some quarters because of the poten- overcoming the chicken-and-egg barriers remain formidable tially deleterious effects of ethanol on engine components, challenges for these vehicles. Some vehicle manufacturers particularly marine engines, although this is not an issue for have indicated that such vehicles will be available for the FFVs. However, the aggressive policies and subsidies have market in the 2015-2016 time frame, and Hyundai began led to about 7 percent replacement of gasoline-energy use leasing a fuel-cell vehicle in California in 2014. However, in light-duty vehicles from less than 1 percent in 2000, and the higher costs of these vehicles compared with convention- this demonstrates that sustained efforts by the federal gov- al vehicles will be substantial, and thus their deployment will ernment can have demonstrable effects in the market (Gru- require subsidies and other new technologies to overcome enspecht 2013). the initial cost barrier (NRC 2008, 2013b).

Compressed Natural Gas Hybrid Electric Vehicles

Another alternative-vehicle system that garnered inter- In 1999, HEVs were introduced into the U.S. automotive est in the 1990s and one that is also used worldwide are ve- market (ANL 2014). A federal income tax incentive for HEVs hicles using compressed natural gas (CNG). They offer air existed between 2000 and 2010. The original tax incentives quality advantages in urban areas and can be fueled from provided a tax deduction of up to $2,000, but the Energy Poli- domestic sources of natural gas, which seemed plentiful and cy Act of 2005 increased it to a maximum of $3,400 and con-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 138 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 123 verted it into a tax credit. The deductions and credits were the tion, it was difficult for the committee to determine which maximums granted for the most fuel-efficient vehicles. HEVs incentives would be the most effective and should be pur- with lesser fuel-economy received more modest tax credits. sued. However, on the basis of its review of the barriers to The tax credits were available for the first 60,000 vehicles PEV adoption and current and past federal and state incen- sold by a manufacturer, after which time the tax credits would tive programs, the committee offers the following recom- expire. In addition to the federal income tax credits, states of- mendations: fered a wide array of other consumer incentives, including income tax credits, sales-tax reductions or exemptions, access Recommendation: Federal financial incentives to purchase to high-occupancy-vehicle (HOV) lanes, reduced registration PEVs should continue to be provided beyond the current fees, and exemptions from emissions testing, similar to the in- production volume limit as manufacturers and consumers centives now offered for PEVs. experiment with and learn about the new technology. The The effectiveness of the purchase incentives for HEVs federal government should re-evaluate the case for incen- has been extensively studied in the United States and else- tives after a suitable period, such as 5 years. Its re-evaluation where. As noted above, the most important finding from the should consider advancements in vehicle technology and literature is that, immediate purchase incentives (such as progress in reducing production costs, total costs of owner- a sales-tax exemption or instant rebate) are more effective ship, and emissions of PEVs, HEVs, and ICE vehicles. than tax credits or deductions because consumers appear to focus on up-front price and highly discount long-term cost Recommendation: Given the research on effectiveness of savings (Diamond 2009; Chandra et al. 2010; Gallagher and purchase incentives, the federal government should consider Muehlegger 2011). With immediate incentives, buyers do converting the tax credit to a point-of-sale rebate. not have to wonder whether they will qualify for the credit when they file taxes in the next year or estimate its value Recommendation: Given the sparse research on incentives given their income bracket. With immediate incentives, the other than financial purchase incentives, research should be purchase price can be adjusted at the time of sale. A study of conducted on the variety of consumer incentives that are (or Canadian experience with tax rebates for HEVs, which were have been) offered by states and local governments to deter- established at the point of sale, found that they were highly mine which, if any, have proven effective in promoting PEV effective (Chandra et al. 2010). deployment.

Lessons Learned from Past Incentive Programs REFERENCES

The past incentive programs for alternative-fuel tech- Al-Alawi, B.M., and T.H. Bradley. 2014. Analysis of cor- nologies indicate that the market for advanced technology porate average fuel economy regulation compliance adoption needs to be cultivated to progress beyond early scenarios inclusive of plug in hybrid vehicles. Applied adopters. Sustained efforts and economic incentives that Energy 113:1323-1337. create a profitable business case, however, can have demon- ANL (Argonne National Laboratory). 2014. “Light Duty strable effects. The ethanol example is one where the regula- Electric Drive Vehicles Monthly Sales Updates.” http:// tory mandate was successful at advancing an alternative-fuel www.transportation.anl.gov/technology_analysis/ technology; the percentage of ethanol in the domestic gaso- edrive_vehicle_monthly_sales.html. Accessed August line supply by volume increased from less than 1 percent in 20, 2014. 2000 to 10 percent in 2011. Using fleets to encourage main- API/WRI (American Petroleum Institute and World Re- stream adoption does not appear to be particularly effective sources Institute). 1990. Methanol as an Alternative Fuel (for example, in the case of CNG vehicles). Cost reduction Choice: An Assessment. Washington, DC: International and technology advances will continue to evolve as product Energy Program, Foreign Policy Institute, the Paul H. volumes increase, but in the meantime, financial incentives Nitze School of Advanced International Studies, Johns are needed to make a technology more cost-competitive. The Hopkins University, Washington, DC. hybrid example, with a U.S. adoption rate still below 4 per- Beltramello, A. 2012. “Market Development for Green Cars.” cent, shows that even with financial incentives and substan- OECD Green Growth Papers. No. 2012-03. Paris: OECD tial technology advances, moving the deployment from suc- Publishing. cessful regional and niche markets to mainstream adoption Brand, C., J. Anable, and M. Tran. 2013. Accelerating the remains a challenge. transformation to a low carbon passenger transport system: The role of car purchase taxes, feebates, road taxes, RECOMMENDATIONS and scrappage incentives in the UK. Transportation Re- search Part A: Policy and Practice 49: 132-148. Because there has not been any extensive research on Burke, P., and S. Nishitateno. 2013. Gasoline prices, gasoline regulatory or incentive programs that promote PEV adop- consumption, and new vehicle fuel economy: Evidence

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 139 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

124 Overcoming Barriers to Deployment of Plug-in Electric Vehicles from a large sample of countries. Energy Economics 36: Greene, D.L. 2012. “Alternative Fuels: What Have We 363-370. Learned?” Presentation at ARPA-e Transportation Be- CARB. 2013. Final Regulation Order §1962.2 Zero Emis- havior and New Technology Workshop, Berkeley, CA, sion Vehicle Standards for 2018 and Subsequent Model March 16. http://arpa-e.energy.gov/sites/default/files/do Year Passenger Cars, Light-Duty Trucks, and Medium- cuments/files/Alternative%20Fuels%20-%20What%20 Duty Vehicles. http://www.arb.ca.gov/msprog/zevprog/ Have%20We%20Learned.pdf. 2011zevreg/attachb3.pdf. Gruenspecht, H. 2013. “Biofuels in the United States: Con- Chandra, A., S. Gulati, and M. Kandlikar. 2010. Green driv- text and Outlook.” Presented at the Biofuels Workshop, ers or free riders? An analysis of tax rebates for hybrid Institute of Medicine, National Academy of Sciences, vehicles. Journal of Environmental Economics and Man- Washington, DC, January 24. http://www.eia.gov/press- agement 60(2): 78-93. room/presentations/howard_01242013.pdf. Diamond, D. 2009. The impact of government incentives for Huang, E. 2010. “Do Public Subsidies Sell Green Cars? hybrid-electric vehicles: Evidence from US states. En- Evidence from the U.S. ‘Cash for Clunkers’ Program.” ergy Policy 37(3): 972-983. Discussion Paper #2010-17. Harvard Kennedy School, DOE (U.S. Department of Energy). 2011. One Million Elec- Belfer Center, Cambridge, MA. December. tric Vehicles by 2015: February 2011 Status Report. Hwang, R. 2009. “U.S. Alternative Fuels Policies: Lessons http://www1.eere.energy.gov/vehiclesandfuels/pdfs/1_ Learned and Future Directions.” Presentation at Senate million_electric_vehicles_rpt.pdf. Briefing on Alternative Fuels. Washington, DC, Febru- DOE. 2014a. “Vehicle Cost Calculator.” Alternative Fuels ary 6. Data Center. http://www.afdc.energy.gov/calc/. IEA (International Energy Agency). 2013. Global EV Out- DOE. 2014b. “State Laws and Incentives.” Alternative Fuels look: Understanding the Electric Vehicle Landscape to Data Center. http://www.afdc.energy.gov/laws/state. 2020. http://www.iea.org/publications/globalevoutlook_ Doyle, A., and N. Adomaitis. 2013. “Norway Shows the Way 2013.pdf. with Electric Cars, but at What Cost? Reuters, March 13. IRS (Internal Revenue Service). 2009. “Qualified Vehicles http://www.reuters.com/article/2013/03/13/us-cars-nor Acquired after 12-31-2009.” Last reviewed or updated way-idUSBRE92C0K020130313. November 17, 2014. http://www.irs.gov/Businesses/Qua Edwards, A. 2013. “The PEV Customer: How to Overcome lified-Vehicles-Acquired-after-12-31-2009. Potential Sales Barriers.” Presentation to the Commit- Jin, L., S. Searle, and N. Lutsey. 2014. “Evaluation of State- tee on Overcoming Barriers to Electric-Vehicle Deploy- Level U.S. Electric Vehicle Incentives.” White Paper pre- ment, Washington, DC, December 3. sented to the International Council on Clean Air Transpor- EIA (U.S. Energy Information Administration). 2015. “Retail tation. http://www.theicct.org/sites/default/files/publica Premium Gasoline Prices, Selected Countries.” http:// tions/ICCT_state-EV-incentives_20141030.pdf. www.eia.gov/countries/prices/gasolinewithtax.cfm. Keddie, E. 2013. “California’s Zero Emission Vehicle (ZEV) EPA/NHTSA (U.S. Environmental Protection Agency and Regulation.” Presentation to the Committee on Over- National Highway Traffic Safety Administration). 2012a. coming Barriers to Electric-Vehicle Deployment, Irvine, 2017 and Later Model Year Light-Duty Vehicle Green- CA, May 8. house Gas Emissions and Corporate Average Fuel Econ- Krause, R.M., S.R. Carley, B.W. Lanea, and J.D. Graham. omy Standards. Federal Register 77(199): October. http:// 2013. Perception and reality: Public knowledge of plug- www.gpo.gov/fdsys/pkg/FR-2012-10-15/pdf/ in electric vehicles in 21 U.S. cities. Energy Policy 63: 2012-21972.pdf. 433-440. EPA/NHTSA. 2012b. Joint Technical Support Document: Leiby, P., and J. Rubin. 2004. “Understanding the Transi- Final Rulemaking for 2017-2025 Light-Duty Vehicle tion to New Fuels and Vehicles: Lessons Learned from Greenhouse Gas Emission Standards and Corporate Analysis and Experience of Alternative Fuel and Hybrid Average Fuel Economy Standards. Report No. EPA- Vehicles.” In The Hydrogen Energy Transition: Moving 420-R-12-901. Washington, DC. Towards the Post-Petroleum Age in Transportation, ed- EV Norway. 2014. “Powered by Nature: A Guide to Norway, ited by D. Sperling and J. Cannon, 191-212. Waltham, the World’s Leading EV Market.” http://www.evnor- MA: Elsevier Academic Press. way.no/. Accessed January 31, 2015. McNutt, B., and D. Rodgers. 2004. “Lessons Learned from Gallagher, K.S., and M. Muehlegger. 2011. Giving green to 15 Years of Alternative Fuels Experience—1988 to get green? Incentives and consumer adoption of hybrid 2003.” In The Hydrogen Energy Transition: Moving To- vehicle technology. Journal of Environmental Econom- wards the Post-Petroleum Age in Transportation, edited ics and Management 61(1): 1-15. by D. Sperling and J. Cannon, 165-179. Waltham, MA: Gallagher, K.S., and R. Steenblick. 2013. Technology-Push Elsevier Academic Press. and Demand-Pull Policies to Support Electric Propulsion. METI (Ministry of Economy, Trade, and Industry). 2013. OECD Report COM/TAD/ENV/JWPTE, Paris, France. “Our Policy about Promoting EVs.” Presentation to the

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 140 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Incentives for the Deployment of Plug-in Electric Vehicles 125 Committee on Overcoming Barriers to Electric-Vehicle Ozaki, R., and K. Sevastyanova. 2011. Going hybrid: An anal- Deployment, Tokyo, Japan, December 9. ysis of consumer purchase motivations. Energy Policy Mock, P., and Z. Yang. 2014. “Driving Electrification: A 39(5): 2217-2227. Global Comparison of Fiscal Incentive Policy for Elec- Prante, G. 2007. “Most Americans Don’t Itemize on Their Tax tric Vehicles.” White paper submitted to the Internation- Returns.” Tax Foundation, July. http://taxfoundation. al Council of Clean Transportation, Washington, DC, org/article/most-americans-dont-itemize-their-tax-re- May. http://www.theicct.org/sites/default/files/publica- turns. tions/ICCT_EV-fiscal-incentives_20140506.pdf. Robertson, B., and L.K. Beard. 2004. “Lessons Learned in Morrow, W.R., K.S. Gallagher, G. Collantes, and H. Lee. the Deployment of Alternative Fueled Vehicles.” In The 2010. Analysis of policies to reduce oil consumption Hydrogen Energy Transition: Moving Towards the Post- and greenhouse-gas emissions from the U.S. transporta- Petroleum Age in Transportation, edited by D. Sperling tion sector. Energy Policy 38(3): 1305-1320. and J. Cannon, 181-190. Waltham, MA: Elsevier Aca- NAS/NAE/NRC (National Academy of Sciences, National demic Press. Academy of Engineering, and National Research Coun- RTI/UMTRI (Research Triangle Institute and University cil). 2009. America’s Energy Future: Technology and of Michigan Transportation Research Institute). 2009. Transformation. Washington, DC: The National Acad- Automobile Industry Retail Price Equivalent and In- emies Press. direct Cost Multipliers: Report Prepared for the U.S. NRC (National Research Council). 2002. Effectiveness and Environmental Protection Agency. RTI Project Number Impact of Corporate Average Fuel Economy (CAFÉ) 0211577.002.004. Washington, DC. Standards. Washington, DC: The National Academies Sallee, J.M. 2011. The surprising incidence of tax credits Press. for consumers. American Economic Journal: Economic NRC. 2008. Transitions to Alternative Transportation Tech- Policy 3(2): 189-219. nologies—A Focus on Hydrogen. Washington, DC: The Small, K. 2012. Energy policies for passenger motor vehi- National Academies Press. cles. Transportation Research Part A: Policy and Prac- NRC. 2010a. Transitions to Alternative Transportation Tech- tice 46(6): 874-889. nologies: Plug-in Hybrid Electric Vehicles. Washington, State ZEV Programs. 2013. “State Zero-Emission Vehicle DC: The National Academies Press. Programs—Memorandum of Understanding.” Signed NRC. 2010b. Review of the Research Program of the October 24, 2013. http://arb.ca.gov/newsrel/2013/8s_ FreedomCAR and Fuel Partnership, Third Report. zev_mou.pdf. Washington, DC: The National Academies Press. Tellis, G. 2013. “Toyota’s Gamble on the Prius.” Financial NRC. 2011a. Assessment of Fuel Economy Technologies Times Management. Financial Times, March 4. http:// for Light-Duty Vehicles. Washington, DC: The National www.ft.com/intl/cms/s/0/146ad23c-7230-11e2-89fb- Academies Press. 00144feab49a.html#axzz3IJ1hL64u. NRC. 2011b. Renewable Fuel Standard: Potential Econom- Tesla. 2013. “Electric Vehicle Incentives Around the World.” ic and Environmental Effects of U.S. Biofuel Policy. http://www.teslamotors.com/incentives/japan. Accessed Washington, DC: The National Academies Press. February 14, 2013. NRC. 2013a. Review of the Research Program of the U.S. Yeh, S., and J. Witcover. 2012. Status Review of California’s DRIVE Partnership, Fourth Report. Washington, DC: Low Carbon Fuel Standard. Series UCD-ITS-RP-12-33. The National Academies Press. University of California, Davis. August. NRC. 2013b. Transitions to Alternative Vehicles and Fuels. ZEV Program Implementation Task Force. 2014. Multi-State Washington, DC: The National Academies Press. ZEV Action Plan. March. http://www.nescaum.org/docu Ohnsman, A. 2013. “Tesla Tops California Green-Car Credit ments/multi-state-zev-action-plan.pdf/. Sales in Past Year.” Bloomberg Businessweek, October 16. Zhang, X., K. Wang, Y. Hao, J. Fan, and Y. Wei, 2013. The http://www.businessweek.com/news/2013-10-16/musk-s impact of government policy on preference for NEVs: -tesla-tops-california-green-car-credit-sales-in-past-year. The evidence from China. Energy Policy 61: 382-393.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 141 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 142 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Appendixes

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 143 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 144 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Biographical Information on the Committee on Overcoming Barriers to Electric-Vehicle Deployment

JOHN G. KASSAKIAN, Chair, is a professor of electri- Office, and deputy assistant secretary in the U.S. Department cal engineering and former director of the Massachusetts of Energy. He holds a doctorate in business administration Institute of Technology (MIT) Laboratory for Electromag- from Harvard University, MS degrees in nuclear engineering netic and Electronic Systems. His expertise is in the use of and management from the Massachusetts Institute of Tech- electronics for the control and conversion of electric energy, nology, and a BS from the U.S. Military Academy. He served industrial and utility applications of power electronics, elec- in the Army in Vietnam. tronic manufacturing technologies, and automotive electric and electronic systems. Before joining the MIT faculty, he JEFF DOYLE is a principal with D’Artagnan Consulting, served in the U.S. Navy. Dr. Kassakian is on the boards of LLP, a firm that specializes in the use of advanced technolo- directors of a number of companies and has held numerous gies to achieve important transportation policy and finance positions with the Institute of Electrical and Electronic En- objectives. Prior to that, Mr. Doyle served as director of gineers (IEEE), including as founding president of the IEEE public-private partnerships for the Washington State De- Power Electronics Society. He is a member of the National partment of Transportation (WSDOT), where he oversaw a Academy of Engineering, a fellow of IEEE, and a recipi- program that develops partnerships with the private sector ent of the IEEE William E. Newell Award for Outstand- to advance transportation projects, programs, and policies. ing Achievements in Power Electronics (1987), the IEEE His office implemented the West Coast Electric Highway, a Centennial Medal (1984), and the IEEE Power Electronics partnership project that provided the nation with its first bor- Society’s Distinguished Service Award (1998). He is a co- der-to-border network of fast-charging stations for electric author of the textbook Principles of Power Electronics and vehicles. Mr. Doyle also served as co-chair of Washington’s has served on a number of National Research Council com- Plug-in Vehicle Task Force, as a member of the Puget Sound mittees, including the Electric Power/Energy Systems Engi- Regional Council Electric Vehicle Infrastructure Advisory neering Peer Committee, the Committee on Assessment of Committee, and as a member of the National Cooperative Solid State Lighting, the Committee on Review of the 21st Highway Research Program Project Panel 20-83 (04), “Ef- Century Truck Partnership Phase 2, the Committee on Re- fects of Changing Transportation Energy Supplies and Alter- view of the Research Program of the Partnership for a New native Fuel Sources on Transportation.” Other public-private Generation of Vehicles, and the Committee on Review of the partnership projects include redeveloping public ferry termi- FreedomCAR and Fuel Research Program. He has an ScD in nals, providing transit-oriented development with advanced electrical engineering from MIT. traveler-information systems at state-owned park-and-ride facilities, and implementing alternative finance and funding DAVID L. BODDE is an engineering professor and senior mechanisms for transportation infrastructure development fellow in Clemson University’s International Center for Au- and maintenance. Prior to joining WSDOT, Mr. Doyle served tomotive Research. Before joining Clemson University, Dr. as staff director and senior legal counsel to the Transporta- Bodde held the Charles N. Kimball Chair in Technology and tion Committee in the Washington legislature, where his Innovation at the University of Missouri–Kansas City. He work focused on transportation policy, finance, and freight serves on the boards of directors of several energy and tech- mobility issues. He is a member of the Washington State nology companies, including Great Plains Energy and the Bar Association and serves on the Supervisory Committee Commerce Funds. His executive experience includes vice of a state-chartered credit union in Washington. Mr. Doyle president of Midwest Research Institute and president of earned a BA in political science from Western Washington MRI Ventures, assistant director of the Congressional Budget University and a JD from Seattle University.

129

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 145 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

130 Overcoming Barriers to Deployment of Plug-in Electric Vehicles GERALD GABRIELSE is Leverett Professor of Physics at Intergovernmental Panel on Climate Change, which won the Harvard University. His previous positions include assistant 2007 Nobel Peace Prize. Mr. Hwang serves or has served and associate professor, University of Washington, and chair on numerous advisory panels, including the California Plug- of the Harvard Department of Physics. His research focuses in Electric Vehicle Collaborative, the National Academy of on making accurate measurements of the electron magnetic Sciences Committee on Fuel Economy, the U.S. EPA Mobile moment and the fine structure constant and on the precise Source Technical Review Subcommittee, the California Air laser spectroscopy of helium. Dr. Gabrielse also leads the Resources Board’s Alternative and Renewable Fuels and Ve- International Antihydrogen TRAP (ATRAP) Collaboration, hicles program, the California Hydrogen Highway Network whose goal is accurate laser spectroscopy with trapped anti- Advisory Panel, the Automotive X Prize, and the Western hydrogen atoms. His many awards and prizes include fellow- Governors’ Association Transportation Fuels for the Future ship of the American Physical Society, the Davisson-Germer Initiative. Before joining NRDC, he was the director of the prize of the American Physical Society, the Humboldt Re- Union of Concerned Scientists transportation program. He search Award (Germany, 2005), and the Tomassoni Award has also worked for the U.S. Department of Energy at Law- (Italy, 2008). Harvard University awarded him its George rence Berkeley National Laboratory and the California Air Ledlie Research Prize and its Levenson Teaching Prize. Resources Board as an air pollution engineer and was in- His hundreds of outside lectures include a Källén Lecture volved in forecasting residential and industrial energy de- (Sweden), a Poincaré Lecture (France), a Faraday Lecture mand, permitting of hazardous waste incinerators, and eval- (Cambridge, U.K.), a Schrödinger Lecture (Austria), a Zach- uating toxic air emissions from landfills. He is currently on ariasen Lecture (University of Chicago), and a Rosenthal the National Research Council Committee on Fuel Economy Lecture (Yale). He is a member of the National Academy of of Light-Duty Vehicles, Phase 2. Mr. Hwang has an MS in Sciences and has participated on many National Research mechanical engineering from the University of California, Council committees, including the Committee on Review of Davis, and a master’s degree in public policy from the Uni- the U.S. DRIVE Research Program, Phase 4, and the Com- versity of California, Berkeley. mittee on Review of the FreedomCAR and Fuel Research Program, Phase 3. He has a BS from Calvin College and an PETER ISARD is a consultant on economic policy issues MS and a PhD in physics from the University of Chicago. and held various managerial positions with the International Monetary Fund (IMF) from 1985 to 2008, primarily in the KELLY SIMS GALLAGHER (committee member until Research Department. Dr. Isard played a lead role in help- June 2014) is an associate professor of energy and environ- ing Lithuania design an economic transformation program in mental policy of the Fletcher School, Tufts University. She 1991 and 1992 and spent the 2002-2003 academic year at the directs the Energy, Climate, and Innovation research program University of Maryland. He retired in June 2008 as deputy of the Center for International Environment and Resource Pol- director of the IMF Institute, the department that provides icy. She is also a senior associate and a member of the Board training on economic policy making for member-country of Directors of the Belfer Center for Science and International officials. Before joining the IMF in 1985, he spent 1970 in Affairs of Harvard University, where she previously directed the Research Department of the IMF, taught at Washington the Energy Technology Innovation Policy research group. University in St. Louis in 1971-1972, held research and man- Broadly, she focuses on energy and climate policy in the agement positions at the Federal Reserve Board from 1972 United States and China. She is particularly interested in the to 1985, and spent a year during 1979 and 1980 at the Bank role of policy in spurring the development and deployment of for International Settlements. Dr. Isard is the author of nu- cleaner and more efficient energy technologies domestically merous articles in academic journals, primarily on exchange and internationally. She speaks Spanish and basic Mandarin rates and monetary policy strategies. He is also the author of Chinese and is a member of the Council on Foreign Relations. two books—Exchange Rate Economics (1995) and Global- She is the author of China Shifts Gears: Automakers, Oil, Pol- ization and the International Financial System (2005)—and lution, and Development (MIT Press, 2006), editor of Acting editor of several others. He has an undergraduate degree in in Time on Energy Policy (Brookings Institution Press, 2009), mathematics from the Massachusetts Institute of Technology and author of numerous academic articles and policy reports. and a PhD in economics from Stanford University. A Truman Scholar, she has an MA in Law and Diplomacy and a PhD in international affairs from the Fletcher School and an LINOS JACOVIDES is professor of electrical engineering AB from Occidental College. at Michigan State University. From 1998 to 2007, he served as director of Delphi Research Laboratories. Dr. Jacovides ROLAND HWANG is the transportation program direc- joined General Motors Research and Development in 1967 tor for the Natural Resources Defense Council (NRDC) and and became department head of electrical engineering in works on sustainable transportation policies. He is an expert 1985. His research was in the interactions between power on clean vehicle and fuels technologies and was a member electronics and electric machines in electric vehicles and lo- of the Technology Assessment and Economics Panel of the comotives. He later transitioned to Delphi with a group of re-

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 146 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Appendix A 131

searchers from General Motors to set up the Delphi Research Week, Reuters, Bloomberg News, the Los Angeles Times, the Laboratories. He is a member of the National Academy of Detroit News, the Detroit Free Press, the Wall Street Jour- Engineering and has served on numerous National Research nal, and National Public Radio, for her coverage of new Council committees, including the National Cooperative product launches and the balance sheet conditions of manu- Highway Research Program’s Panel on Effects of Changing facturers and brands. Prior to her work at IHS, Ms Lindland Transportation Energy Supplies and Alternative Fuel Sourc- worked at AlliedSignal in Rumford, Rhode Island, where she es on State Departments of Transportation, the Committees forecasted products, such as Bendix brakes. A life-long auto- on Assessment of Technologies for Improving Light-Duty motive enthusiast, she began her career as a staff accountant Vehicle Fuel Economy (Phases 1 and 2), the Committee on with Mercedes-Benz Credit Corporation in Norwalk, Con- Review of the U.S. Drive Research Program, Phase 4, the necticut. Ms Lindland holds a double major in accounting Committee on Electric Vehicle Controls and Unintended Ac- and business administration from Gordon College, Wenham, celeration, and the Committee on Review of the Freedom- Massachusetts. She is a former board member of the Soci- CAR and Fuel Research Program, Phase 3. Dr. Jacovides is ety of Automotive Analysts, the International Motor Press a fellow of IEEE and the Society of Automotive Engineers Association, and Motor Press Guild and was accepted into (SAE) International and served as president of the Industry Strathmore's 2001 Who's Who in American Business. Applications Society of IEEE in 1990. He received a BS in electrical engineering and an MS in machine theory from RALPH D. MASIELLO is the senior vice president and in- the University of Glasgow and a PhD in generator control novation director of DNVGL, Inc. In recent years, his focus systems from the Imperial College, University of London. has been on electric market and transmission operator business models and systems, including cost-benefit analyses of ULRIC KWAN is a senior managing consultant with IBM, paradigms for models, systems, and operations. He has also concentrating on helping utilities around the globe enable developed technology and strategic plans for market opera- customer-oriented systems of engagement. Previously, he tors and automation and smart grid roadmaps for several in- worked for Pacific Gas and Electric, where he was a leader in dependent system operators. His current interests include the the electric vehicle and demand response fields. In this role, market and utility applications of advanced storage devices he was responsible for the development of business and regu- for ancillary markets, reliability, and energy economics; the latory justification for utility involvement in electric-vehicle grid integration of electric vehicles; and the development of deployments, integration of electric vehicles and customers advanced building-to-grid concepts. He has provided expert into the electricity market, creation of a long-term contract testimony before Congress on metering systems and mar- between an automaker and utility for demand response, and ket operations and cosigned a Supreme Court amicus curiae development of the first integrated point-of-sale rate calcula- brief on transmission access and native load service. He was tor for electric-vehicle customers. Earlier, he worked at Sie- recently appointed to the Department of Energy Electric- mens as an energy engineer and LCG Consulting as an elec- ity Advisory Council. Dr. Masiello is a fellow of the IEEE tricity wholesale market consultant and forecaster. Mr. Kwan and has served as chairman of Power System Engineering, has participated in numerous regulatory proceedings at the as chairman of Power Industry Computing Applications, on California and federal electricity regulatory bodies on how the editorial board of IEEE Proceedings and on the advisory to integrate customers into the wholesale market and serves board of IEEE Spectrum magazine. He is the winner of the or has served on numerous advisory panels, including the 2009 IEEE PES Concordia award for power system analysis North American Electric Reliability Corporation (NERC) and is a member of the National Academy of Engineers. He advisory team for demand response, the California Plug-in received his PhD from the Massachusetts Institute of Tech- Electric Vehicle Collaborative, and the California Electric nology in electrical engineering. Transportation Coalition. Mr. Kwan has a BS and MS in mechanical engineering from the University of Calgary and JAKKI MOHR is the Regents Professor of Marketing at Stanford University, respectively. the University of Montana–Missoula (UM). An international expert and innovator in marketing of high-technology prod- REBECCA LINDLAND is a senior fellow with the King ucts and services, she has achieved international acclaim for Abdullah Petroleum Studies and Research Center and leads Marketing of High-Technology Products and Innovations its work on transportation policy, technology, and consumer (coauthor with S. Sengupta and S. Slater, with European and demand. She was formerly the director of research for IHS India/Southeast Asia editions and translations into Chinese, Automotive where she was responsible for evaluating and Portuguese, and Korean). Motivated by the desire to apply assessing vehicle manufacturers that participate in the U.S. the promise of new technologies to solve social and global and Canada marketplaces. She has a particular interest in problems, Dr. Mohr has provided training to companies and how manufacturers’ decisions reflect consumer values. As universities worldwide in strategic market planning to com- a member of IHS Automotive, Ms Lindland was frequently mercialize innovation. She has received numerous teach- quoted in the media, including the New York Times, Business ing awards—including the Outstanding Marketing Teacher

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 147 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

132 Overcoming Barriers to Deployment of Plug-in Electric Vehicles Award (presented by the Academy of Marketing Science), RICHARD TABORS is president of Across the Charles and the Carnegie Foundation CASE Professor of the Year, and is director of the Utility of the Future Project at the MIT Ener- the Most Inspirational Teacher of the Year Award at the Uni- gy Initiative. Until July 2012, he was vice president of Charles versity of Montana—and the Distinguished Scholar Award, River Associates (CRA) in the Energy & Environment Prac- the John Ruffatto Memorial Award, and the Dennison Presi- tice. He founded the engineering-economics consulting firm dential Faculty Award for Distinguished Accomplishment. of Tabors Caramanis & Associates (TCA) in 1989 to provide Dr. Mohr served as a Fulbright senior specialist in Montevi- economic, regulatory, and financial analytic support to the deo, Uruguay. Her research has received national awards and restructuring of the U.S. and international electric power in- has been published in the Journal of Marketing, the Strategic dustry. TCA was sold to CRA in 2004. He was a researcher Management Journal, the Journal of the Academy of Mar- and member of the faculty at Harvard University from 1970 keting Science, the Journal of Public Policy and Marketing, to 1976 and was at Massachusetts Institute of Technology as and others. In research sponsored by the Marketing Science a senior lecturer in technology management and policy and a Institute, she studies how companies use biomimicry (inno- research director in power systems from 1976 to 2004. He is vations inspired by nature that are based on underlying bio- a visiting professor of electrical engineering at the University logic mechanisms) to solve technical and engineering chal- of Strathclyde in Glasgow. His research and development ac- lenges, the basis of her TEDxSanDiego talk in 2011. Before tivities at MIT led to his authorship or coauthorship of over joining UM in fall 1997, Dr. Mohr was an assistant professor 80 articles and books, including Spot Pricing of Electricity, at the University of Colorado Boulder (1989-1997). Before on which the economic restructuring of the electric utility beginning her academic career, she worked in Silicon Val- wholesale and retail markets is based. Dr. Tabors continues ley in advertising for Hewlett-Packard's Personal Computer his directing and consulting activities in regulation, litigation, Group and TeleVideo Systems. Dr. Mohr received her PhD and asset evaluation in the power industry with a focus on from University of Wisconsin–Madison. development of future platforms and pricing structure of the smart grid. He received a BA in biology from Dartmouth and MELISSA SCHILLING is a professor of management and an MS and a PhD in geography and economics from the Max- organizations at New York University Stern School of Busi- well School of Syracuse University. ness. Dr. Schilling teaches strategic management, corporate strategy, and technology and innovation management. She TOM TURRENTINE is director of the California Energy is widely recognized as an expert in innovation and strategy Commission’s Plug-in Hybrid and Electric Vehicle Research in high-technology industries. Her textbook, Strategic Man- Center at the Institute of Transportation Studies, University agement of Technological Innovation (now in its fourth edi- of California, Davis. For the last 20 years, Dr. Turrentine has tion), is the top innovation-strategy text in the world and is been researching consumer response to alternative fuels, ve- available in seven languages. Her research in innovation and hicle technologies, road systems, and policies that have envi- strategy has earned her such awards as the National Science ronmental benefits. His most recent work includes multiyear Foundation's CAREER Award and the Best Paper in Man- projects to study consumer use of plug-in electric vehicles, agement Science and Organization Science for 2007 Award. including the BMW Mini E, Prius PHEV conversions, the Her research has appeared in leading academic journals, Nissan Leaf, GM Volt, PHEV pickups, and specially designed such as the Academy of Management Journal, the Academy energy-feedback displays in vehicles. He and his researchers of Management Review, Management Science, Organization are studying BEV and PHEV driver travel patterns and use Science, the Strategic Management Journal, and the Jour- of infrastructure and are developing planning tools to advise nal of Economics and Management Strategy and Research on deployment of infrastructure and optimal ways to integrate Policy. She sits on the editorial review boards of Organi- plug-in electric vehicles into California’s grid. He and his zation Science and Strategic Organization. Dr. Schilling re- team wrote “Taking Charge,” a plan for California to develop ceived her BS in business administration from the University a PEV market, which is the blueprint for the California PEV of Colorado, Boulder, and her PhD in strategic management Collaborative. He holds a PhD in anthropology. from the University of Washington.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 148 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Meetings and Presentations

FIRST COMMITTEE MEETING Electric Vehicle Charging Services October 28-29, 2012 Richard Lowenthal, Founder and CTO, ChargePoint

EV Everywhere: Overview and Status The Complete Electric Vehicle Charging Solution Patrick B. Davis, Program Manager, Vehicle Technologies Michael Krauthamer, Director, Mid-Atlantic Region, eVgo Program, U.S. Department of Energy (DOE) Better Place Update EV Everywhere Grand Challenge: Charging Infrastructure Jason Wolf, Vice President for North America, Enabling Flexible EV Design Better Place Lee Slezak, Technology Manager, Vehicle Systems, Vehicle Technologies Program, DOE The DOE Vehicle Technologies Analysis Toolbox and EV Everywhere Target-Setting DOE AVTA: The EV Project and Other Light-Duty Electric Jacob Ward, Vehicle Technologies Analysis Manager, DOE Drive Vehicle Activities James Francfort, Principal Investigator, Advanced The Need for Public Investments to Support the Plug-in Vehicle Testing Activity, Idaho National Laboratory Electric Vehicle Market Nick Nigro, Manager, Transportation Initiatives, SECOND COMMITTEE MEETING Center for Climate and Energy Solutions December 17-19, 2012 Research Insights from the Nation’s Highest Residential Charging Infrastructure Needs Concentration of Electric Vehicles Marcus Alexander, Manager, Vehicle Systems Analysis, Brewster McCracken, President and CEO, Electric Power Research Institute (EPRI) Pecan Street Inc.

General Motors: National Research Council Electric Vehicle Initiatives in the Houston-Galveston Region Britta K. Gross, Director, Advanced Vehicle Allison Carr, Air Quality Planner, Houston-Galveston Commercialization Policy, General Motors Area Clean Cities Coalition

Overcoming Barriers to Deployment of Electric Vehicles The EV Project Deployment Barriers Mike Tamor, Executive Technical Leader, Energy Systems Donald Karner, ECOtality North America and Sustainability, Ford Motor Company New Models of Mobility and EV Deployment Overcoming Barriers to Electric-Vehicle Deployment: Jack Hidary, Global EV Leader, Hertz Barriers to Deployment, an OEM Perspective Joseph Thompson, Project Manager-Technology Electric Vehicle Infrastructure Demonstration Projects: Planning, Nissan Lessons Learned Rick Durst, Transportation Electrification Project The Electrification Coalition: Revolutionizing Manager, Portland General Electric Transportation and Achieving Energy Security Jonna Hamilton, Vice President for Policy, Electrification Coalition 133

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 149 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

134 Overcoming Barriers to Deployment of Plug-in Electric Vehicles THIRD COMMITTEE MEETING Workplace Charging Programs – Nissan January 25-26, 2013 Tracy Woodard, Senior Director for Government Affairs, Nissan No open session presentations were held during this meeting. EV Charging at Lynda.com FOURTH COMMITTEE MEETING Dana Jennings, Facilities Supervisor, Lynda.com, Inc. May 8-9, 2013 Panel Discussion on Workplace Charging California’s Zero-Emission-Vehicle (ZEV) Regulation Katie Drye, Sarah Olexsak, Ali Ahmed, Tracy Woodard, Elise Keddie, Manager, Zero Emission Vehicle Dana Jennings Implementation Section, California Air Resources Board Technical, Manufacturing, and Market Issues Associated Electric-Vehicle Deployment: A Long-Term Perspective with xEV Batteries Chuck Shulock, President, Shulock Consulting Suresh Sriramulu, Vice President, Battery Technology, TIAX LLC Perspective on the Electrification of the Automotive Fleet: The Prius and Beyond Consumers’ Thoughts, Attitudes, and Potential Toyota Motor Corporation Acceptance of Electric Vehicles Chris Travell, Vice President of Automotive Research, Consumer Behavior and Attitudes Concerning PEV Adoption Maritz Research Ed Kim, Vice President, Industry Analysis, AutoPacific SIXTH COMMITTEE MEETING Selling Plug-in Electric Vehicles December 3-4, 2013 Paul Scott, EV Specialist, Downtown LA Nissan The PEV Customer: How to Overcome Potential San Diego Gas & Electric Plug-in Electric Vehicle Landscape Sales Barriers John H. Holmes, Research & Development, Alexander Edwards, President, Strategic Vision Asset Management & Smart Grid Projects, San Diego Gas & Electric Panel Discussion: Dealer Perspective on Plug-in Electric Vehicles FIFTH COMMITTEE MEETING Tammy Darvish, Vice President, DARCARS August 13-14, 2013 Automotive Group

DOE Electric Vehicle Activities Update Neil Kopit, Director of Marketing, Criswell Automotive, Gaithersburg, MD Jake Ward, Program Analyst, Vehicles Technology Program, DOE Greg Brown (via teleconference), General Manager, Serra Chevrolet, Southfield, MI Vehicle Choice Modeling for Advanced Technology Doug Greenhaus, Chief Regulatory Counsel, Environment, and Electric Vehicles Health and Safety, National Automobile Dealers David Greene, Corporate Fellow, National Transportation Association Research Center, Oak Ridge National Laboratory PEV Deployment in the Defense Department: Local Barriers to Plug-in Electric Vehicle (PEV) Deployment Barriers and Strategies Katie Drye, Transportation Project Manager, Camron Gorguinpour, Executive Director, Plug-in Advanced Energy Electric-Vehicle Program, Department of Defense

Workplace Charging Challenge: Part of the EV Massachusetts Electric Vehicle Roadmap Everywhere Grand Challenge Mark Sylvia, Commissioner, Massachusetts Department Sarah Olexsak, Energy Project Specialist, DOE of Energy Resources

Workplace Electric Vehicle Charging FEDEX Experience Ali Ahmed, Senior Manager, Workplace Resources, Russ Musgrove, Managing Director, FedEx Express Global Energy Management and Sustainability, Cisco Systems, Inc. Frito-Lay Experience Steve Hanson, Fleet Sustainability Manager, Frito-Lay

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 150 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Appendix B 135 Panel Discussion on Fleet Deployment Reporting on Site Visits to Japan Camron Gorguinpour, Mark Sylvia, Russ Musgrove, Roland Hwang, Member, Committee on Overcoming and Steve Hanson Barriers to Electric-Vehicle Deployment, Transportation Program Director, Natural Resources Defense Council SEVENTH COMMITTEE MEETING February 25-26, 2014 EIGHTH COMMITTEE MEETING May 6-7, 2014 The Future of Automobile Battery Recycling Linda Gaines, Transportation Systems Analyst, Center for Stationary Wireless Charging of PEVs: Near-Term Barriers Transportation Research, Argonne National Laboratory John Miller, JNJ Miller plc

Li-Ion Technology Evolution for xEVs: How Far Car2Go: Electric Vehicles and Car Sharing and How Fast? Mike Cully, U.S. Regional Manager, Car2Go Menahem Anderman (via WebEx), President, Advanced Automotive Batteries DOE Vehicle Electrification Activities Patrick Davis, Program Manager, Electric Vehicle Charging Infrastructure Usage Observed Vehicle Technologies, DOE in Large-Scale Charging Infrastructure Demonstrations John Smart, Electric Vehicle Test Engineer, Reporting on Site Visits to Europe Energy Storage & Transportation Systems, Jeff Doyle, Member, Committee on Overcoming Idaho National Laboratory Barriers to Electric-Vehicle Deployment, Director Public/Private Partnerships, Washington State What Electric-Vehicle Drivers Want in a Charging Department of Transportation Network (and What They Actually Need) Michael Nicholas, Institute of Transportation Studies, NINTH COMMITTEE MEETING University of California, Davis July 16-17, 2014

Understanding Electric Vehicle Market Barriers: No open session presentations were held during this meeting. An Automotive Manufacturer’s Perspective William P. Chernicoff, Manager, Energy and Environmental Research, Toyota Motors TENTH COMMITTEE MEETING North America, Inc. October 23-24, 2014

EV Infrastructure Financing Solutions No open session presentations were held during this meeting. John Rhow, Kleiner Perkins

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 151 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

International Incentives

This appendix provides some information about the in- the point of sale to reduce the purchase price directly. PEVs centive programs in Japan, France, Norway, Germany, the qualify for the highest bonus of €6,300. The bonus-malus Netherlands, and China. system generated deficits in its first few years (2008-2010) owing to unexpectedly high demand for the lower-CO2 JAPAN emitting vehicles but led to substantial reductions in the

CO2 emissions of new vehicles sold in France (Beltramello In fiscal year (FY) 2013, the Japanese government of- 2012). Average new light-duty vehicle CO2 per kilometer fered rebates for 18 different makes and models of plug-in moved from being the fourth lowest to the lowest in the Eu- electric vehicles (PEVs) available in Japan. For the Nissan ropean Union since the program started in 2007 (Brand et al. Leaf, the FY 2013 rebate was ¥780,000 (about $7,800) based 2013). The bonus-malus system is periodically updated, with on a 2013 target price of ¥2,850,000 (about $28,500) (METI the most recent revision having become effective in January 2013).1 The committee predicts that the rebates will decline 2014. to ¥520,000 (about $5,200) in FY 2014, ¥260,000 (about The bonus-malus system appears to be an effective con- $2,600) in FY 2015, and zero in FY 2016. In addition to the sumer incentive. According to the French government, the rebates, PEV purchasers are also exempt from the vehicle ac- French market for PEVs and hybrid electric vehicles (HEVs) quisition tax (about 5 percent of the purchase price) and from represented 3.1 percent of the global passenger vehicle mar- the vehicle weight or tonnage tax (Nelson and Tanabe 2013). ket in France. Compared with 2012, sales of PEVs increased The acquisition tax is waived through March 2015, and the by 50 percent and sales of HEVs increased by 60 percent. In weight tax is waived through April 2015 (Tesla 2013). The total, 8,779 PEVs were registered in France in 2013. Sales vehicle weight or tonnage tax exemption is applicable once, increased by more than 50 percent compared with the 5,663 at the time of the first mandatory inspection, which occurs vehicles registered in 2012. 3 years after the vehicle purchase. PEV owners also enjoy a substantial discount on the annual automobile tax, which can NORWAY otherwise range from ¥29,500 to ¥111,000, depending on the vehicle’s engine displacement. Finally, some prefectures and The government of Norway has made a firm commitment cities offer additional incentives at time of purchase. to battery electric vehicles (BEVs), motivated in part by the desire to reduce the greenhouse gas (GHG) emissions of its FRANCE transportation fleet. Because almost 100 percent of Norway’s electricity is generated from hydroelectric power, a transition In 2007, France introduced a fee-bate (bonus-malus) to BEVs would decarbonize the passenger vehicle fleet almost system for vehicle purchases based on the carbon dioxide entirely. Forty percent of Norway’s GHG emissions currently (CO2) emissions of the vehicle. The policy levies a fee de- come from the transportation sector, and 60 percent of those pending on the CO2 emission performance of the vehicle come from road transport (Deshayes 2011). ranging from €150 to €8,000 and provides a rebate ranging According to a recent study of an incentive scheme from €150 to €6,300.2 The dealer can advance the bonus at scheduled to last through 2017 (Doyle and Adomaitis 2013), the Norwegian government provides tax breaks of up to 1 The fiscal year for the national Japanese budget cycle runs from $11,000 over the lifetime of a PEV, or about $1,400 per year. June to May. The tax breaks include no purchase tax, no annual registra- 2 For more specific break-downs on the bonus-malus system, see: tion tax, and no value-added tax (VAT) (Doyle and Adomai- http://www.developpement-durable.gouv.fr/Bonus-Malus-2014. tis 2013). As part of the scheme, commuters do not pay road html. 136

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 152 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

Appendix C 137 tolls, worth $1,400 annually, and they receive free parking then he or she must pay income tax on the total. The bijtel- worth $5,000. PEVs are permitted in bus lanes and have ling tax is assessed on the basis of grams of CO2 per kilome- access to free public charging at 466 parking spots in Oslo ter, and for high-emitting vehicles, the tax rate is 25 percent. (Doyle and Adomaitis 2013). For BEVs, the bijtelling tax rate is 4 percent. BEV buyers As of the beginning of 2013, PEV sales accounted for 3 also enjoy a purchase tax incentive, whereby through 2017 percent of total passenger car sales, a much higher fraction they pay no tax for vehicles with low CO2 emissions and than in most countries. A total of 12,000 PEVs had been sold are exempt from a vehicle-use tax, which is normally based in Norway as of 2013, with about half in the Oslo region on weight and kilometers driven. Employees are therefore (Ingram 2013a). Nonetheless, some (for example, Doyle and motivated to encourage their employers to buy them BEVs. Adomaitis 2013) have criticized the incentive program be- The federal government is also providing a purchase sub- cause it encourages families to purchase a PEV as a second sidy for BEV taxis and delivery vans used in urban areas to car and rely on their gasoline- or diesel-powered vehicle for help cope with urban air pollution and noise. Amsterdam, longer-range trips. However, even if that is the practice, fam- Arnhem, The Hague, Rotterdam, and Utrecht add an addi- ilies might be driving more electric miles during the course tional purchase subsidy for taxis and delivery vans (€5,000) of everyday life. Although the programs could prove to be an and trucks (€40,000) and are particularly motivated as no environmental benefit, Norway might not be able to sustain new construction may occur in the city until air pollution has such a financial commitment. It spends about $13,600 in tax been reduced (Dutch Ministry of Economic Affairs 2014). incentives to reduce CO2 emissions by just one tonne. This The incentives are especially important to consumers be- cost is much higher than the prevailing price of CO2 on the cause Dutch electricity prices are high (€0.28/kWh), so the European Union emissions trading market (Ingram 2013b). consumer incentive to use electricity as a fuel is minimal.

GERMANY CHINA

Germany does not currently offer consumer incentives Beginning in 2006, China made a major push toward and is instead relying on a demonstration program in four PEVs. Given China’s heavy reliance on coal to generate major regions. German vehicle manufacturers are investing electricity, the main environmental benefits for the country heavily in hydrogen fuel-cell vehicles (FCVs), and, other could be cleaner air in some cities and a reduction in noise than BMW, they have been slow to embrace PEVs. pollution. However, according to a recent analysis by Ji et al. (2012), replacing gasoline vehicles with PEVs in China with

THE NETHERLANDS its current electricity supply mix will result in higher CO2 emissions and increased mortality risk from PM2.5 (particu- The Netherlands has extensive consumer incentives for late matter less than 2.5 micrometers in diameter) in most PEVs and at one time these incentives equaled as much as Chinese cities. In any case, the Chinese government views 85 percent of the price of a new plug-in hybrid electric ve- a shift to PEVs to be beneficial to China’s energy security. hicle (PHEV), although they have since been scaled back. China is a net importer of coal and its current reserve-to-pro- Unsurprisingly, the Netherlands has become a hot market duction ratio of coal is only 31 years (BP 2013). The energy for PEV manufacturers and is Tesla’s second biggest market security benefits are therefore not apparent. As of March besides the United States after Norway. The Dutch govern- 2013, there were about 28,000 PEVs registered in China, of ment is especially motivated to support BEVs because most which about 80 percent were public buses. of the larger cities in the Netherlands experience severe ur- As of 2010, there were 135 million electric bicycles in ban air pollution. Municipal governments are also keen to China (Jie and Hagiwara 2013).3 China is already the largest reduce urban noise, especially in the evenings, and find that electric bicycle producer and consumer, accounting for about noise reduction from BEV taxis and delivery vans greatly 90 percent of the global market. The Chinese government improves the quality of city life (Nissan 2012). The Dutch research and development program for clean, light-duty ve- government also views BEV deployment as consistent with hicles initially focused almost equally on FCVs, BEVs, and its climate change goals and strategy. Not having significant PHEVs. In China’s Five-Year Plan (2006-2010), however, domestic vehicle production, there is little resistance to im- the government’s emphasis shifted strongly to BEVs. porting BEVs from abroad. The Chinese central government has subsidized the de- The tax incentive structure is unique among all the ployment of PEVs since 2009. Some local governments in countries examined because corporate buyers overwhelm- 25 pilot cities also provided subsidies on top of the central ingly dominate the Dutch new-vehicle market, and most new government subsidies discussed below, mostly to support the vehicles are bought by firms for their employees. Employees must pay income tax (bijtelling) for vehicles received from 3 As of 2008, 970 invention patents had been applied for through their employers. For example, 25 percent of the value of a the State Intellectual Property Organization (SIPO) based on the new vehicle is added to an employee’s personal income, and research of the Chinese government’s Energy-Saving and New En- ergy Vehicle Programme (Ouyang 2009).

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-3; Source: Overcoming Barriers to Deployment of Plug-in Electric Vehicles Page 153 of 153 Overcoming Barriers to Deployment of Plug-in Electric Vehicles

138 Overcoming Barriers to Deployment of Plug-in Electric Vehicles purchase of public transportation vehicles, such as buses and norway.html. locally stated-owned taxis. It has been alleged that some lo- Doyle, A., and N. Adomaitis. 2013. Norway shows the way cal governments have imposed “buy local” provisions so that with electric cars, but at what cost? Reuters, March 13. the local PEV firms benefit at the expense of PEV companies http://www.reuters.com/article/2013/03/13/us-cars-nor- elsewhere in China and around the world (Zeng 2013). way- idUSBRE92C0K020130313. The Chinese government allowed six cities to experi- Dutch Ministry of Economic Affairs. 2014. Electromobility ment with subsidies to individual consumers who purchased in the Netherlands: Highlights 2013. http://www.rvo.nl/ PEVs starting in 2013.4 In those cities, the local government sites/default/files/2014/04/Electromobility%20in%20 is allowed to provide purchase incentives, and the central the%20Netherlands%20Highlights%202013.pdf. government will also provide up to RMB 50,000 (about Ingram, A. 2013a. “Electric Cars: 12 Percent of All New-Car $8,000) for the purchase of a PHEV and RMB 60,000 (about Sales In Norway Last Month.” Green Car Reports, Decem- $9,600) for the purchase of a BEV. Beijing has announced ber 3. http://www.greencarreports.com/news/1088856_ that it will also subsidize BEVs at a rate of RMB 60,000 electric-cars-12-percent-of-all-new-car-sales-in-nor- (about $9,600), while Shanghai will provide a subsidy of way-last-month. RMB 20,000 (about $3,000) for a PHEV and RMB 50,000 Ingram, A. 2013b. “How Much Is Norway Paying to Pro- (about $8,000) for a BEV. Changchun will offer RMB 40,000 mote Electric Cars?” Green Car Reports, March 15. (about $6,400) for a PHEV and RMB 45,000 (about $7,200) http://www.greencarreports.com/news/1082915_how- for a BEV. Shenzhen will offer RMB 30,000 (about $4,800) much-is-norway-paying-to-promote-electric-cars. for a PHEV and RMB 60,000 (about $9,600) for a BEV. He- Ji, S., C. Cherry, M. Bechle, Y. Wu, and J. Marshall. 2012. Elec- fei has not yet set individual rates but has set aside a budget tric vehicles in China: Emissions and health impacts. Envi- of RMB 800 million (about $128 million) for subsidies. To ronmental Science and Technology 46(4): 2018-2024. qualify for the subsidies, there are minimum battery require- Jie, M., and Y. Hagiwara. 2013. “In Ghosn We Trust Tested as ments (at least 15 kWh for a BEV and at least 10 kWh for Nissan Electric Push Falters.” Bloomberg, March 20. http:// a PHEV). As of 2014, fewer than 70,000 PEVs were on the www.bloomberg.com/news/2013-03-20/in-ghosn-we- road in China, far from the target set by the government of trust-tested-as-nissan-electric-push-falters.html. 500,000 by 2015 (Bloomberg News 2014). METI (Ministry of Economy, Trade, and Industry). 2013. “Our Policy about Promoting EVs.” Presentation to the REFERENCES Committee on Overcoming Barriers to Electric-Vehicle Deployment, Tokyo, Japan, December 9. Beltramello, A. 2012. “Market Development for Green Cars.” Nelson, T., and M. Tanabe. 2013. “Japan Continues to Offer OECD Green Growth Papers. No. 2012-03. OECD Pub- Electric Vehicle Incentives.” Dashboard Insights, Septem- lishing, Paris. ber 12. http://www.foley.com/japan-continues-to-offer- Bloomberg News. 2014. “China to Exempt Electric Cars From electric-vehicle-incentives-09-12-2013/. 10% Purchase Tax.” Bloomberg Business Week, July 9. Nissan. 2012. “Amsterdam’s TAXI-E Takes Zero Emission BP. 2013. “Statistical Review of World Energy.” http://www. Push to the Streets.” http://reports.nissan-global.com/EN/? bp.com/en/global/corporate/about-bp/energy-econom- p=3348. ics.html. Accessed July 16, 2014. Ouyang, M. 2009. “Development of EVs in China.” Presen- Brand, C., J. Anable, and M. Tran. 2013. Accelerating the tation at U.S.-China EV Forum, Beijing, China. Sep- transformation to a low carbon passenger transport sys- tember 28. tem: The role of car purchase taxes, feebates, road taxes, Tesla. 2013. “Electric Vehicle Incentives Around the World.” and scrappage incentives in the U.K. Transportation Re- http://www.teslamotors.com/incentives/japan. Accessed search Part A: Policy and Practice 49:132-148. February 14, 2013. Deshayes, P. 2011. “Electric Cars Take Off in Norway.” Phys. Zeng, J. 2013. “Subsidies for Green Cars Not Bringing Ex- org, May 10. http://phys.org/news/2011-05-electric-cars- pected Results.” China Daily, April 1. http://www.chi- nadaily.com.cn/business/greenchina/2013-04/01/con- 4 The six cities permitted to provide additional incentives are Bei- tent_16364018.htm. jing, Changchun, Hangzhou, Hefei, Shanghai, and Shenzhen.

Copyright © National Academy of Sciences. All rights reserved. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 1 of 8

Contents lists available at SciVerse ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

Electric vehicles and the electric grid: A review of modeling approaches, Impacts, and renewable energy integration

David B. Richardson n

Department of Geography & Program in Planning, University of Toronto, 100 St. George St., Room 5047, Toronto, Ontario, Canada M5S 3G3 article info abstract

Article history: Electric vehicles (EVs) and renewable energy sources offer the potential to substantially decrease Received 20 August 2012 carbon emissions from both the transportation and power generation sectors of the economy. Mass Received in revised form adoption of EVs will have a number of impacts and beneﬁts, including the ability to assist in the 9 November 2012 integration of renewable energy into existing electric grids. This paper reviews the current literature on Accepted 12 November 2012 EVs, the electric grid, and renewable energy integration. Key methods and assumptions of the literature Available online 11 December 2012 are discussed. The economic, environmental and grid impacts of EVs are reviewed. Numerous studies Keywords: assessing the ability of EVs to integrate renewable energy sources are assessed; the literature indicates Electric vehicle that EVs can signiﬁcantly reduce the amount of excess renewable energy produced in an electric Solar energy system. Studies on wind–EV interaction are much more detailed than those on solar photovoltaics (PV) Renewable energy and EVs. The paper concludes with recommendations for future research. Wind energy & Smart grid 2012 Elsevier Ltd. All rights reserved.

Contents

1. Introduction ...... 247 2. Electric vehicles ...... 248 2.1. Vehicles and energy sources ...... 248 2.2. Charging and grid connections ...... 248 3. EV-grid models...... 249 4. EV impacts and performance...... 249 4.1. Economic impacts ...... 249 4.2. Environmental impacts ...... 250 4.3. Grid impacts...... 250 5. EVs and renewable energy ...... 250 5.1. Wind energy...... 250 5.2. Solar energy ...... 251 5.3. Biomass energy ...... 252 6. Conclusion ...... 252 Acknowledgments ...... 253 References ...... 253

1. Introduction this century: peak oil, climate change, and energy independence. Electricity generation and transportation account for over 60% of The world’s transportation and electric power generation global primary energy demand; a majority of the world’s coal sectors are directly linked to some of the key driving issues of demand is for electricity generation and a majority of the world’s oil demand is for transportation [1]. Alternative vehicle technologies, such as electric vehicles (EVs), are being developed to n Tel.: þ1 647 457 1121; fax: þ1 416 946 3886. reduce the world’s dependence on oil for transportation and limit E-mail address: [email protected] transportation related CO2 emissions. Likewise, renewable energy

1364-0321/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.rser.2012.11.042 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid 248 D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 2 of 8 sources are being developed and deployed to displace fossil fuel Fuel cell vehicles (FCVs) are another type of electric vehicle, in based electricity generation, reducing greenhouse gas emissions that the fuel cell generates electricity through an electrochemical as well as the emission of other pollutants such as nitrous oxides process in the fuel cell stack. FCVs have an onboard fuel source,

(NOx) and sulfur dioxide (SO2). The integration of the transporta- such as natural gas or hydrogen, and can either be fully reliant on tion and electricity sectors, in combination with EVs and renew- the fuel cell or designed with a battery in a hybrid arrangement able energy, offers the potential to significantly reduce the like an HEV or PHEV. Future visions of a hydrogen economy world’s dependence on fossil fuels and the consequent emission involve the use of FCVs for transportation; if the hydrogen is of greenhouse gases. created through the electrolysis of water using renewable elec- There are a number of barriers to the large-scale integration of tricity or from biomass sources then the FCVs would be utilizing renewables into the electricity system [2]. Renewable energy renewable sources as well. Currently, the vast majority of the sources, such as wind and solar photovoltaic (PV) electricity, tend world’s hydrogen is produced from fossil fuel sources and the to be variable in supply with no correlation to changes in demand. creation of a sustainable hydrogen economy still faces a number Whereas natural gas turbines can be ramped up and down to follow of hurdles [3]. While hydrogen from electrolysis is an important fluctuations in demand, renewable energy sources like wind and potential use for renewable electricity, the transition to a hydro- solar are only available when the wind is blowing or the sun is gen economy is too broad a topic to be considered in this paper shining. A variety of strategies have been developed to manage EV technologies offer opportunities for a transportation sector supply fluctuations of varying timescales; these include storage, powered by renewable energy. To the extent that traditional dispatchable loads (or demand response), and alternative generating transportation fuels can be replaced by sustainably grown bio- capacity [3]. Electric vehicles with an electric grid connection can fuels, such as ethanol or biodiesel, HEVs can be run from renew- support all of these strategies; therefore the wide-spread adoption able energy sources. PHEVs can also use biofuels in their internal of EVs could play an important role in the integration of renewable combustion engine, while both PHEVs and BEVs can be comple- energy into existing electricity systems [4]. tely operated with renewables if charged with renewable elec- The basic goal of this paper is to review and assess the tricity from the grid. As such, the vehicle technologies that will be literature that discusses the impacts of electric vehicles on the considered for this paper are those with the capacity to store electric grid, with the main focus on the integration of renewable electrical energy from the grid: PHEVs and BEVs (from here on energy into the electricity system. Section 2 gives an overview of referred to jointly as EVs). EVs, including the key concepts that are pertinent to vehicle interaction with the grid. Section 3 discusses and compares the 2.2. Charging and grid connections modeling approaches used in the literature to analyze EVs and the electric grid. Some general impacts and benefits of EVs on the The battery of an electric vehicle can be recharged from the electricity system are presented in Section 4. Section 5 gives a grid with varying measures of external control, labeled here as more thorough review of the literature on electric vehicles and charge plans. A simple, or unconstrained, charge plan is a system renewable energy integration. Section 6 concludes by summariz- in which the vehicle immediately begins recharging as soon as it ing the key results of the review and identifying some key is connected to the grid. A delayed charge plan offsets the battery knowledge gaps that could inform future research projects. charging by a set amount of time, for example three hours. Nighttime charge plans delay charging to occur over the course of the night when electricity prices are lowest, with the battery 2. Electric vehicles fully charged for use in the morning. Smart charging implies some measure of intelligent control over the charging of the vehicle by 2.1. Vehicles and energy sources the utility or system operator. This can either be direct charging, through direct control of the vehicle, or indirect charging by An electric vehicle will be defined, for the purposes of this designing the vehicle to respond to price signals. Dallinger and paper, as any vehicle in which some or all of the driving energy is Wietschel [6] suggest that indirect charging is a more promising supplied through electricity from a battery. In a conventional concept as it is more likely to lead to consumer acceptance than internal combustion engine vehicle (ICEV), gasoline or diesel fuel direct external control. is combusted to create mechanical energy that provides the The idea behind smart charging is to charge the vehicle when power to move the vehicle forward. A number of EV technologies it is most beneficial, which could be when electricity is at its are currently in use or under development, as discussed in lowest price, demand is lowest, when there is excess capacity, or Jorgensen [5]. A hybrid electric vehicle (HEV) has a small electric based on some other metric. The rate of charge can be varied battery that supplies electricity to the drivetrain in order to within certain limits set by the driver; the most basic limit being optimize the operating efficiency of the combustion engine. The that the vehicle must be fully charged by morning. Lunz et al. [7] battery in an HEV can be charged by the engine or through suggest that one focus of smart charging should be to manage captured kinetic braking energy from a process called regenera- battery performance and lifetime, which can improve the lifetime tive braking. HEVs are more fuel efficient than ICEVs, but economics of the battery. ultimately the vehicle is fully powered by liquid fuels. A plug-in A vehicle-to-grid (V2G) capable EV is one that is able to store hybrid electric vehicle (PHEV) is similar in concept to an HEV, but electricity and then return it to the electric grid. V2G power is an with a larger battery and a grid connection. The grid connection interesting concept that was first proposed by Kempton and allows the battery to be charged with electricity and the larger Letendre [8]. The authors suggested that V2G could be used to battery size enables the car to drive a significant distance in all- generate a profit for vehicle owners if the power was used under electric mode. An all-electric range of twenty miles can be certain conditions to provide valuable services to the electric grid. denoted through the notation PHEV-20, and a forty mile These services include regulation (second by second balancing of all-electric range would be PHEV-40. A battery electric vehicle demand and supply), spinning reserve, and peak power provision. (BEV) is fully powered by grid electricity stored in a large onboard The energy could in theory be supplied from the battery of a BEV battery. EVs use energy much more efficiently than ICEVs; a or PHEV, from the engine of a PHEV in generator mode, or from traditional ICEV fuel efficiency is 15–18%, while a BEV can be as the fuel cell of an FCV [9]. Pang et al. [10] suggest that vehicle-to- high as 60–70% efficient [5]. building (V2B) technology is closer to being a viable option than July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 3 of 8 249

V2G; under a V2B scenario the EV would offer demand manage- compared with multiple EV penetration levels to produce a ment and outage management services to the building. A V2G/ sensitivity analysis for the effects of EVs. V2B capable EV could store renewably generated electricity The influence of charge plans and different charging scenarios during periods of low demand or excess supply and provide it is addressed in a number of ways in the literature. Some studies back to the grid, or building, when required. focus exclusively on simple charging [27] or smart charging [16,21]. Others compare just two charging strategies, such as simple charging versus delayed charging [28]. The most prevalent modeling approach is to compare a wide range of charging 3. EV-grid models strategies, with some combination of simple, delayed, nighttime, smart, and smart with V2G being considered [4,23,29]. The body of literature that discusses EVs and the electric grid The availability of accurate and detailed driving patterns is a is primarily based on models, likely due to a scarcity of actual major issue in the creation of accurate and useful EV/electric grid systems of an appropriate size for study. A few single vehicle models [26]. In order to know the available EV storage and proof-of-concept tests have been conducted for V2G vehicles discharge capacity at any one time it is necessary to know the [11,12], but no systems level empirical evaluations have been number of available (parked and plugged in) vehicles and the carried out. The models used in the literature can be broadly amount of energy stored in each battery. The battery energy level, divided into two categories: (1) long-term, system scale planning or state-of-charge (SOC), depends on how far the vehicle has been models and (2) hourly time-series models. driven since it was fully charged, the vehicle efficiency, and the The long-term planning models are run on a regional or battery size and characteristics. From a computational perspec- national scale over the course of many decades and generally tive it is too difficult to model the driving habits and battery SOC optimize the mix of electric generating units in a system given a of each individual vehicle, so some models choose to aggregate all set of boundary conditions, which can include the integration of the vehicle batteries into one large unit and use historical data to EVs. Time-series models take hourly historical data of electricity predict vehicle availability [4,23]. Kristofferson et al. [21] state supply and demand as well as driving data in order to assess the that this method allows the vehicles to charge faster than would ability of the system to match supply and demand over the short actually be realistic; therefore they construct 30 aggregate driving and medium term. These models are generally run over the patterns out of historical data in order to more accurately course of one week up to one year. represent the vehicle fleet. Wang et al. [29] aggregate vehicles There are other discussions of EVs and the electricity system by size and battery range. Whatever the method, a balance that do not involve the use of models, including a review of socio- between computational ease and real-world accuracy must be technical barriers to mass adoption [13], an outline for a transi- found, and many authors insist that more work must be done to tion path and aggregator framework [14] and a discussion of produce accurate vehicle and energy availability functions. business models and policy options for grid integration [15]. These papers provide insight into the practical considerations for integrating the transportation and electricity sectors. How- 4. EV impacts and performance ever, at this point in time the most effective way to predict the impacts of EVs on the electrical system is through models; There is clearly a wide range of input parameters that can be therefore some of the key model inputs and constraints in the used for the models, and there is an equally wide range of output literature will now be discussed. variables that the models can measure. The model outputs can be One of the more interesting input variables for the models is broadly categorized as economic, environmental, grid perfor- the selection of PHEVs versus BEVs. Of the 42 studies that mance, and renewable energy-related. Some key findings for the included detailed analysis which were reviewed for this paper, first three categories of outputs will be briefly discussed in this 18 (43%) analyzed PHEVs exclusively, 15 (36%) analyzed BEVs, section before a more thorough and comprehensive review of and 9 (21%) of the studies analyzed both vehicle types. In the studies that focused on the integration of renewable energy into PHEV studies, different all-electric ranges could be modeled [16]. the electricity system is conducted in Section 5. In the BEV studies, BEVs were often selected in order to use data and characteristics from actual production vehicles [17,18]. 4.1. Economic impacts In two of the studies that modeled both BEVs and PHEVs, the models were allowed to endogenously favor vehicle technologies TheeconomicimpactsofEVsaregenerally examined from two given some optimization constraints; in both cases the models perspectives: that of a vehicle owner, and that of the electricity chose PHEVs as preferable to BEVs [19,20]. In a different study, system. The lifecycle economics of EVs are expected to improve with BEVs were assumed to have much lower daily driving ranges than better battery technology and mass production. Currently, BEVs cost PHEVs [21]. From this, it can be seen that PHEVs are slightly more to purchase than PHEVs, and both cost more than traditional preferred by researchers over BEVs, likely due to lower battery ICEVs [30]. However, the fuel and operating costs of EVs are much costs and an extended driving range. lower than ICEVs, due to the high efficiency of an electric motor. Thiel The penetration, or market share, of EVs in the models was et al. [31] estimate that the payback time for a BEV, in comparison to selected in a number of different ways. A few studies chose a set a cheaper ICEV, is currently 20 years, but it should drop to less than number of EVs and analyzed the effects of this level of penetration five years by 2030. A similar result is found by Faria et al. [32],while in isolation [4,22]. Others evaluated a range of EV penetration Contestabile et al. [33] estimate that the total lifetime costs of scenarios [21,23]. Goransson et al. [24] took a novel approach, ownership for all vehicle types will converge by 2030. analyzing varying fractions of the total electricity demand that is A growing body of studies has assessed the economic viability of attributed to EVs. The literature predicting the future market V2G participation in different markets [18,34–38]. Yearly profit from share for EVs offers a wide range of values and scenarios, these studies ranges from a loss of $300 per vehicle per year to a providing scant insight to modelers on an appropriate penetration profit of over $4600, with most indicating a profit in the $100–300 level to select [25]. Green et al. [26] suggest that models should range. This level of profit may not be enough to induce participation, have one scenario run with a 0% market share for EVs, which either by individuals or by an aggregator organization. Assuming that would be considered as a baseline scenario. This could then be both EVs and V2G power services are technologies that governments July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid 250 D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 4 of 8 want adopted, the question remains of how best to encourage both transmission capacity. When the vehicles use a smart charging consumer and business participation in these markets from a policy plan, studies indicate that EVs levelize the overall load, make perspective. This has been discussed for the Ontario, Canada market better use of baseload units, and require no extra installed by Richardson [39]. capacity [21,46]. Hajimiragha et al. [47,48] estimate that In general, adding EVs to the electrical grid will increase 500,000 PHEVs that could be introduced in Ontario, Canada system costs due to increased fuel use and transmission losses, without adversely affecting the electric grid. In the US, the however the selection of charging strategies greatly influences the existing grid capacity would allow for 73% of the light-duty fleet overall effect on system costs. Kiviluoma and Meibom [22] to be converted to PHEVs [49]. estimate a system cost in Denmark of h263/vehicle/year using a Other impacts of EVs on the distribution grid include increased simple charge plan, while smart charging vehicles have a system wear on transformers, transmission bottlenecks and power quality cost of h36/vehicle/year, a savings of h227/vehicle/year. Similarly, issues; these and other more technical issues are thoroughly Wang et al. [29] conclude that smart charging saves $200,000/ reviewed by Green et al. [26]. There are conflicting findings on the week compared to simple charging in a future Illinois electricity effect of EVs on distribution networks. Ma et al. [50] find that EVs system with a high share of wind energy. Lynch et al. [40] andV2Gcanbecontrolledinsuchawayastohaveaminimal calculate the system savings for the PJM and Midwest ISO (MISO) impact on distribution system losses and voltage fluctuations. Other markets in the US, finding that the savings from off-peak charging studies have developed distribution level charge plans designed to versus on-peak are highly dependent on the regional generating maintain power quality and avoid distribution congestion problems mix. For the MISO, savings from smart charging are small due to that could result from widespread adoption of EVs [51,52]. Gong an excess capacity of coal generation, while much larger savings et al. [53] assess the effects of EVs on a residential distribution are realized in PJM as this market has a high reliance on more transformer; the effects are negligible at low EV penetrations, but expensive peaking natural gas plants. A study of the Spanish there is excess equipment wear with increasing vehicle numbers. A electric grid indicates that the marginal cost of electricity is study on UK distribution systems by Papadopolous et al. [54] reduced up to a certain target level of EV penetration, beyond indicates that high numbers of EVs lead to voltage limit violations, which the marginal cost slightly increases [41]. transformer overloads and increased line losses. They suggest that network reinforcements, embedded generation and EV charge 4.2. Environmental impacts management strategies are needed to safely integrate large numbers of EVs into distribution networks.

CO2 emissions are the most commonly measured output used to assess the environmental impacts of switching to EVs powered by the grid. Juul and Meibom [20] calculate that the integration of 5. EVs and renewable energy the electric power and transportation sectors in Denmark reduces transportation related CO2 emissions by 85%. Lund and Kempton The ability of EVs to assist the integration of renewable energy [4] ﬁnd that the use of EVs decreases CO2 emissions compared to sources into the existing power grid is potentially the most ICEVs even if there is no wind energy present in the generation transformative impact on the electricity system. The literature mix. Hadley [28] examines the introduction of EVs in Virginia and on this subject is primarily focused on the analysis of wind energy the Carolinas in the US where fossil fuel plants account for two- and solar energy, with wind energy receiving much greater thirds of total generation capacity; even under a simple charge attention and more detailed analysis. A few papers have com- plan EVs reduce CO2 emissions by roughly 10% compared to the pared the use of biomass energy for electricity in EVs as opposed base case with gasoline vehicles. Goransson et al. [24] examine a to biofuels. The models that study the interaction between wind-thermal power system and found that CO2 emissions slightly renewable energy, generally wind, and EVs tend to measure the increase under a simple charging strategy but decrease with smart amount of renewable capacity that EVs can accommodate, or the charging and V2G power. EVs were analyzed in three regions of effect on system performance that results from integrating EVs

China, with CO2 reductions occurring under all scenarios, even in into an electricity system with a large fraction of renewable regions that rely heavily on coal power [42]. generation. Results from a number of studies on the integration

There is debate over what emissions intensity (gCO2e/kWh) of wind energy and EVs will first be discussed, beginning with a should be assigned to the electricity used by electric vehicles when look at the large system scale models and then the hourly time- charging. Most studies use the average grid intensity to reflect a series models. This is followed by an overview of the work that situation in which EVs are widely adopted and should be considered has been done concerning solar energy and EVs, and a discussion as part of the everyday demand profile. Other authors argue for the of bioelectricity in comparison to biofuels. use of marginal intensity (as discussed in Ma et al. [43]), in which It should be noted that a few papers examine the impacts of the emissions of the marginal generating unit are assigned to EV EVs on electricity systems with a large share of wind, but the electricity. In most markets, especially at times of peak demand, this results and discussion are focused on costs [29] or carbon will be a natural gas or coal power plant which will lead to larger emissions [24], and do not directly address renewable energy carbon emissions from EVs. However, even studies that utilize the integration. As such, these papers are not discussed here. marginal mix in different regions find a net carbon benefit in comparison to the use of ICEVs [44,45]. In general, it can be seen 5.1. Wind energy that EVs reduce total CO2 emissions even in electricity systems with a high fraction of fossil fuel generation, due to the high efficiency of A number of studies examine the large-scale, long-term an electric motor in comparison to an internal combustion engine. impact of EVs on the ability of electric grids to integrate wind energy. Short and Denholm [16] model the effect of large-scale 4.3. Grid impacts adoption of PHEVs on the integration of wind energy into the US electricity mix. Installed wind capacity increases by 243 GW, or EVs affect the performance, efficiency, and required capacity of 6% of total generation, when the vehicle fleet is converted to 50% the electric grid, especially if vehicle charging is unconstrained. PHEVs under a smart charging plan. Hadley [28] found that peak loads will increase under a simple Turton and Moura [19] use a global energy model that charging strategy, requiring extra investment in generation and forecasts the integration and impacts of EVs and V2G over the July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 5 of 8 251 period from 2000 to 2100. The authors find that the installed with charged batteries. These facilities offer an excellent oppor- renewable energy capacity increases by 30–75% with V2G capable tunity for wind energy integration as they would house a large EVs due to their ability to store intermittent energy and discharge number of battery packs onsite; the battery charging could be it back to the grid when required. managed to account for fluctuations in wind power output. Juul and Meibom [20] examine the integration of the electric Hodge et al. [27] simulate electricity supply and demand over power and transportation systems in Denmark for the year 2030 a six month summer period in California. The output variable from the perspective of power plant investments. The integration measured is the fraction of power supplied from renewable and of these two sectors, through the use of EVs, results in a wind sources. The results indicate that the addition of V2G has significant increase in offshore wind power capacity and a little effect on the fraction of energy supplied by wind, especially decrease in combined cycle natural gas plants, biomass electricity as higher penetrations of wind and other renewable sources are plants, and onshore wind power capacity. Overall, the increase in reached. This paper uses a ‘realistic’ scenario, with low levels of wind power generation exceeds the total demand for energy by PHEV integration and simple charging, which has been shown to the transportation sector. be the least effective manner in which to integrate EVs into the Borba et al. [55] take an interesting approach to modeling EVs grid. It offers no base case with zero PHEV integration, so there is and wind energy. The Brazilian power sector is modeled from no measure of how PHEVs on their own influence the integration 2010 to 2030, with an assumed 16-fold increase in wind generat- of wind power. ing capacity in the northeast. The authors then calculate the size Finally, one of the more detailed and complex studies on EVs of PHEV fleet that could be charged using the excess wind energy and renewable energy is presented by Dallinger and Wietschel [6]. production. Since the excess production varies seasonally, occur- It is the only paper reviewed so far to consider the combination of ring primarily between January and June, the authors assume that wind and solar PV, which in a projected German electricity mix the vehicles drive on locally produced ethanol for the remainder comprise 50% of capacity in 2030. The paper examines the of the year. Over 1.6 million vehicles could be powered in this influence of charging strategies, but uses an indirect charging manner by 2030. strategy based on consumer price response; as such, pooled Other studies assess the EV-wind energy interaction on an groups of vehicles must make price forecasts and compare hourly timescale. Kempton and Tomic´ [56] evaluate the use of predicted energy costs to vehicle demand. The model also electric vehicles to provide regulation and reserve services in considers grid fees in order to avoid overloading distribution conjunction with wind power in the US. They calculate that in a system transformers. The paper finds that EVs can make a positive scenario with 50% power production from wind, 3.2% of the US contribution to balancing renewable sources, with over 50% of the vehicle fleet would need to be fully electric to provide the yearly excess renewable production being absorbed by EVs. required regulation services and 38% of the fleet would be needed Overall, these studies indicate that the introduction of EVs has to provide operating reserves. The paper also looks at historical the potential to increase the amount of wind energy capacity time-series wind data from eight dispersed sites in the Midwest. installed in a regional or national electricity system. More Out of 6919 h of data, they identify 342 low power events where specifically, EVs can absorb excess wind energy production that power output would be less than 20% of capacity. BEVs have the would otherwise be wasted or curtailed, which improves the ability to provide enough reserve power for events less than two economics of wind energy generation. The larger the cumulative hours in duration, but they cannot cover longer-scale shortages, EV storage capacity, either through more vehicles or larger which can last fourteen to twenty-two hours. batteries, the more wind energy that can be accommodated into A study by Ekman [23] assesses the relationship between wind the system. Furthermore, EVs with V2G power can supply this energy production, power consumption, and electric vehicle energy back to the grid, which allows a further integration of charging patterns in Denmark. Smart charging EVs are found to wind energy into the generation mix. It should be noted that the significantly reduce the excess wind energy generation, with the marginal benefit of V2G in comparison to the overall benefits of potential to reduce the required amount of non-wind backup EVs appear to be relatively small. capacity, depending on the vehicle penetration level. The model demonstrates that EVs can be used to aid in the integration of 5.2. Solar energy wind energy, but cannot completely make up for intermittent supply from wind. The literature on solar energy and EVs is much more diverse Lund & Kempton [4] evaluate the integration of wind power than the previously reviewed studies that integrate wind energy and BEVs in national energy systems, in this case Denmark and a and vehicles. Electricity from solar PV can be produced anywhere; hypothetical country identical to Denmark that does not use this provides more interesting methods to directly integrate combined heat and power (CHP) plants. In the CHP case with energy production and use in EVs. A few representative studies 50% wind penetration, changing from 100% ICEVs to 100% BEVs of these varying approaches are discussed. However, it should be with V2G reduces excess wind electricity by a factor of two. The noted that the depth of analysis in this field is not nearly as strong introduction of EVs in the non-CHP system has a larger effect as with wind and electric vehicles. because CHP is more efficient with fuel use and contributes to Birnie [59] proposes the idea of solar PV arrays built over excess electricity. The study also finds that larger vehicle batteries parking lots to provide daytime charging to commuter vehicles. reduce the amount of excess wind electricity produced. The paper broadly sketches out what such a system would look A paper by Bellekom et al. [57] examines the introduction of like and assesses the potential energy production from a single wind power and EVs, both separately and combined, in the Dutch parking space. Assuming New Jersey solar irradiation data, a PV electricity system. Four gigawatts of wind can be introduced module efficiency of 14%, and a 15 m2 parking space, the average without problems under the no EV scenario; this increases to summertime production would be 12.6 kWh and the winter 10 GW with the introduction of 1 million EVs. Smart charging is average would be 3.78 kWh. This would be enough to meet most deemed necessary to avoid capacity problems. driving needs in the summer, but not during the winter. The A model that coordinates the operation of an EV battery paper does not assess the economic practicality of such a system, switching station with wind energy production is presented by the utility system benefits, or the environmental benefits. Gao et al. [58]. Battery switching stations are proposed as a Li et al. [60] conduct a similar analysis based on data from method to extend the range of EVs by replacing depleted batteries Alberta, Canada. The authors calculate the size of solar PV panel July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid 252 D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 6 of 8 needed to provide the daily energy requirements to a PHEV-40. were modeled as a means to meet domestic hot water demand They find that 20 m2 of panel would be sufficient to provide while using excess PV electricity. The study indicates that if electricity for 40 miles of range on the best sunshine days in mid- 30GW of solar capacity was installed in the area, just over July. However, 78 m2 of PV panel would be necessary to provide 10 TWh of annual production would be in excess. One million enough electricity in December. The oversized PV panel would EVs and 1 million HPs could reduce this excess production by produce 67 kWh of excess electricity on the best summer days, approximately 30%, while five million of each could absorb which could be sold to the grid to offset the costs of the panel. virtually all the excess production. Though HPs were never Large-scale deployment of parking lot solar car chargers is modeled separately from EVs, the marginal benefit of adding analyzed by Neumann et al. [61]. This study introduces solar car EVs to the system appears to be greater than that of HPs. ports over all the available large parking lots in a medium-sized Swiss city; the authors find that 14–50% of the city’s passenger 5.3. Biomass energy transportation energy demand could be provided through solar energy under the proposed system. Biomass energy differs from wind and solar in that it can be PV parking lot charging and other business models to charge stored and used when needed. Liquid biofuels are most EVs with solar energy are discussed by Letendre [62]. Parking lot commonly proposed for use as an alternative vehicle fuel, but chargers could be grid-connected or stand-alone units, sized to bioelectricity offers a number of advantages over biofuels. Bioe- meet a daily PV demand. The business models are not analyzed in lectricity can be obtained using a number of biomass feedstocks, detail in this paper. Solar PV is estimated to be a cheaper fuel per including forestry and agricultural residues, woody energy crops, vehicle kilometer than gasoline, especially as PV module prices and whole tree harvesting. These feedstocks can be directly decrease and gasoline stays around $4/gallon (USD) or higher. combusted, or co-fired with coal, in a boiler, or they can be Gibson and Kelly [63] and Kelly and Gibson [64] examine the gasified into a syngas and used in a simple turbine or combined technical feasibility of directly charging vehicle batteries with cycle power plant. Bioelectricity fuel pathways tend to give a solar PV panels. This would allow EVs to be charged using higher energy return on energy investment (EROEI) compared to electricity generated on-site, avoiding transmission losses from biofuel processes [3]. distant power plants or wind farms. Furthermore, converting DC A number of recent studies indicate that bioelectricity for use solar electricity to AC grid electricity results in energy losses of in a vehicle is a more effective use of biomass than conversion to around 10%; directly charging the batteries from PV panels avoids biofuels. Schmidt et al. [69] assess the production and use of those losses. These two studies provide a proof of concept for this multiple types of biofuels in Austria compared to bioelectricity; approach, demonstrating the safety and viability of this charging the results indicate that greenhouse gas emissions, land use scheme. effects, and the amount of required biomass feed stocks are all One method of taking advantage of direct battery charging reduced using electric vehicles as compared to biofuels. Campbell from solar PV would be to combine it with parking lot chargers, as et al. [70] find that the gross average driving output, in kilometers described earlier. Alternatively, PV systems could be mounted driven per hectare of biomass production, is 112% greater for directly on the vehicle as an auxiliary power source, also called bioelectricity than for biofuels. Furthermore, the average net vehicle-integrated PV (VIPV). This has been done by universities greenhouse gas offset for switchgrass production is 108% greater in solar car competitions for years, in which solar energy is the from bioelectricity than biofuels. Ohlrogge et al. [71] assert that primary power source for the vehicle [65]. Solar cars are not biomass for electricity in EVs displaces twice as much petroleum intended for commercial purposes; however VIPV could be used as biofuels. Electric vehicles using bioelectricity in Ontario, with existing hybrid and electric vehicles to improve efficiency. Canada are found to have a higher EROEI, and lower fuel costs Letendre [65] estimates this could improve vehicle efficiency by and GHG emissions in comparison with ICEVs using biofuels [72]. 10–20%, while Giannouli and Yianoulis [66] suggest that the Thus, given constraints on land availability for biomass produc- payback time for a VIPV system, in avoided fuel costs, would be tion, there is a clear benefit to producing electricity for transpor- just over four years. They also suggest that a VIPV system could be tation instead of biofuels. used for other purposes, such as running the vehicle air condi- tioner to keep the vehicle cool while parked. Letendre et al. [67] useaverysimplemethodtoestimatehow 6. Conclusion much firm capacity a combination of solar PV and V2G-enabled EVs couldprovideintheCaliforniamarket.Theideaisthatvehicle A number of positive impacts can be expected from the batteries would form a short-term buffer for PV output. The calcula- introduction of EVs, including lower vehicle operating costs, tions do not consider transportation demand or any system analysis. reduced CO2 emissions, and the ability to support and contribute Kempton and Tomic´ [56] discuss the use of PV solar electricity to grid power quality and stability if the right infrastructure is to supply peak energy in the US through storage in EVs with V2G adopted. Perhaps most significant, though, is the ability of EVs to power. Peak electricity production from PV panels occurs at mid- assist in the integration renewable energy sources into the day, a few hours before the daily peak in electricity demand. This electric grid. This has the potential to reduce the carbon emissions means that electricity generated at the solar peak would need to from both power generation and transportation. It should be be stored for a few hours before use to meet peak demand. noted that while EVs can substantially reduce some of the Assuming adequate PV capacity to supply all US peak supply negative impacts of large-scale renewable deployment, other (162 GW, or one-fifth of US generating capacity), they estimate methods and technologies are likely necessary to completely that 26% of the US vehicle fleet would be required to store the integrate a high penetration of renewable energy. peak solar electricity and then provide it to the grid a few The existing literature is fairly unanimous and conclusive in its hours later. assessment that EVs can increase the amount of renewable Only one study has been found that employs the more energy that can be brought online while reducing the negative rigourous methodology as the previously discussed studies on consequences for the grid. This is better documented, and more wind energy and EVs. Zhang et al. [68] analyze the integration of conclusive, for wind energy than for solar, as EVs can potentially PV power in conjunction with EVs and heat pumps (HP) in the be charged at off-peak times when otherwise unwanted wind Kansai Area of Japan. Air source heat pump water heater systems energy can be used to charge the vehicle batteries. This appears to July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 7 of 8 253 be the major contribution of EVs under current assumptions. V2G [17] Kempton W, Kubo T. Electric-drive vehicles for peak power in Japan. Energy plays a limited role in improving the penetration of renewables in Policy 2000;28:9–18. [18] Tomic´ J, Kempton W. Using fleets of electric-drive vehicles for grid support. the literature, most likely due to excessive battery degradation Journal of Power Sources 2007;168:459–68. which results in a relatively high cost of providing V2G power. [19] Turton H, Moura F. Vehicle-to-grid systems for sustainable development: an Another common theme that appears in almost every paper on integrated energy analysis. Technological Forecasting & Social Change EVs and the grid is the importance of some form of smart 2008;75:1091–108. [20] Juul N, Meibom P. Optimal configuration of an integrated power and charging. Smart charging reduces system costs by avoiding extra transport system. Energy 2011;36:3523–30. investment in peak generating units, transmission and distribu- [21] Kristofferson T, Capion K, Meibom P. Optimal charging of electric drive tion systems. Furthermore, it allows EVs to be used as distributed vehicles in a market environment. Applied Energy 2011;88:1940–8. [22] Kiviluoma J, Meibom P. Methodology for modeling plug-in electric vehicles in storage mechanism for absorbing excess renewable energy. While the power system and cost estimates for a system with either smart or dumb the savings are demonstrable, a comprehensive economic argu- electric vehicles. Energy 2011;36:1758–67. ment in favor of smart chargers has yet to be produced; this topic [23] Ekman C. On the synergy between large electric vehicle fleet and high wind penetration—an analysis of the Danish case. Renewable Energy is worthy of future research. The impacts and interaction between 2011;36:546–53. solar energy and EVs is another area requiring detailed analysis, [24] Goransson L, Karlsson S, Johnsson F. Integration of plug-in hybrid electric as solar energy charging stations could be a focus on future vehicles in a regional wind-thermal power system. Energy Policy 2010;38:5482–92. infrastructure investment. [25] Oak Ridge National Laboratory. A comparative study of emerging vehicle EVs offer many potential benefits to the electric grid, including technology assessments. Oak Ridge, TN: Oak Ridge National Laboratory; the ability to integrate intermittent renewable energy sources. 2011. [26] Green R, Wang L, Alam M. The impact of plug-in hybrid electric vehicles on It is important to understand the potential, limits, and impacts of distribution networks: a review and outlook. Renewable Sustainable Energy combining the transportation and electricity sectors through EVs Review 2011;15:544–53. and renewable energy. This can inform policies and infrastructure [27] Hodge B, Huang S, Shukla A, Pekny J, Reklaitis G. The effects of vehicle-to-grid planning in order to maximize the environmental and economic systems on wind power integration in California. Computer Aided Chemical Engineering 2010;28:1039–44. benefits of the two technologies, while at the same time reducing [28] Hadley S. Impact of plug-in hybrid vehicles on the electric grid. Oak Ridge, the world’s greenhouse gas emissions and its dependence on TN: Oak Ridge National Laboratory; 2006. fossil fuels. [29] Wang J, Liu C, Ton D, Zhou Y, Kim J, Vyas A. Impact of plug-in hybrid electric vehicles on power systems with demand response and wind power. Energy Policy 2011;39:4016–21. [30] Thomas C. Fuel cell and battery electric vehicles compared. International Journal of Hydrogen Energy 2009;34:6005–20. Acknowledgments [31] Thiel C, Perojo A, Mercier A. Cost and CO2 aspects of future vehicle options in Europe under new energy policy scenarios. Energy Policy 2010;38: The author would like to thank Dr. Danny Harvey and 7142–51. [32] Faria R, Moura P, Delgado J, de Almeida A. A sustainability assessment of Dr. William Gough of the University of Toronto, as well as two electric vehicles as a personal mobility system. Energy Conversion and anonymous reviewers, for their comments on this paper. Management 2012;61:19–30. [33] Contestabile M, Offer G, Slade R, Jaeger F, Thoenees M. Battery electric vehicles, hydrogen fuel cells and biofuels. Which will be the winner? Energy & Environmental Science 2011;4:3754–72. References [34] Andersson SL, Elotsson A, Galus M, Goransson¨ L, Karlsson S, Johnsson F, et al. Plug-in hybrid electric vehicles as regulating power providers: case studies of [1] IEA. Key world energy statistics. Paris: International Energy Agency; 2011. Sweden and Germany. Energy Policy 2010;38:2751–62. [2] Boyle G, editor. London: Earthscan; 2007. [35] Quinn C, Zimmerle D, Bradley T. The effect of communication architecture on [3] Harvey LDD. Energy and the new reality 2: carbon-free energy supply. the availability, reliability, and economics of plug-in hybrid electric vehicle- London: Earthscan; 2010. to-grid ancillary services. Journal of Power Sources 2010;195:1500–9. [4] Lund H, Kempton W. Integration of renewable energy into the transport and [36] Peterson S, Whitacre J, Apt J. The economics of using plug-in hybrid electric electricity sectors through V2G. Energy Policy 2008;36:3578–87. vehicles battery packs for grid storage. Journal of Power Sources [5] Jorgensen K. Technologies for electric, hybrid and hydrogen vehicles: 2010;195:2385–92. electricity from renewable energy sources in transport. Utilities Policy [37] Sioshansi R, Denholm P. The value of plug-in hybrid electric vehicles as grid 2008;16:72–9. resources. The Energy Journal 2010;31:1–23. [6] Dallinger D, Wietschel M. Grid integration of intermittent renewable energy [38] White C, Zhang K. Using vehicle-to-grid technology for frequency regulation sources using price-responsive plug-in electric vehicles. Renewable Sustain- and peak-load reduction. Journal of Power Sources 2011;196:3972–80. able Energy Reviews 2012;16:3370–82. [39] Richardson D. Economic analysis of vehicle-to-grid (V2G) business models [7] Lunz B, Yan Z, Gerschler J, Sauer D. Influence of plug-in hybrid electric vehicle and tariff structures in Ontario, Canada. Energy Policy. p. 32, submitted for charging strategies on charging and battery degradation costs. Energy Policy publication. 2012;46:511–9. [40] Lynch T, Michelin M, Jongejan A, Leahy T. Is smart charging policy for electric [8] Kempton W, Letendre S. Electric vehicles as a new power source for electric vehicles worthwhile? Energy Policy 2012;41:259–68. utilities. Transportation Research Part D 1997;2:157–75. [41] Fernandes C, Frıás P, Latorre J. Impact of vehicle-to-grid on power system [9] Kempton W, Tomic´ J. Vehicle-to-grid power fundamentals: calculating operation costs: the Spanish case study. Applird Energy 2012;96:194–202. capacity and net revenue. Journal of Power Sources 2005;144:268–79. [42] Wu Y, Yang Z, Lin B, Liu H, Wang R, Zhou B, et al. Energy consumption and [10] Pang C, Dutta P, Kezunovic M. BEVs/PHEVs as dispersed energy storage for CO2 emission impacts of vehicle electrification in three developed regions of V2B uses in the smart grid. IEEE Transactions on Smart Grid 2012;3(1): China. Energy Policy 2012;48:537–50. 473–82. [43] Ma H, Balthser F, Tait N, Riera-Palou X, Harrison A. A new comparison [11] Brooks A. Vehicle-to-grid demonstration project: grid regulation ancillary between the life cycle greenhouse gas emissions of battery electric vehicles service with a battery electric vehicle. Sand Dimas, CA, California: Air and internal comubstion vehicles. Energy Policy 2012;44:160–73. Resources Board and the California Environmental Protection Agency; 2002. [44] McCarthy R, Yang C. Determining marginal electricity for near-term plug-in [12] Kempton W., Udo W., Huber K., Komara K., Letendre S., Baker S. et al. A test of and fuel cell vehicle demands in California: impacts on vehicle greenhouse vehicle-to-grid (V2G) for energy storage and frequencey regulation in the gas emissions. Journal of Power Sources 2010;195:2099–109. PJM system. University of Delaware; 2008. [45] Elgowainy A, Burnham A, Wang M, Molburg J, Rousseau A. Well-to-wheels [13] Sovacool B, Hirsch R. Beyond batteries: an examination of the benefits and energy use and greenhouse gas emissions analysis of plug-in hybrid electric barriers to plug-in hybrid electric vehicles (PHEVs) and a vehicle-to-grid vehicles. Lemont, IL: Argonne National Laboratory; 2009. (V2G) transition. Energy Policy 2009;37:1095–103. [46] Denholm P, Short W. An evaluation of utility system impacts and benefits of [14] Guille C, Cross G. A conceptual framework for the vehicle-to-grid (V2G) optimally dispatched plug-in hybrid electric vehicles. Golden, CO: National implementation. Energy Policy 2009;37:4379–90. Renewable Energy Laboratory; 2006. [15] Andersen P, Mathews J, Rask M. Integrating private transport into renewable [47] Hajimiragha A, Can˜izares C, Fowler M, Elkamel E. Optimal transition to plug- energy policy: the strategy of creating intelligent recharging grids for electric in hybrid electric vehicles in Ontario, Canada, considering the electricity-grid vehicles. Energy Policy 2009;37:2481–6. limitations. IEEE Transactions on Industrial Electronics 2010;57:690–701. [16] Short W, Denholm PA. Preliminary assessment of plug-in hybrid electric [48] Hajimiragha A, Can˜izares C, Fowler M, Moazeni S, Elkamel A. A robust vehicles on wind energy markets. Golden, CO: National Renewable Energy optimization approach for planning the transition to plug-in hybrid electric Laboratory; 2006. vehicles. IEEE Transactions on Power Systems 2011;26:2264–74. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-4; Source: Richardson Paper on Electric Vehicles and the Grid 254 D.B. Richardson / Renewable and Sustainable Energy Reviews 19 (2013) 247–254 Page 8 of 8

[49] Kintner-Meyer M, Scheider K, Pratt R. Impacts assessment of plug-in hybrid [61] Neumann H, Schar¨ D, Baumgartner F. The potential of photovoltaic carports vehicles on electric utilities and regional U.S. power grids part 1: technical to cover the energy demand of road passenger transport. Progress in analysis. Richland, WA: Pacific Northwest National Laboratory; 2007. Photovoltaics: Research and Applications 2012;20:639–49. [50] Ma Y, Houghton T, Cruden A, Infield D. Modeling the benefits of vehicle-to- [62] Letendre S. Solar electricity as a fuel for light vehicles. In: Proceedings grid technology to a power system. IEEE Transactions on Power Systems of the 2009 American solar energy society annuual conference, Boulder, CO; 2012;27(2):1012–20. 2009. [51] Sundstrom O, Binding C. Flexible charging optimization for electric vehicles [63] Gibson T, Kelly N. Solar photovoltaic charging of lithium-ion batteries. considering distribution constraints. IEEE Transactions on Smart Grid Journal of Power Sources 2010;195:3928–32. 2012;3(1):26–37. [64] Kelly N, Gibson T. Solar photovoltaic charging of high voltage nickel metal [52] Richardson P, Flynn D, Keane A. Local versus centralized charging strategies hydride batteries using DC power conversion. Journal of Power Sources for electric vehicles in low voltage distribution systems. IEEE Transactions on 2011;196:10430–41. Smart Grid 2012;3(2):1020–8. [65] Letendre S. Vehicle integrated photovoltaics: exploring the potential. [53] Gong Q, Midlan-Mohler S, Marano V, Rizzoni G. Study of PEV charging on In: Proceedings of the 23rd annuual electric vehicles symposium. Anaheim, residential distribution transformer life. IEEE Transactions on Smart Grid CA; 2007. 2012;3(1):404–12. [66] Giannouli M, Yianoulis P. Study on the incorporation of photovoltaic systems [54] Papadopolous P, Skarvelis-Kazakos S, Grau I, Cipcigan L, Jenkins N. Electric vehicles’ impact on British distribution networks. IET Electrical Systems in as an auxiliary power source for hybrid and electric vehicles. Solar Energy Transportation 2012;2(3):91–102. 2012;86:441–51. [55] Borba B, Szklo A, Schaeffer R. Plug-in hybrid electric vehicles as a way to [67] Letendre S, Perez R, Herig C. Battery-powered electric-drive vehicles providing maximize the integration of variable renewable energy in power systems: buffer storage for PV capacity value. In: Proceedings of the 2002 American solar the case of wind generation in northeastern Brazil. Energy 2012;37:469–81. energy society annual conference. Boulder, CO; 2002. [56] Kempton W, Tomic´ J. Vehicle-to-grid power implementation: from stabilizing the [68] Zhang Q, Tezuka T, Ishihara K, Mclellan B. Integration of PV power into future grid to supporting large-scale renewable energy. Journal of Power Sources low-carbon smart electricity systems with EV and HP in Kansai Area, Japan. 2005;144:280–94. Renewable Energy 2012;44:99–108. [57] Bellekom S, Benders R, Pelgrom¨ S, Moll H. Electric cars and wind energy: two [69] Schmidt J, Gass V, Schmid E. Land use changes, greenhouse gas emissions, problems, one solution? A study to combine wind energy and electric cars in and fossil fuel substitution of biofuels compared to bioelectricity production 2020 in The Netherlands. Energy 2012;45(1):859–66. for electric cars in Austria. Biomass Bioenergy 2011;35:4060–74. [58] Gao Y, Zhao K, Wang C. Economic dispatch containing wind power and [70] Campbell J, Lobell D, Field C. Greater transportation energy and GHG offsets electric vehicle battery swap station. In: Proceedings of the IEEE power from bioelectricity than ethanol. Science 2009;324:1055–7. engineering society transmission and distribution conference; 2012. p. 1–7. [71] Ohlrogge J, Allen D, Berguson B, DellaPenna D, Schachar-Hill Y, Stymme S. [59] Birnie D. Solar-to-vehicle (S2V) systems for powering commuters of the Driving on biomass. Science 2009;324:1019–20. future. Journal of Power Sources 2009;186:539–42. [72] Richardson D. A comparison of biomass energy sources and fuel pathways [60] Li X, Lopes L, Williamson S. On the suitability of plug-in hybrid electric for vehicles: an Ontario, Canada case study. Energy Sources Part A. p. 16, vehicles (PHEV) charging infrastructure based on wind and solar energy. In: http://dx.doi.org/10.1080/15567036.2012.721861, in press. Proceedings of the IEEE power & energy society general meeting; 2009. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 1 of 53

The Market for Electric Vehicles: Indirect Network Eﬀects and Policy Design

Shanjun Li Lang Tong Jianwei Xing Yiyi Zhou∗

May 2016

Abstract

The market for plug-in electric vehicles (EVs) exhibits indirect network effects in that there is interdependence between consumer decision of EV purchase and investor decision of charging station deployment. Through a stylized model, we demonstrate that indirect network effects on both sides of the market lead to feedback loops which could amplify shocks and alter the diffusion process of the new technology. The relative strength of indirect network effects on the two sides of the market has important implications on policy design. Based on quarterly EV sales and charging station deployment in 353 Metro areas from 2011 to 2013, our empirical analysis finds indirect network effects on both sides of the market with those on the EV demand side being stronger. The federal income tax credit of up to $7,500 for EV buyers contributed to about 40% of the EV sales during 2011-2013 with feedback loops explaining 40% of that increase. A policy of equal-sized spending but subsidizing charging station deployment could have been more than twice as effective in promoting EV adoption.

Keywords: electric vehicles, indirect network eﬀects, policy design JEL classiﬁcation: Q48, Q58, L62

∗Shanjun Li is an Associate Professor, School of Applied Economics and Management, Cornell Univer- sity, [email protected]; Lang Tong is Irwin and Joan Jacobs Professor, School of Electric and Computer Engineering, Cornell University, [email protected]; Jianwei Xing is a PhD student, Dyson School of Ap- plied Economics and Management, Cornell University, [email protected]; Yiyi Zhou is an Assistant Professor, Department of Economics and College of Busines, Stony Brook University, [email protected]. We thank seminar participants at the University of Michigan, 2015 International Industrial Organization Con- ference, 3rd Northeast Workshop on Energy and Environmental Policy, and 2015 National Tax Association conference for helpful comments. We are grateful to the Editor Dan Phaneuf and two anonymous reviewers for their detailed and constructive comments and suggestions. Shanjun Li and Lang Tong acknowledge the support by the National Science Foundation under Grant CNS-1248079. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 2 of 53

1 Introduction

The electrification of the transportation sector through the diffusion of plug-in electric vehicles (EVs), coupled with cleaner electricity generation, is considered a promising pathway to reduce air pollution from on-road vehicles and to strengthen energy security. The U.S. transportation sector contributes to nearly 30% of U.S. total greenhouse gas emissions, over half of carbon monoxide and nitrogen oxides emissions, and about a quarter of hydrocarbons emissions in recent years. It also accounts for about three-quarters of U.S. petroleum consumption. Different from conventional gasoline vehicles with internal combustion engines, plug-in electric vehicles (EVs) use electricity stored in rechargeable batteries to power the motor and the electricity comes from external power sources. When operated in all-electric mode, EVs consume no gasoline and produce zero tailpipe emissions. But emissions shift from on-road vehicles to electricity generation, which uses domestic fuel source. The environmental benefit critically depends on the fuel source of electricity generation.1 Since the introduction of the mass-market models into the U.S. in late 2010, monthly sales of EVs have increased from 345 in December 2010 to 13,388 in December 2015.2 Despite the rapid growth, the market share of electric cars is still small: the total EV sales only made up 0.82% of the new vehicle market in 2015. In the 2011 State of the Union address, President Obama set up a goal of having one million EVs on the road by 2015. Based on the actual market penetration, the goal was met less than halfway.3 As a new technology, EVs face several significant barriers to wider adoption including the high purchase cost, limited driving range, the lack of charging infrastructure and long charging time. Although EV owners can charge their vehicles overnight at home, given the limited driving range, consumers may still worry about running out of electricity before reaching their destination. This issue of range anxiety could lead to reluctance to adopt EVs especially when public charging stations are scarce. At the same time, private investors have less incentive to build charging stations if the size of the EV fleet and the market potential are small. The interdependence between the two sides of the market (EVs and charging stations) can be characterized as indirect network effects (or the chicken-and-egg problem):

1Holland et al. (2015) find considerable heterogeneity in environmental benefits of EV adoption depending on the location and argue for regionally differentiated EV policy. 2From 1996 to 1998, GM introduced over 1000 first-generation EVs (EV1) in California, mostly made available through leases. In 2003, GM crushed their EVs upon the expiration of the leases. 3Similar national goals exist in many other countries: Chinese government set up a goal of half a million EVs on the road by 2015 and five million by 2020. German government developed an initiative to reach one million EVs by 2020.

1 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 3 of 53 the benefit of adoption/investment on one side of the market increases with the network size of the other side of the market. The objective of this study is to empirically quantify the importance of indirect network effects on both sides of the EV market and examine their policy implications. This is important for at least two reasons. First, while industry practitioners and policy makers often use the chicken-and-egg metaphor to characterize the challenge faced by this technology, we are not aware of any empirical analysis on this issue. Examining the presence and the magnitude of indirect network effects is important in understanding the development of the EV market. If indirect network effects exist on both sides of the market, feedback loops arise. The feedback loops could exacerbate shocks, whether positive or negative, on either side of the market (e.g., gasoline price changes or government interventions) and alter the diffusion path. Ignoring feedback loops could lead to under-estimation of the impacts of policy and non-policy shocks in this market. Second, indirect network effects could have important policy implications. As we describe below, policy makers in the U.S. and other countries are employing a variety of policies to support the EV market. When promoting consumer adoption of this technology, they can subsidize EV buyers or charging station investors or a combination of the two. Both our theoretical and empirical analysis show that the nature of indirect network effects largely determines the effectiveness of different policies. Therefore, understanding indirect network effects could help develop more effective policies to promote EV adoption. Taking advantage of a rich data set of quarterly new EV sales by model and detailed information on public charging stations in 353 Metropolitan Statistical Areas (MSAs) from 2011 to 2013, we quantify indirect network effects on both sides of the market by estimating two equations: a demand equation for EVs that quantifies the effect of the availability of public charging stations on EV sales; and a charging station equation that quantifies the effect of the EV stock on the deployment of charging stations. Recognizing the endogeneity issue due to simultaneity in both equations, we employ an instrumental variable strategy to identify indirect network effects. To estimate the network effects of charging stations on EV adoption, we use Bartik (1991)-style instrument for the endogenous number of electric charging stations, which interacts national charging station deployment shock with local market conditions: number of grocery stores and supermarkets. To estimate the network effects of EV stock on charging station deployment, we use current and historic gasoline prices to instrument for the endogenous cumulative EV sales. Across various specifications, our analysis finds statistically and economically significant indirect network effects on both

2 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 4 of 53 sides of the market. The estimates from our preferred specifications show that a 10% increase in the number of public charging stations would increase EV sales by about 8% while a 10% growth in EV stock would lead to a 6% increase in charging station deployment. With the parameter estimates, we examine the effectiveness of the federal income tax credit program which provides new EV buyers a federal income tax credit of up to $7,500.4 Our simulations show that the $924.2 million subsidy program contributed to 40.4% of the total EV sales during this period. Importantly, our analysis shows that feedback loops resulting from indirect network effects in the market accounted for 40% of that sales increase, a significant portion. Our simulations further show that if the $924.2 million tax incentives were used to build charging stations instead of subsidizing EV purchase, the increase in EV sales would have been twice as large. The better cost-effectiveness of the subsidy on charging stations relative to the income tax credit for EV buyers is due to (1) strong indirect network effects on EV demand; and (2) low price sensitivity of early adopters. This study directly contributes to the following three strands of literature. First, our study adds to the emerging literature on consumer demand for electric vehicles. CBO (Congressional Budget Office) estimates the effect of income tax credits for EV buyers based on previous research on the effects of similar tax credits on traditional hybrid vehicles and finds that the tax credit could contribute to nearly 30% of future EV sales. DeShazo et al. (2014) use a state-wide survey of new car buyers in California to estimate price elasticities and willingness to pay for different vehicles and then simulate the effect of different rebate designs. They estimate that the current rebate policy in California that offers all income classes the same rebate of $2,500 for BEVs and $1,500 for PHEVs lead to a 7% increase in EV sales. Using market-level sales data, our study offers a first analysis to quantify the role of indirect network effects in the market and their implications on government subsidies. Second, our study fits into the rich literature on indirect network effects. Previous work on indirect network effects dates back to early theoretical studies such as Rohlfs (1974), Katz and Shapiro (1985) and Farrell and Saloner (1985). Our paper is also related to the emerging

4Throughout our analysis, we treat the tax credit as a full-amount rebate due to the lack of household- level data in our analysis. EV buyers are more aﬄuent than average vehicle buyers and their tax liability is likely to be over $7,500. According to California Plug-in Electric Vehicle Owner Survey (2014), among buyers of conventional new vehicles, 15% of households have annual household income over $150,000 while among EV buyers, that share is 54%. https://energycenter.org/clean-vehicle-rebate-project/ vehicle-owner-survey/feb-2014-survey.

3 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 5 of 53 literature on two-sided markets that exhibit indirect network effects5. Theoretical work includes Rochet and Tirole (2006), Caillaud and Jullien (2003), Armstrong (2006), Hagiu (2006), and Weyl (2010), and empirical work includes the PDA and compatible software market by Nair et al. (2004), the market of CD titles and CD players by Gandal et al. (2000), the yellow page industry by Rysman (2004), and the video game industry by Clements and Ohashi (2005), Corts and Lederman (2009), Lee (2013), and Zhou (2014). In this strand of literature, our study is closest to Corts (2010) in topic which extends the literature to the automobile market and studies the effect of the installed base of flexible-fuel vehicles (FFV) on the deployment of E85 fueling stations. Corts (2010) only focuses on indirect network effects on one side of the market and does not look at that the effect of E85 fueling stations on FFV adoption. Third, our analysis contributes to the rich literature on the diffusion of vehicles with advance fuel technologies (e.g., hybrid vehicles) and alternative fuels (e.g., FFVs). Kahn (2007), Kahn and Vaughn (2009), and Sexton and Sexton (2014) examine the role of consumer environmental awareness and signaling in the market for traditional hybrid vehicles. Heutel and Muehlegger (2012) study the effect of consumer learning in hybrid vehicle adoption focusing on different diffusion paths of Honda Insight and Toyota Prius. Several recent studies have examined the impacts of government programs both at the federal and state levels in promoting the adoption of hybrid vehicles including Beresteanu and Li (2011), Gal- lagher and Muehlegger (2011) and Sallee (2011). Both hybrid vehicles and EVs represent important steps in fuel economy technology. Environmental preference, consumer learning and government policies are likely to be all relevant in the EV market. Our paper focuses on the key difference between these two technologies: indirect network effects in the EV market. Huse (2014) examines the impact of government subsidy in Sweden on consumer adoption of FFVs and the environmental impacts when consumers subsequently choose to use gasoline instead of ethanol due to low gasoline prices. Based on naturalistic driving data, Langer and McRae (2014) show that a larger network of E85 fueling stations would reduce the time cost of fueling and hence increase the adoption of FFVs. Section 2 briefly describes the industry and policy background of the study and the data. Section 3 presents a simple model of indirect network effects and uses simulations to show how feedback loops amplify shocks. Section 4 lays out the empirical model. Section 5

5Although exhibiting indirect networks, the EV market diﬀers from the canonical two-sided markets in that there is no well-deﬁned platform for buyers and sellers to interact. The automakers sell EVs to consumers directly. Public charging stations serve as a backup to home charging (e.g., a complementary good). The automakers do not charge charging stations loyalty fees or membership fees as is often the case in a two-sided market.

4 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 6 of 53 presents the estimation results. In section 6, we present the policy simulations and compare the existing income tax credit policy with an alternative policy. Section 7 concludes.

2 Industry and Policy Background and Data

In this section, we ﬁrst present industry background focusing on important barriers to EV adoption and then discuss current government policies. Next we present the data used in the empirical analysis.

2.1 Industry Background

Tesla Motors played a significant role in the come-back of electric vehicles by introducing Tesla Roadster, an all-electric sport car in 2006 and beginning general production in March 2008. However, the model had a price tag of over $120,000, out of the price range for average buyers. Nissan Leaf ($33,000) and Chevrolet Volt ($41,000) were introduced into the U.S. market in December 2010, marking the beginning of the mass market for EVs. There are currently two types of EVs: battery electric vehicles (BEVs) which run exclusively on high-capacity batteries (e.g., Nissan LEAF), and plug-in hybrid vehicles (PHEVs) which use batteries to power an electric motor and use another fuel (gasoline) to power a combustion engine (e.g., Chevrolet Volt). As depicted in Figure 1, quarterly EV sales increased from less than 2,000 in the first quarter of 2011 to nearly 30,000 in the last quarter of 2013 while the number of public charging stations has increased from about 800 to over 6,000. Nevertheless, the EV market is still very small: EV sales only made up 0.82% (or 113,889) of the total new vehicle sales in the U.S. in 2015 and there are only about 12,500 public charging stations as of March 2016, compared to over 120,000 gasoline stations. There are several commonly-cited barriers to EV adoption. First, EVs are more expensive than their conventional gasoline vehicle counterparts. The manufacturer’s suggested retail prices (MSRP) for the 2015 model of Nissan Leaf and Chevrolet Volt are $29,010 and $34,345, respectively, while the average price for a comparable conventional vehicle (e.g., Nis- san Sentra, Chevrolet Cruze, Ford Focus and Honda Civic) is between $16,000 and $18,000. A major reason behind the cost differential is the cost of battery. As the battery technology improves, the cost should come down. In addition, lower operating costs of EVs can significantly offset the high initial purchase costs.6 A recent study by EPRI (2013) compares the

6For a regular EV such as Nissan Leaf, the fuel cost of traveling 100 miles is about $3.6 assuming that it takes 30 kWh to drive the distance and the electricity price is 12 cents per kWh. For a conventional gasoline vehicle, the fuel cost is about $14 assuming the fuel economy of 25 MPG and gasoline price at $3.5 per gallon.

5 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 7 of 53 lifetime costs (including purchase cost less incentives, maintenance, and operation) of vehicles of different fuel types and finds that under reasonable assumptions, higher capital costs are well balanced by savings in operation costs: EVs are typically within 10% of comparable hybrid and conventional gasoline vehicles. The second notable barrier to EV adoption is the limited driving range. BEVs have a shorter range per charge than conventional vehicles have per tank of gas, contributing to consumer anxiety of running out of electricity before reaching a charging station. Nissan LEAF, the most popular BEV in the U.S. has an EPA-rated range of 84 miles on a fully charged battery in 2015. Chevrolet Volt has an all-electric range of 38 miles, beyond which it will operate under gasoline mode. This range is sufficient for daily household vehicle trips but may not be enough for longer distance travels. The third barrier, closely related to the second, is the lack of charging infrastructure. A large network of charging stations can reduce range anxiety and allow PHEVs to operate more under the all-electric mode to save gasoline.7 There are two types of public charging stations: 240 volt AC charging (Level 2 charging) and 500 volt DC high-current charging (DC fast charging), with the former being the dominant type. The installation of charging stations involves a variety of costs including charging station hardware, other materials, labor and permit. A typical Level 2 charging station for public use has 3-4 charging units and costs about $ 27,000 while a DC fast charging station costs over $50,000.8 Charging stations can be found at workplace parking lots, shopping centers, grocery stores, restaurants, dealers and existing gasoline stations, a point that we will come back to when constructing the instrument for the number of charging stations in the EV demand estimation. Owners of charging stations are often motivated by a variety of considerations such as boosting their sustainability credentials, attracting customers for their main business, and providing a service for employees. Charging stations are often managed by one of the major national operators such as Blink, ChargePoint, and eVgo. The fourth barrier is the long charging time. It takes much longer to charge EVs than

7According to California Plug-in Electric Vehicle Owner Survey (2014), 71% of EV owners expressed dissatisfaction with public charging infrastructure, coming down from 83% in 2012.https://energycenter. org/clean-vehicle-rebate-project/vehicle-owner-survey/feb-2014-survey. 8According to the charging station cost report by U.S. Department of Energy Vehicle Technologies Oﬃce (2015), the cost of a level-2 EV charging unit for public use is between $3,000 and $6,000, and the installation fee is from $600 to $12,700 per unit. Use the average equipment cost ($ 4,500) and installation fee ($3,000) per unit, the total cost of installing a charging station of an average size (3.6 charging units) comes at $27,000. This estimate does not include future maintenance and operating cost and is therefore a lower bound estimate. Footnote 28 provides a upper bound estimate which includes those costs. http://www. afdc.energy.gov/uploads/publication/evse_cost_report_2015.pdf.

6 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 8 of 53 to ﬁll up gasoline vehicles. A BEV may not be able to get fully charged overnight if just using a regular 120 Volt electric plug (e.g., it takes 21 hours for Nissan LEAF to get fully charged). To get faster charging, BEV drivers either need to install a charging station at home or go to public charging stations. It takes 6-8 hours to fully charge a Nissan Leaf at a Level 2 charging station and only 10-30 minutes at a DC fast charging station.9 Unlike BEVs, PHEV batteries can be charged not only by an outside electric power source, but by the internal combustion engine as well. Having the second source of power may alleviate range anxiety but the shorter electric range limits the fuel cost savings from EVs.

2.2 Government Policy

The diffusion of electric vehicles together with a clean electricity grid can be an effective combination in reducing local air pollution, greenhouse gas emissions and oil dependency. The EV technology is widely considered as representing the future of passenger vehicles. The International Energy Agency projects that by 2050, EVs have the potential to account for 50% of the light duty vehicle sales.10 Many countries around the world have developed goals to develop the EV market and provide support to promote the diffusion of this technology (Mock and Zhang 2014).11 To reduce the price gap between EVs and their gasoline counterparts, the Energy Im- provement and Extension Act of 2008, and later the American Clean Energy and Security Act of 2009 grant federal income tax credit for new qualified EVs. The minimum credit is $2,500 and the credit may be up to $7,500, based on each vehicle’s battery capacity and the gross vehicle weight rating. Moreover, several states have established additional state-level incentives to further promote EV adoption such as tax exemptions and rebates for EVs and non-monetary incentives such as HOV lane access, toll reduction and free parking. California through the Clean Vehicle Rebate Project offers $2,500 rebate to BEV buyers and $1,500 rebate to PHEV buyers. In addition, federal, state and local governments provide funding to support charging station deployment. For example, the Department of Energy provided

9Consumers do not need to wait for the battery to get fully charged before operating their vehicles again. They can recharge batteries by a certain amount depending on the duration of their stay at the charging locations while working, shopping or running errands. Public charging stations mainly serve as a backup or complementary charging option to alleviate EV drivers’ range anxiety. A concern towards DC fast charging is that it can reduce battery life due to the nature of charging. In addition, DC fast charging on a large scale can create demand spikes on the local electricity grid and exacerbate peak demand. 10Hydrogen vehicles (not yet mass-produced) will account for the majority of the remainder. https://www.iea.org/publications/freepublications/publication/EV PHEV Roadmap.pdf. 11The Chinese government provides rebate of over $9,000 to BEV buyers and nearly $8,000 for PHEV buyers. The UK government oﬀers a grant of up to $7,800 to EV buyers. In Japan, EV buyers were eligible for a subsidy of up to $10,000 in 2013 and $8,500 in 2014.

7 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 9 of 53

ECOtality Inc. $115 million grant to build residential and public charging stations in 22 U.S. cities in collaboration with local project partners. Government intervention in this market could be justified from the following perspectives. First, indirect network effects in the EV market represent a source of market failure since the marginal consumer/investor only consider the private benefit in their decision and the network size on both sides is less than optimal (Liebowitz and Margolis 1995; Church et al. 2002). In addition, given the nature of the market, each side of the market is unlikely to internalize the external effect on the other side through market transactions. If EVs are produced by one automaker, the automaker would have incentive to offer a charging station network to increase EV adoption. Nissan and GM are the two early producers of EVs but more and more auto makers are entering the competition. Nissan is a large owner of charging stations but GM is not.12 Second, the external costs from gasoline consumption in the U.S. and many countries around the world are not properly reflected by the gasoline tax (Parry and Small 2005; Parry et al. 2014). Compared to conventional gasoline vehicles, EVs offer environmental benefits when the electricity comes from clean generation such as renewables. In regions that depend heavily on coal or oil for electricity generation, EVs may not demonstrate an environmental advantage over gasoline vehicles.13 The electricity generation continues to become cleaner around the world due to the adoption of abatement technologies (e.g., scrubbers), the deployment of renewable generation and the switch from coal to natural gas. In addition, technologies are being developed to integrate EVs and renewable electricity generation such as solar and wind. The integration of the intermittent energy source with EV charging not only can help EVs fully realize its environmental benefits but also can leverage EV batteries as a storage facility to address the issue of intermittency and serve as energy buffer (Lund and Kempton 2008). Third, technology spillovers among firms often exist especially in the early stage of new technology diffusion (Stoneman and Diederen 1994). The development of the EV technology requires significant costs but the technology knowhow once developed can spread through

12Tesla is building its own proprietary network for Telsa owners only. This suggests that they recognize the importance of charging stations in EV adoption but this would create duplicate systems. 13 Zivin et al. (2014) estimate marginal CO2 emissions of electricity production that vary by location and time of the day and they find that charging EVs in some regions (the upper Midwest) during the recommended off-peak hours of midnight to 4am even generates more carbon emissions than the average conventional gasoline vehicle on the road. The environmental benefit of EVs under different fuel mix of electricity generation is still an active research topic and a critical element that has not been well understood in the literature is what types of vehicles (hybrid or gasoline vehicles) EVs replace.

8 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 10 of 53 many channels including worker migration and the product market. Bloom et al. (2013) estimate that the social returns to R&D are larger than the private returns due to positive technology spillovers, implying under-investment in R&D. In addition, the social returns to R&D by larger ﬁrms are larger due to stronger spillovers.

2.3 Data

We construct a panel dataset consisting of quarterly EV sales by vehicle model and the number of charging stations available at 353 MSAs from 2011 to 2013. Table 1 presents summary statistics of the variables used in our regression analysis. Data on quarterly vehicle sales of each EV model in each MSA is purchased from IHS Automotive. The sales data include 17 EV models: 10 BEVs and 7 PHEVs. Due to diﬀerent introduction schedules, there were two vehicle models in our 2011 data: Nissan LEAF and Chevrolet Volt. The 2012 data include four more vehicle models: Ford Focus EV, Mitsubishi i-MiEV, Fisker Karma, and Toyota Prius Plug-in. The 2013 data include 11 additional models: Honda Accord Plug-in, Ford C-Max Energi, Cadillac ELR, Honda Fit EV, Fiat 500E, Smart ForTwo Electric Drive, Tesla Model S, Porsche Panamera, Toyota RAV4, Chevrolet Spark EV, and Ford Transit Connect EV. In 2013, the top four EV models are Nissan Leaf, Chevrolet Volt, Tesla Model S and Toyota Prius plug-in with market shares (sales) of 25.8% (22,610), 24.4% (23,094), 17.4% (18,650) and 9.4% (12,088), respectively. For our analysis, we focus on the 353 MSAs (out of 381 MSAs in total) for which observations are available in all three years, and their EV sales accounted for 83% of the national EV sales during our data period. Panel A of Figure 1 depicts the spatial patten of EV ownership (the number of EVs per million people) in the last quarter of 2013. It shows that large urban areas have a higher concentration of EVs. The MSA with the highest concentration is San Jose-Sunnyvale-Santa Clara, CA with 5,608 EVs per million people by the end of 2013. The next two MSAs are both nearby: San Francisco-Oakland-Fremont and Santa Cruz-Watsonville. The MSA with the lowest concentration is Laredo, TX with only 36 EVs per million people (9 EVs with a population of a quarter of a million). We obtain detailed information on locations and open dates of all charging stations from the Alternative Fuel Data Center (AFDC) of the Department of Energy. By matching the ZIP code of each charging station to an MSA and using the station open date, we construct the total number of public charging stations available in each quarter for each MSA. Panel B of Figure 2 shows the spatial distribution of charging stations (the number of charging stations per million people). The pattern is very similar to what we observe in Panel A for

9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 11 of 53

EV ownership. The correlation coeﬃcient between the two variables is 0.63, partly reﬂecting the interdependence of EVs and charging stations. The top three MSAs with the most charging stations per million people are Corvallis, OR, Olympia, VA and Napa, CA with 210, 170, 117 public charging stations per million people, respectively. These three MSAs are the number 11th, 5th, and 6th in terms of the EV concentration in Panel A. We collect data on state-level incentives such as tax credits and rebates for both electric vehicles and charging stations from AFDC. From the American Chamber of Commerce cost- of-living index database, we collect quarterly gasoline prices for each MSA from 2008 to 2013. Household demographics are collected from the American Community Survey.

3 A Model of Indirect Network Eﬀects

In this section, we use a stylized model to illustrate indirect network effects on both sides of the market (EV demand and charging station investment) and to show how indirect network effects give rise to feedback loops. We then conduct simulations to shed light on how the effectiveness of different types of policies (e.g., subsidizing EV purchases versus charging station investment) hinges on the relative magnitude of indirect network effects on the two sides as well as consumer price sensitivity. The results from the simulations provide theoretical basis for our empirical findings based on real-world data.

3.1 Model Setup and Properties

We assume that EV sales qt(Nt, pt, xt) depends on the number of public charging stations in

the market (Nt), the price of EV (pt) and other product characteristics combined (xt) that aﬀect consumers’ choice such as the fuel cost.14 The installed base of EVs is the cumulative Pt sum of EV sales minus scrappage by the time t, denoted by Qt = h=1 qh ∗ st,h, where st,h is the survival rate at time t for EVs sold in time h. The number of charging stations that

have been built Nt(Qt, zt) depends on the EV market size Qt and other variables combined

zt that might aﬀect the ﬁxed cost of investment. To facilitate the illustration, we specify the following functions for EV demand and charging station deployment:

ln(qt) = β1ln(Nt) + β2ln(pt) + β3xt, (1)

ln(Nt) = γ1ln(Qt) + γ2zt. (2)

14We assume there is only one EV model to ease exposition. Our empirical analysis is at the vehicle model level and uses a richer speciﬁcation.

10 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 12 of 53

The EV demand equation arises from a discrete choice model of vehicle demand and follows closely the logit model using the market-level data as in Berry (1994). The charging station equation can be derived from an entry model as in Gandal et al. (2000) and we derive an empirical counterpart to this equation for our speciﬁc context in the appendix.

β1 and γ1 capture the magnitude of the indirect network eﬀects on the two sides. Feedback

loops (or two-way feedback) arise if both β1 and γ1 are non-zero. Intuitively, a shock to the system, for example an increase in xt would change EV sales qt, which would in turn aﬀect the installed base Qt+1. This would then lead to changes in the number of charging stations Nt+1 and hence aﬀect qt+1. The impact would circle back and forth between these two equations.

If both β1 and γ1 are positive, positive feedback loops would arise and they can amplify the shocks (either positive or negative) in either side of the market such as tax credit for EV purchases or subsidy on charging station investment. β2 (negative) is the price elasticity of demand and captures consumer price sensitivity. To understand the property of the system such as the existence of the steady state and t−h its property, we assume that the survival rate st,h is δ , where δ < 1. Further assume

pt = p, xt = x and zt = z. Substituting equation (2) into equation (1), we have:

ln(qt) − β1γ1ln(qt + δQt−1) = β1γ2z + β2ln(p) + β3x. (3)

The right-hand side β1γ2z + β2ln(p) + β3x is constant with respect to qt and we denote c it by c. During period 1, Qt−1 = 0. Solving the equation, we obtain q1 = exp( ). To 1−β1γ1 ∗ ∗ ∗ solve for the steady state solution (q , N ), we use the steady state condition qt = qt+1 = q and ﬁnd:

∗ c − β1γ1ln(1 − δ) −β1γ1ln(1 − δ) q = exp( ) = q1 ∗ exp( ). 1 − β1γ1 1 − β1γ1

∗ c − β1γ1ln(1 − δ) N = exp[γ1 − γ1ln(1 − δ) + γ2z]. 1 − β1γ1 The stock of EVs in the steady state Q∗ = q∗/(1 − δ) where the outﬂow of EVs due to scrappage is equal to the inﬂow from new EV sales.15 To examine the stability of the steady

state, we write Qt−1 = qt−1 + δQt−2 and substitute it into equation (3): ln(qt) − β1γ1ln(qt + 2 δqt−1 + δ Qt−2) = c. This deﬁnes an implicit function of qt = G(qt−1). When β1γ1 < 1, it

15Alternatively, the steady state can be equivalently expressed in terms of (Q∗, N ∗). Our specification rules out (0,0) as another steady state solution. Having multiple equilibria is often a signature property of two-sided markets with indirect network effects due to self-confirming expectations (e.g., Gandal et al. (2000)). From an empirical perspective, our specification is without loss of generality since the empirical studies in this literature often assume that the non-zero stable solution plays out in the data.

11 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 13 of 53

can be shown that G(0) > 0,G0() > 0, and G00() < 0. Therefore, the steady state solution is stable as shown in Figure 3. In our following policy analysis, we take β1γ1 < 1, which is also conﬁrmed in our empirical analysis. The partial eﬀect of vehicle price p on EV sales in the steady state is: ∂q∗ c − β γ ln(1 − δ) β = exp( 1 1 ) 2 , ∂p 1 − β1γ1 1 − β1γ1

where β2 < 0. When β1γ1 < 1, this partial eﬀect is negative as economic theory would suggest. Similarly, the changes in other demand side factors captured by x and the changes

in the factors in the charging station equation z will both shift G(qt−1) in Figure 3 up or down and hence aﬀect the steady state solution of EV sales.

3.2 Implications on Policy Choices

Now we conduct simulations to understand how feedback loops magnify policy shocks and

their implications on policy choices. We ﬁx pt, xt, and zt in equations (1) and (2) and assume certain values for model parameters as reported in Table 2. We then solve for qt, Qt, and

Nt sequentially for each period. Because of the positive feedback loops (by assuming both

β1 and γ1 being positive), EV sales and the number of charging stations will keep growing naturally until they reach the steady state where the inflow of new vehicles equals the outflow of vehicles due to scrappage. To examine how positive feedback loops could amplify a policy shock, we simulate a scenario where all EV buyers are provided with a $7,500 subsidy for the first five periods and no more subsidy is offered afterwards.

As shown in Panel (a) in Figure 4, due to both the price effect (captured by β2) and the indirect network effects (captured by β1 and γ1), the subsidy increases EV sales substantially compared with the no-policy case during the first five periods. When the subsidy terminates, EV sales continue to increase through feedback loops but with a smaller magnitude. The sales increase due to subsidy gets smaller as feedback loops diminish and the two growth paths eventually overlap. In both cases, the path of EV sales converges to the same steady state but the policy shock makes the system converge to the steady state more quickly: indirect network effects expedite this process through positive feedback loops. Figure 4 Panel (b) depicts a similar pattern in the dynamic path of charging station deployment. With the positive policy shock on the EV purchase side, the stock charging stations increases quickly for the first five periods and continues to grow at a decreasing rate after the policy. It eventually converges to the same steady state as in the no-policy scenario. This two graphs demonstrate that feedback loops from indirect network effects magnify a

12 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 14 of 53

shock to any side of the system and alter the convergence process on both sides. The existence of indirect network effects on both sides of the market could have important policy implications. To foster the development of the EV market, policy makers can choose to subsidize consumers for EV purchase directly (policy 1) or to subsidize charging station investment (policy 2). We conduct simulations to examine the relative cost-effectiveness of these two policy options. Policy 1 provides EV buyers with a subsidy of $7,500 per EV in the first five periods. Policy 2 uses the same account of total funding as in policy 1 to build charging stations. We compare the cumulative sales increase over time due to these two policies (with a 5% annual discount rate).16 To examine the implication of relative strength of indirect network effects on policy choices, we vary the ratio of β1/γ1 by holding β1 constant while changing γ1. Figure 5 depicts, for any given price sensitivity β2 (say -1.5), as β1/γ1 increases (i.e., indirect network effects in EV demand become relatively stronger), the second policy (subsidy on charging stations) becomes more and more effective measured by the increase in cumulative sales over time. The two policies are equivalent when β1/γ1 is 1 given the price elasticity of -1.5. In addition to relative strength of indirect network effects on both sides, the policy comparison also depends on the price elasticity of EV demand. When consumers are more sensitive to prices (e.g., going from -1.5 to -1.6), the policy of subsidizing charging stations becomes relatively less effective for a given β1/γ1. This finding is illustrated by the outward shift of the curve when the price elasticity changes to -1.4 and -1.6. The result is intuitive: if consumers are less price-sensitive, it would take a larger subsidy on EV purchases in order to push consumers to buy EVs, hindering the effectiveness of the policy. To summarize, the policy of subsidizing charging stations becomes more effective relative to the policy of subsiding EV purchases when indirect network effects on the EV demand become stronger (holding network effects on the charging station side constant) or when consumers are less sensitive to price. These findings offer a theoretical foundation for the policy comparison after our empirical analysis.

16For policy 1, we subtract the EV price by $7,500 to simulate the counterfactual sales in the first five periods. The total expenditure of the subsidy policy is then calculated by multiplying $7,500 with the total EV sales for the five periods. By assuming the cost of building one charging station to be $27,000, we obtain the total number of charging stations that could be built with the same amount of funding. We assume that the investment occurs evenly each year over the first five periods and we add the number of charging stations that could be funded to Nt to simulate the counterfactual outcomes under policy 2.

13 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 15 of 53

4 Empirical Framework

To investigate indirect network effects on both sides of the market, we estimate: (1) a EV demand equation that examines the effect of charging stations on EV sales; and (2) a charging station equation that estimates the effect of EV fleet on charging station deployment. These equations build upon equations (1) and (2) in the theoretical model above.

4.1 EV Demand

To describe the empirical demand model of EVs, let k index an EV model such as Nissan Leaf and Chevrolet Volt, m index a market (MSA), and t index a year-quarter. We estimate the following equation:

0 ln(qkmt) = β0 + β1ln(Nmt) + β2Xkmt + Tt + δkm + εkmt, (4)

17 where qkmt is the sales of EV model k in market m and year-quarter t. Nmt denotes the total number of public charging stations that have been built in the MSA by the end of a given quarter.18 We use the number of charging stations instead of the total number of charging outlets to represent the availability of charging infrastructure but the qualitative

findings remain if we use the number of charging units. ln(Nmt) captures the effect of charging stations on electric vehicle purchases and the log form allows the effect to be

diminishing. Xkmt is a vector of related covariates including the effective purchase price, personal income and other control variables. The effective purchase price of a model is defined as the manufacturer’s suggested retail price (MSRP) less the related subsidies (tax credits and tax rebates at both federal and state levels). We also include a full set of year-quarter (e.g., the first quarter of 2011) fixed effects and MSA-model (e.g., Nissan Leaf in San Francisco) fixed effects in equation (4). Year-quarter

ﬁxed eﬀects Tt control for national demand shock for EVs common across MSAs such as

consumer awareness. MSA-model fixed effects δkm not only control for time-invariant product attributes such as quality and brand loyalty that could affect vehicle demand but also control for time-invariant local preference for green products (Kahn 2007; Kahn and Vaughn 2009) and demand shocks for each model (e.g., a stronger preference or dealer presence for Nissan

17This empirical speciﬁcation is taken to be consistent with our theoretical model and to ease results interpretation. The logit model from Berry (1994) implies that the dependent variable would be ln(skit) − ln(s0mt) where skmt is the market share of model k in market m and time t and s0mt is the share of consumers who are not purchasing an EV. These two speciﬁcations provide almost identical parameter and elasticity estimates (see Table 5.B). 18 In the estimation, we add one to qkmt, Nmt to deal with zero values for some of the observations. Our results are robust to excluding observations with zero values on qkmt or Nmt and using ln(qkmt) and ln(Nmt).

14 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 16 of 53

Leaf in San Francisco). εkmt is the unobserved demand shocks that are time-varying and market-specific (for example, unobserved local government subsidy for purchasing EVs or market-specific promotions for a vehicle model that vary over time). It is well documented in the vehicle demand literature that failing to control for unobserved product attributes could lead to downward bias in the price coefficient estimates (for example Berry et al. (1995); Beresteanu and Li (2011)). MSA-model fixed effects absorbs both observed and unobserved vehicle attributes variations which are time-invariant and what is left is the variation of vehicle attributes over time. Since most of the EV models in our sample appear for only one year and there is little variation of the observed attributes for the models that appear for more than one year, we believe that using MSA-model fixed effects could control for unobserved product attributes and alleviate the need to use the methodology developed in Berry et al. (1995) to deal with price endogeneity where they only have national-level sales data (i.e., one market).19 The price coefficient is identified from the fact that effective EV prices vary across markets and over time due to state-level subsidies and temporal price variations. Although we include a rich set of control variables, the charging station variable is still endogenous due to simultaneity: the unobserved time-varying and market-specific demand shocks could affect charging station investment decisions and hence the stock of charging stations. To deal with the endogeneity, we use the IV strategy and a valid IV needs to be correlated with the number of charging stations in an MSA (the endogenous variable) but not correlated with the unobserved shocks to EV demand. The IV we employ is the interaction term between the number of grocery stores and supermarkets in an MSA in 2012 with the number of charging stations in all MSAs other than the MSA corresponding to a given observation (lagged for one quarter). Grocery stores and supermarkets are a major owner of charging stations and they build charging stations to attract customers and boost green credentials among other reasons. These places could be good sites for public charging stations because EV drivers can charge their vehicles while shopping. Nissan has been actively partnering with grocery store owners to build charging stations. Kroger, the country’s largest grocery store owner has installed about 300 charging stations in their stores across the country. Our data show that the number of grocery stores in an MSA is positively

19The methodology in Berry et al. (1995) uses a contracting mapping technique to first back out product- level fixed effects (mean utility) in the first stage and then uses IV strategy to estimate the remaining preference parameters based on the assumption that observed product attributes are not correlated with observed product attributes, which could be a strong assumption (Klier and Linn 2012). As a robustness check, we also include electric range and electric mpg in one of the alternative specifications and the results are qualitatively the same.

15 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 17 of 53 correlated with the number of charging stations. However, the number of grocery stores does not vary with time in our sample period and it is therefore absorbed by the MSA fixed effects. To introduce temporal variation, we multiply it with the lagged number of existing charging stations in all MSAs other than the MSA corresponding to a given observation, which captures the national-level trend in charging station investment due to aggregate shocks such as temporal variations in costs, investor confidence and federal incentive programs. The construction of this IV is similar in spirit to the Bartik instrument used in the labor literature to isolate local labor demand changes (Bartik 1991). The intuition for the IV is that national shocks to charging station investment (captured by the lagged number of charging station in all MSAs other than own) have disproportional effects on charging station investment across MSAs: MSAs with a larger number of grocery stores and supermarkets (hence better endowment of good sites for charging stations) will be affected by these national shocks more than others, leading to variations in charging stations across MSAs. Our first-stage results in Table 4 to be discussed below show that the interaction term has a positive and highly statistically significant impact on charging station investments. We argue that this instrument should satisfy the exogeneity assumption. The number of grocery stores and supermarkets is unlikely to affect EV sales directly. There might be common unobservables that influence both the EV sales and the number of grocery stores, especially at the cross-sectional level. However, our model controls for MSA fixed effects and should capture these time-invariant unobservables. At the temporal dimension, EV sales vary from year to year but the number of grocery stores is very stable given the maturity of the industry. In fact, the number of grocery stores is measured at the end of 2012. The temporal variation in the IV comes from the total (lagged) number of charging stations in all MSAs other than the own city. Time fixed effect would control for time-varying common shocks across MSAs. Excluding home city’s charging stations also removes the concern that one MSA’s installation of large amount of charging stations could overly influence the estimation results. Our IV strategy leverages the interaction term between a national-level variable with only temporal variation and a MSA-level variable with only spatial variation. The rationale behind the IV is that different MSAs have different pre-existing conditions/ability to absorb national shocks to charging station investment such as changes in macro-economic conditions and costs. One might be concerned that different MSAs may have different susceptibility to unobservable demand shocks at the national level and the number of grocery stores could

16 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 18 of 53 be correlated with this susceptibility for some reason. To address this concern, we include a variety of MSA-level controls interacting with the time trend. We use the sales of hybrid vehicles in 2007 (several years before EVs entered the market) to proxy for preference heterogeneity for greener vehicles or environmental friendliness. We also include personal income, the share of college graduates among residents, the share of commuters driving to work, the share of commuters using public transport to work, and the share of white residents. We use the interactions of these variables with the time trend to control for potential heterogeneity in diﬀusion path of EVs across MSAs. Our results are robust to the inclusion of these controls, providing further support that our IV is a valid exclusion restriction. In some of the robustness checks, we use local policy variables such as subsidies on charging stations as additional IVs and obtain similar results. We do not use them in our benchmark speciﬁcations due to the concern that local policies whether subsidizing charging stations or EV purchases could be a response to local unobserved demand shocks and hence be endogenous.

4.2 Charging Station Deployment

We derive the empirical model of charging stations investment from an entry model presented in the Appendix where the proﬁt depends on both the installed base of EVs and the total number of charging stations in a market. Under certain functional form assumptions, the total number of charging stations in a free-entry equilibrium is given by the following equation:

EV 0 ln(Nmt) = γ0 + γ1ln(Qmt ) + γ2Zmt + Tt + ϕm + ςmt, (5) where Nmt denotes the stock of public charging stations that have been built in market m EV by time t and Qmt denotes the installed base of EVs by time t. The vector of covariates

Zmt include the state-level tax credit given to charging station investors measured as the percentage of the building cost, a dummy variable indicating whether there exists public grants or funding to build charging infrastructure, the interaction term of number of grocery stores in a MSA in 2012 with the lagged number of charging stations in all MSAs other than own (the instrument in the EV demand equation), and other control variables.

We also include a full set of time and MSA fixed effects. Tt denotes year-quarter fixed effects to control for time-varying common shocks to charging station investment across MSAs such as macro-economic conditions. Market fixed effects ϕm control for time-invariant and MSA-specific preferences for charging stations. For example, some MSAs may be “greener” than others and invest more on alternative fuel infrastructure. Similarly, MSAs with a higher

17 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 19 of 53 population density and limited private installment of charging stations may have more public charging stations. ζmt is the unobserved shock to charging station investment, for instance, the unobserved local policies to support the charging station building. In the estimation, we EV add one to Nmt and Qmt to deal with zero values for some of the observations. We obtain EV similar results by dropping these observations and use ln(Nmt) and ln(Qmt ) instead.

The issue of endogeneity due to simultaneity also arises in this equation. Both Nmt and EV Qmt are stock variables but the inflows to each variable are determined at the same time. As a result, time-varying and MSA-specific shocks to investment decisions (the error term in the equation) could be correlated with current EV sales which is part of the installed base. The instrument variables emerge more naturally in this equation. In particular, we instrument for the installed base of EVs with a set of current and past gasoline price variables. The fuel cost savings from driving EVs depend on the price difference between gasoline and electricity, which varies across locations. In MSAs with higher gasoline prices, consumers may have a stronger incentive to purchase EVs 20. Because the installed base of EVs is the cumulative sales of EVs, we include gasoline prices not only in the current quarter but annual gasoline prices in the past three years as instruments. For example, for the installed base of EVs in the 2nd quarter in 2013, we use the gasoline price in the 2nd quarter in 2013, the average gasoline price in 2012, the average gasoline price in 2011, and the average gasoline price in 2010 as instrumental variables. These gasoline price variables (including current and past gasoline prices) should affect the installed base, which is confirmed in the first-stage regression in Table 7 to be discussed below. But they are unlikely to affect investment decisions directly (i.e., other than through the installed base). Since we include both time and MSA fixed effect, the remaining variation in gasoline prices is largely driven by how time-varying crude oil prices interact with market conditions that are likely time-invariant during our data period (e.g., market structure in wholesale and retail gasoline markets and distance to refineries). These interactions lead to time-varying and MSA-specific differences in gasoline prices, which are unlikely to be correlated with charging station investment directly. The decision of charging station investment hinges on, among other things, the EV market potential (proxied by the installed base of EVs) and the fixed costs of investment. Fixed costs of charging station investment include the cost of equipment (chargers) and labor cost, neither of which is likely to be correlated with gasoline price variations (after controlling for MSA and time fixed effects). The operating costs of the charger largely depend on electricity prices. There is no direct link between

20A report on the ownership cost of EVs by Electric Power Research Institute (2013) ﬁnds that the increases and decreases in gasoline prices will have a signiﬁcant impact on the relative costs of PEVs.

18 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 20 of 53 electricity and gasoline prices (after controlling for common shocks such as national economic conditions using time ﬁxed eﬀects).

5 Estimation Results

We ﬁrst present parameter estimates for equation (4) and (5). We then discuss the indirect network eﬀects implied by these parameter estimates.

5.1 Regression Results for EV Demand

Columns (a) to (e) in Table 3 report the ordinary least squares (OLS) estimation results for five different specifications where we add more control variables successively. Column (a) includes only six explanatory variables. Column (b) adds in year-quarter fixed effects to control for time-varying common unobservables across MSAs. Column (c) further adds vehicle model fixed effects to control for unobserved product attributes such as quality and brand loyalty that affect consumer demand. Column (d) includes rich MSA-model fixed effects to control for both unobserved product attributes and MSA-specific demand shocks for different EV models. Column (e) adds two additional variables to control for potential heterogeneity in the diffusion pattern across MSAs. The first variable is the interaction term between the sales of hybrid vehicles in 2007 (to proxy for preference for green vehicles) and the time trend and the second variable is the interaction between average personal income and the time trend. Column (f) implements a GMM estimation strategy and uses the interaction term of number of grocery stores and supermarkets in a MSA in 2012 with the lagged number of charging stations in all the other MSAs as the instrument for the number of charging stations. Given the log-log specification, the coefficient estimates can be interpreted as elasticities. All the specifications provide intuitive and statistically significant coefficient on the key variables of interests: EV demand increases with a larger network of charging stations, a lower vehicle price, and more home charging stations funded by the EV project supported by the DOE, and higher income. The coefficient on the charging station variable captures indirect network effects from charging station investment on EV demand. The GMM results show that a 10% increase in charging stations would result in a 8.4% increase in the EV sales, which is higher than all the OLS estimates (ranging from 1.8% to 5%). This suggests that the number of charging stations is negatively correlated with the unobserved shocks to EV demand, leading to downward bias in OLS. One example of unobserved shocks is local EV incentives that local governments provide to compensate for the lack of public charging

19 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 21 of 53

stations. Another example is the home charging incentives from local electric utilities. Many local utilities offer a rebate for installing a home charging station and a discounted rate for home EV charging as part of the demand-side management program. Local governments often partner with local utilities to provide more generous home charging incentives when there is a lack of private investment in public charging stations. The price coefficients changes from -0.470 to -0.817 from columns (b) to (c) after vehicle model fixed effects are included. This is consistent with our discussion above: vehicle model fixed effects control for unobserved product attributes which could be positively correlated with prices. Ignoring unobserved product attributes will bias the price coefficient toward zero. Going from Columns (c) to (d) where MSA-model fixed effects are included, the EV demand function changes from being inelastic with an price elasticity of -0.817 to being elastic with a price elasticity of -1.378. MSA-model fixed effects control for MSA-specific time- invariant demand shocks (such as environmental preference) and these demand shocks could affect state-level tax incentives. For example, higher incentives are used to counter negative demand shocks. Hence MSA fixed effects could control for the potential endogeneity in state- level tax incentives. The GMM results provide a price elasticity of -1.288. Although this is at the lower end of the price elasticity estimates in the literature on automobiles, we believe that the magnitude is reasonable compared with the literature for two reasons.21 First, EV buyers are more affluent and hence less price-sensitive compared with average vehicle buyers. Second and perhaps more importantly, EV buyers can be characterized as early adopters and one can argue that many of them choose EVs out of their strong environmental concerns and/or making a statement by driving an EV as has been documented in the case of hybrid vehicles (e.g., Kahn (2007), and Sexton and Sexton (2014)).22 Lower fuel cost is one of the major benefits of EVs and higher gasoline prices will have a positive impact on EV adoption by increasing future fuel cost savings from driving EVs in place of conventional vehicles. To capture the heterogeneous impact of gasoline price on the demand of the two types of EVs, two interaction terms of quarterly gasoline prices with BEV and PHEV dummy variables are included. The results from all the specifications find a positive and statistically significant effect of gasoline price on BEV purchase. While the

21Berry, Levinsohn and Pakes (1995) estimate the price elasticities ranging from -3 to -7 for vehicle models in 1990 with more expensive models having the smaller price elasticities (in magnitude). The lower end of price elasticities for vehicles models in 2006 in Beresteanu and Li (2011) is also around -3. 22California Plug-in Electric Vehicle Owner Survey (2014) shows that EV buyers have higher household income than buyers of gasoline vehicles and that the environmental concern is an important motivator for EV purchase. 38% of Nissan Leaf buyers and 18% Chevy Volt buyers consider the environmental concern to be the top motivator.

20 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 22 of 53 interaction term with PHEV is positive and significant in Columns (a) to (c), Columns (d) to (f) do not find a significant impact of gasoline price on PHEV demand when MSA-model fixed effects are included. Intuitively, BEV drivers could be more sensitive to gasoline prices than PHEV buyers given that BEV run exclusively on electricity while PHEVs run mostly on gasoline for long-distance driving given its short range of battery. That is, PHEV drivers do not make a long-term commitment to an alternative fuel to the extent that BEV drivers do. In addition to gasoline prices, we included electricity prices in previous analysis and the coefficient estimate was small in magnitude and statistically insignificant in all specifications. This could be due to: (1) the operating cost from using electricity is a small portion of vehicle lifetime cost for EVs; and (2) there is not much MSA-specific temporal variation in electricity prices. The coefficient on the interaction term between hybrid sales in 2007 and time trend is positive and statistically significant, implying that MSAs that had more sales of hybrid vehicles in 2007 (proxy for preference for greener products) have faster diffusion of EVs. The results from these regressions imply that the increased availability of public charging stations has a statistically and economically significant impact on EV adoption decisions. Our estimation results confirm that even if most EV drivers can charge vehicles at home, better access to public charging facilities elsewhere is still an important demand factor by, for example, alleviating range anxiety.23 Based on the parameter estimates on charging station and price variables, a back-of-the-envelope calculation shows that the demand effect from having one more charging station from the sample average of 22.6 is equivalent to that from a reduction of EV price by $961 (the average price is $33,127). When the number of charging stations increases to 27.3 (the sample average in 2013), the equivalent price reduction is $795. At the sample maximum of 320 charging stations, one more charging station is only equivalent to $68 price reduction, showing the diminishing effect implied by the log-log functional form.

5.2 Alternative Speciﬁcations for EV Demand

We take the estimates in Column (f) in Table 3 as our baseline specification. To check the robustness of the results, we estimate a variety of different specifications and the results are reported in Table 5. Column (a) includes the interaction term between the number of charging stations and average commute time to work in the MSA to capture the heteroge-

23According to the EV Project report (2013), the percentage of EV home charging for 22 program areas is about about 74% for Nissan Leaf and 80% for Chevrolet Volt.

21 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 23 of 53 neous impact of charging stations across cities with different commuting patterns.24. The positive coefficient estimate on the interaction term suggests that the availability of charging stations has a larger impact on EV demand in the MSAs with longer commute. This is intuitive since in MSAs where people have longer commute, range anxiety would be more of an issue. Across MSAs, the elasticity of charging stations with respect to the EV sales ranges from 0.27 to 1.05 depending on the average commute time. Column (b) adds the quadratic terms of the time trend variables to our baseline specification to allow more flexible time effect. The coefficient estimates on the key parameters are almost intact. Column (c) includes the interaction term of charging stations with a BEV dummy to capture the different impact of charging stations on BEVs and PHEVs. The coefficient estimate on the interaction term is positive while not statistically significant. Column (d) includes two more instruments: the tax credits and the availability of public funding for building charging stations, both of which appear in the charging station equation. We did not include them in our baseline specification because the exogeneity assumption may not hold for the subsidies as they could be a response to unobserved EV demand shocks. Column (e) removes the price variable in the regression to deal with the concern that the price variable especially state-level incentives could be endogenous. Column (f) adds the interaction terms between various demographic variables and the time trend to further control for MSA-level heterogeneity in the diffusion pattern. Across these specifications, the estimated effects of charging station availability on EV demand as well as other parameter estimates are similar to those from the baseline specification in Table 3. Column (g) reports the demand estimation using the logit model as in Berry (1994) and it produces almost identical elasticity estimates as our baseline specification. Some states such as California and Oregon have adopted Zero Emission Vehicle program which requires a certain part of automakers’ sales to be clean fuel vehicles and some automakers have introduced EV models in those regions only to comply with the regulations. To control for more intense competition in those markets due to more EV models introduced, Column (h) in Table 5.B includes ZEV specific time fixed effects and the results are similar to previous results with a modest increase in the coefficient estimate on charging stations. Column(i) uses only BEV sales and the estimates are not systematically different from the estimates using the full sample with BEVs and PHEVs. Columns (j) and (k) increases the lag of the total number of charging stations in all other MSAs to 2 and 3 quarters when

24The average commute time to work at the MSA level is calculated based on combined 2006-2011 samples of American Community Survey. The average commute time is 22.96 with a standard deviation of 3.37, a minimum of 14.59 and maximum of 35.01

22 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 24 of 53 constructing the instrumental variable and there is no substantial change of the estimates except that the coefficient of the charging stations decreased slightly, primarily due to loss of observations. However, increasing the lag to more than three quarters leads to weak IV. As shown in Column (g) in Table 5.B, our demand specifications would yield nearly identical results as the Berry-logit model. With only EV models in our data, our analysis treats all other non-EV models to be in one category (i.e., the outside good). Limiting the choice set and the substitution pattern across choices could potentially impact our estimate of the price elasticity and our policy simulations. EV models represent a different technology that is dramatically from conventional gasoline vehicles, therefore consumers are likely to consider them as a separate category in making purchase decisions especially given that the EV buyers in our data period are often motivated by strong environmental concern according to California Plug-in Electric Vehicle Owner Survey (2014). Nevertheless, some PHEVs do have conventional hybrid counterparts. For example, Toyota Prius-plug in has a hybrid version, Toyota Prius. Considering most of PHEVs have limited electric range (11-38 miles), some consumers may compare PHEVs with hybrid vehicles. Recognizing this, Column (i) in Table 5.B only include BEV models in the regression. The results are very similar to those obtained using the full sample, suggesting that the limitation in our choice set and modeling framework may not have a large impact on the key parameters of interest.25

5.3 Regression Results for Charging Station Deployment

Columns (a) to (d) in Table 6 report the OLS regression results for the charging station equation (5). In Column (a), only the four explanatory variables of interest are included. Column (b) includes year-quarter fixed effects to control for time trends that are common to all MSAs such as federal subsidies for building charging stations that occur during a specific period of time. Column (c) further includes MSA fixed effects to control for time- invariant MSA-level baseline differences in charging station investment. Column (d) adds in the interaction term between the hybrid vehicle sales in 2007 (proxy for environmental friendliness) and time trend to control for heterogeneity in the diffusion pattern of charging stations. All OLS regressions find a positive and statistically significant coefficient for the installed EV base. The estimate results in Column (d) suggest that a 10% increase in the EV fleet size

25EV models only represent less than 0.8% of new vehicle sales in the nation in 2013. Including models of other fuel types would not help us identify the indirect network eﬀects since their demand does not depend on EV charging stations. We believe that micro-level data with the second-choice information is much better suited to assess the substitution pattern between EVs and diﬀerent types of non-EVs than the aggregate data that we currently have. This is an ongoing work of the authors.

23 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 25 of 53 would lead to a 1.2% increase in the number of public charging stations. The GMM results in column (e) show that a 10% increase in EV fleet size would result in a 6.1% increase in charging stations. In Column (f), we add the EV incentives (tax credits and rebates at the federal and state levels) as an additional instrument and the coefficient for charging stations increases from 0.613 to 0.659. We take Column (e) as our baseline IV specification due to the concern that the EV incentives (especially those at the state level) could be endogenous as they could be a response to the unobserved shocks to the deployment of charging stations. The GMM coefficient estimates are higher than all the OLS estimates, suggesting that the installed base of EVs is negatively related to the unobserved shocks to charging station investment, leading to downward bias in OLS. An example of the unobserved shocks is the unobserved local policies: policy makers may design policies to support charging station investment to counteract negative EV demand shocks. The results in Column (e) show that tax credits and the availability of public funding for charging stations have positive but statistically insignificant coefficients. The tax credits and public funding are both at the state level. The dependent variable in the charging station equation, however, only includes publicly-accessible charging stations, which are mainly subsidized by the federal government directly through the federal projects such as the EV Project and ChargePoint Project. Although state-level tax credits and funding also apply to public charging stations, they mostly support the installation of charging stations at workplace and multi-family dwellings, which are usually privately-accessible and are excluded from our analysis. The interaction term of grocery stores with the lagged number of stations in all MSAs other than own has a positive and statistically significant coefficient, consistent with our argument for using it as a relevant instrument in the EV demand equation.

6 Policy Simulations

Our empirical analysis suggests that indirect network eﬀects exist on both sides of the market. In this section, we ﬁrst examine the policy impact of the current federal income tax credit policy for EV buyers and then compare this policy with an alternative policy that subsidizes charging station investment instead.

6.1 Impact of Income Tax Credits

The federal government has adopted several policies to support the EV industry including providing federal income tax credits for EV purchase, R&D support for battery development, and funding for expanding charging infrastructure. The CBO (Congressional Budget Oﬃce)

24 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 26 of 53 estimates that the total budgetary cost for those policies will be about $7.5 billion through 2017. The tax credits for EV buyers account for about one-fourth of the budgetary cost and are likely to have the greatest impact on vehicle sales. Under the tax credits policy, EVs purchased in or after 2010 are eligible for a federal income tax credit up to $7,500. Most popular EV models on the market are eligible for the full amount.26The credit will expire once 200,000 qualified EVs have been sold by each manufacturer. In order to examine the effectiveness of the income tax credit policy in terms of stimulating EV sales, we use our parameters estimates from the two baseline GMM regressions to stimulate the counterfactual sales of EVs that would arise in the absence of the $924.2 million worth of tax credits to EV buyers from 2011 to 2013. The impact of the policy depends not only on the price elasticity of EV demand in the EV demand equation, but also on the magnitude of indirect network effects captured in both equations. We assume in our simulations that the MSRPs will not be affected implying that the consumers previously captured all the subsidies. We believe that this is a reasonable assumption in the EV launch stage when automakers produce EVs likely at a loss since the production level is far below the efficient production scale.27 While more and more states are providing subsidy programs to encourage the adoption of EVs, the retail prices for electric vehicles have actually been decreasing (for the same model) during our sample period likely due to decreasing production cost and the increasing competition. Our simulation results in Table 8 show that EV sales would have been 56,690 less (or 40.44% of the total sales) from 2011 to 2013 without the $924.2 million worth of income tax credit to EV buyers. If we shut down feedback loops, the sales contribution from the tax credit policy would only have been 33,949 (24.2% of the total sales). This implies that feedback loops magnify the policy shock and explain 40% of the sales increase from the policy. The results suggest feedback loops have a multiplier effect of 1.67. CBO finds the policy impact to be 30% of the total EV sales while their study only considers the price effect of the tax credit but not the role of indirect network effects in amplifying the policy effect (CBO Congressional Budget Office). DeShazo et al. (2014) study the California Clean

26The only EV models that are not eligible for the full amount of credits are Honda Accord Plug-in, Ford C-Max Energi, Porsche Panamera, and Toyota Prius plug-in. And their eligible tax credits are $3,626, $4,007, $4,751.8, and $2,500 respectively. In our policy simulation, we remove the tax credits based on different models. 27In a study by Sallee (2011) on the income tax credit on hybrid vehicles after hybrid vehicles were first introduced to the market, he finds consumers captured the majority of the gains for the income tax subsidy. Automobile assembly lines generally operate most efficiently with an output of 200,000 to 250,000 vehicles per platform (Rubenstein 2002). The global sales of the most popular EV model, Nissan Leaf, was only 61,027 in 2014.

25 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 27 of 53

Vehicle Rebate Projects for EVs and ﬁnd a 7% increase in EV sales from the rebate of $1,838 on average. Neither of these studies takes into account indirect network eﬀects and their estimates likely provide the lower bounds of subsidy impacts.

6.2 Policy Comparison

Our stylized model suggests that feedback loops from indirect network effects on both sides of the market have important policy implications. A policy shock on one side of the market would affect the other side. To promote EV adoption based on a variety of rationale as discussed in Section 2.2, policy makers face a problem of optimal policy design in that the tax revenue can be used to subsidize one or both sides of the market. We compare the subsidy policy on EV purchase with an alternative policy of subsidizing charging station investment. The alternative policy uses the same budget of $924.2 million evenly in each quarter during 2011-2013 to install charging stations in all MSAs (proportional to population). As a lower bound estimate of the investment cost of charging stations, we assume the government is only responsible for the purchase and installation of the charging hardware and the charging station company will then operate and maintain the charging stations, as in the case of the EV Project and ChargePoint Project, the two federal charging station support programs. As a robustness check, we also estimate an upper bound investment cost for charging stations assuming the government will also need to maintain and operate those charging stations. The lower bound and the upper bound of the charging station investment cost are $27,000 and $50,244 respectively. 28 The policy comparison between these two policies is provided in Table 9. We assume the policy period from 2011 to 2013 (i.e., no subsidy available in either policy after that). The existing tax credit policy (policy 1) has led to 56,690 more EVs from 2011 to 2013, amounting to $16,303 for one additional EV. The policy effect will continue to exist until the feedback loops die out in 2055. 29 The impact on EV sales from this policy in the long

28According to the charging station cost report by U.S. Department of Energy Vehicle Technologies Oﬃce (2015), the maintenance and operation cost of charging stations include the following components: electricity use, network fee, maintenance and repair, and rents in parking lots. The average electricity consumption is reported to be 6,864 kWh per year with an average network fee of $500 and maintenance fee of $300. Assuming the charging station charges customers $0.39 per kWh (the current charging fee of Blink Network supported by the EV project) and the average annual parking lot rate being $ 1995.12 (the national average parking lot rate in 2012 Colliers International Parking Rate Survey), the estimated operation and maintenance cost less revenue is $805 per charging unit and $2,896 per station (3.6 units per station). Assuming a 10-year life span of a charging station, the discounted total cost of maintenance and operation is $23,244 per station and the total investment cost of a charging station including initial installation is $ 50,244. 29We assume the public charging stations and the installed base of EV drivers keep increasing at a rate that was observed in the last quarter of 2013.

26 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 28 of 53 term would be 184,049, amounting to $5,022 per policy-induced EV purchase. If instead, the government had spent the $924.2 million subsidizing charging stations by purchasing and installing charging infrastructure (policy 2 with lower station cost), EV sales would have increased by 124,904 during these three years. The cumulative impact on EV sales from this policy until year 2055 would be 403,558, amounting to $2,290 per induced EV, only 46% of the unit cost under the existing policy. If the government were also responsible for maintaining and operating those stations (policy 2 with higher station cost), EV sales would have increased by 75,199 during these three years and 267,741 in the long term, amounting to $3,452 per induced EV, still preferable to policy 1. As depicted in Figure 6, policy 2 that subsidizes charging station investment demonstrates a dominant advantage in stimulating EV sales in the early stage of the EV market. The $924.2 million spending during the three years can install about 18,395 to 34,231 charging stations depending on the actual investment cost. This is more than one eighth to one fourth of the total number of gasoline stations in the country and almost one and half to three times of the current total number of public charging stations in the whole country. This large amount of public charging stations should dramatically alleviate or even eliminate range anxiety for potential EV buyers. Our results indicate that building charging stations is a more effective way to boost EV sales in the EV launch stage. As shown in our regression results, indirect network effects on the EV demand side are much stronger than those on the charging station side (coefficient ratio being 1.4) and consumers are not very sensitive to prices (price elasticity being -1.3). As a result, the policy that builds charging stations stimulates EV sales at a much faster pace, consistent with our findings in the model section. The long-run simulations are based on a variety of assumptions and are meant to be illustrative. As the technology improves, the EV driving range is likely to increase, weakening indirect network effects from charging stations to EV demand. In addition, in the longer term, as EVs become more of a serious choice for average vehicle buyers, consumer price sensitivity among EV buyers could increase. Both of these changes would affect the policy outcomes and weaken the effectiveness of policy 2 relative to policy 1. As discussed in Section 5.2, the network size of charging stations has heterogeneous impacts on the EV demand across locations with different average commute time. This implies heterogeneity in the relative strength of indirect network effects on the two sides of the EV market and hence heterogenety in policy comparison. That is, policy 2 may not be always preferred as previous analysis suggests. In MSAs where indirect network effects on EV demand are not strong since drivers have shorter commute and home charging is enough

27 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 29 of 53 to ensure their daily trips, policy 1 that subsidies EV purchase could be more effective. Figure 7 depicts the relative effectiveness of the two policies in promoting EV adoption. The MSA with policy 2 being most effective is New York-New Jersey-Long Island (NY-NJ-PA) where the average commute time is longest, while the MSA with policy 1 being most effective is Grand Forks (ND-MN) where the average commute time is shortest. This suggests that a more elaborate and cost-effective policy design would be to subsidize charging stations in areas with long commute (e.g., large MSAs) but to subsidize EV purchases in areas with short commute (e.g., small MSAs). This type of regionally differentiated policy could be implemented at the federal level but perhaps more feasibly at the state and local levels. For example, in states or cities where average commute is longer, state and local government should focus on building charging station infrastructure while subsidies on EV adoption for example through rebate and HOV lane usage can be implemented in states and cities where average commute is shorter.

7 Conclusion

This study first demonstrates through a stylized model that positive indirect network effects in both EV demand and charging station deployment give rise to feedback loops which amplify shocks to the system and have important policy implications. Although indirect network effects on both sides of the market imply subsidizing either side of the market will result in an increase in both EV sales and charging stations, the relative cost-effectiveness of different subsidy policies depends on consumer price sensitivity for EVs and the relative magnitude of indirect network effects on the two sides of the market. The paper provides to our knowledge the first empirical analysis of indirect network effects in this market and evaluates the impacts of the current federal income tax credit program for EV buyers. Our analysis estimates the elasticity of EV adoption with respect to charging station availability to be 0.84 and the elasticity of charging station investment with respect to the EV installed base to be 0.61. These indirect network effects enhance the effectiveness of the tax credit policy, which has contributed to 40% of the EV sales during 2011 to 2013 and will continue to exhibit a positive effect on the market for many years through feedback loops. Given the relatively strength of indirect network effects on the EV demand side and the low price sensitivity of early adopters, subsidizing charging station deployment would be much more cost-effective than the current policy of subsidizing EV purchases. Our findings offer some insights for policy design to promote the EV technology. First, the policy to expand the charging station network (e.g., through subsidies) would be es-

28 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 30 of 53 pecially effective in the EV launch stage due to low price-sensitivity of early adopters and strong indirect network effects from charging stations on EV demand. Second, our analysis demonstrates that significant spatial differences exist in optimal policy design. Together with the finding from the literature that the environmental benefits from EVs exhibit significant heterogeneity across locations with different fuel mix of electricity generation, the spatial variation in indirect network effects limits one-size-fit-all policies and argues for regionally differentiated policies.

29 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 31 of 53

References

Armstrong, Mark, “Competition in two-sided markets,” The RAND Journal of Eco- nomics, 2006, 37, 668–691.

Bartik, Timothy J., Who Beneﬁts from State and Local Economic Development Policies?, Upjohn Institute for Employment Research, 1991.

Beresteanu, Arie and Shanjun Li, “Gasoline Prices, Government Support, and the Demand for Hybrid Vehicles,” International Economic Review, 2011, 52 (1), 161–182.

Berry, Steven, “Estimating Discrete Choice Models of Product Diﬀerentiation,” RAND Journal of Economics, 1994, 25 (2), 242–262.

, James Levinsohn, and Ariel Pakes, “Automobile Prices in Market Equilibrium,” Econometrica, 1995, 63 (4), 841–890.

Bloom, Nicholas, March Schankerman, and John van Reenen, “Identifying Tech- nology Spillovers and Product Market Rivalry,” Econometrica, 2013, 81, 1347–1393.

Caillaud, Bernard and Bruno Jullien, “Chicken & Egg: Competition among Interme- diation Service Providers,” The RAND Journal of Economics, 2003, 34, 309–328.

CBO (Congressional Budget Oﬃce), “Eﬀects of Federal Tax Credits for the Purchase of Electric Vehicles,” 2012.

Church, Jeﬀrey, Neil Gandal, and David Krause, “Indirect Network Eﬀects and Adoption Externalities,” 2002. Foerder Institute for Economic Research Working Paper.

Clements, Matthew and Hiroshi Ohashi, “Indirect Network Eﬀects and the Product Cycle: Video Games in the US, 1994-2002,” The Journal of Industrial Economics, 2005, 53, 515–542.

Corts, Kenneth, “Building Out Alternative Fuel Retail Infrastructure: Government Fleet Spillovers in E85,” Journal of Environmental Economics and Management, 2010, 59, 219– 234.

and Mara Lederman, “Software Exclusivity and the Scope of Indirect Network Eﬀects in the US Home Video Game Market,” International Journal of Industrial Organization, 2009, 27, 121–136.

DeShazo, J.R., Tamara Sheldon, and Richard T. Carson, “Designing Policy Incen- tives for Cleaner Technologies: Lessons from California’s Plug-in Electric Vehicle Rebate

30 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 32 of 53

Program,” 2014. working paper.

Farrell, Joseph and Garth Saloner, “Standardization, Compatibility and Innovation,” RAND Journal of Economics, 1985, 16 (1), 70–83.

Gallagher, Kelley and Erich Muehlegger, “Giving Green to Get Green? Incentives and Consumer Adoption of Hybrid Vehicle Technology,” Journal of Environmental Economics and Management, 2011, 61, 1–15.

Gandal, Neil, Michael Kende, and Rafael Rob, “The Dynamics of Technological Adoption in Hardware/Software Systems: The Case of Compact Disc Players,” RAND Journal of Economics, 2000, 31, 43–61.

Hagiu, Andrei, “Pricing and Commitment by Two-sided Platforms,” The RAND Journal of Economics, 2006, 37, 720–737.

Heutel, Garth and Erich Muehlegger, “Consumer Learning and Hybrid Vehicle Adop- tion,” 2012. forthcoming, Environmental and Resource Economics.

Holland, Stephen, Erin Mansur, and Nicholas Muller, “Are There Environmental Beneﬁts from Driving Electric Vehicles? The Importance of Local Factors,” 2015. working paper.

Huse, Cristian, “Fast and Furious(and Dirty): How Asymmetric Regulation May Hinder Environmental Policy,” 2014. working paper.

Kahn, Matthew, “Do greens drive Hummers or hybrids? Environmental ideology as a determinant of consumer choice,” Journal of Environmental Economics and Managment, 2007, 54 (2), 129–145.

and Ryan Vaughn, “Green Market Geography: The Spatial Clustering of Hybrid Vehicles and LEED Registered Buildings,” The B.E. Journal of Economic Analysis and Policy, 2009, 9 (2), 1935–1682.

Katz, Michael L. and Carl Shapiro, “Network Externalities, Competition and Compat- ibility,” American Economic Review, 1985, 75 (3), 424–440.

Klier, Thomas and Joshua Linn, “New-vehicle characteristics and the cost of the Corpo- rate Average Fuel Economy standard,” Rand Jounal of Economics, 2012, 43 (1), 186–213.

Langer, Ashley and Shaun McRae, “Fueling Alternatives: Evidence from Naturalistic Driving Data,” 2014. working paper.

Lee, Robin, “Vertical Integration and Exclusivity in Platform and Two-Sided Markets,”

31 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 33 of 53

American Economic Review, 2013, 103, 2960–3000.

Liebowitz, S.J. and Stephen E. Margolis, “Are Network Externalities a New Source of Market Failure,” Research in Law and Economics, 1995, 17, 1–22.

Lund, Hendrik and Willett Kempton, “Integration of renewable energy into the transport and electricity sectors through V2G,” Energy Policy, 2008, 36, 3578–3587.

Mock, Peter and Zifei Zhang, “Driving Electriﬁcation: A Global Comparison of Fiscal Incentive Policy for Electric Vehicles,” 2014. The International Council on Clean Trans- portation White Paper.

Nair, Harikesh, Pradeep Chintagunta, and Jean-Pierre Dube, “Empirical Analysis of Indirect Network Eﬀects in the Market of Personal Digital Assistants,” Quantitative Marketing and Economics, 2004, 2, 23–58.

Parry, Ian and Kenneth Small, “Does Britain or the United States Have the Right Gasoline Tax?,” American Economic Review, 2005, 95 (4), 1276–1289.

, Dirk Heine, Eliza Lis, and Shanjun Li, Getting Energy Prices Right: From Principle to Practice, International Monetary Fund Press, 2014.

Rochet, Jean-Chrles and Jean Tirole, “Two-Sided Markets: A Progress Report,” The RAND Journal of Economics, 2006, 35, 645–667.

Rohlfs, Jeﬀrey, “A Theory of Interdependent Demand for a Communications Service,” Bell Journal of Economics, 1974, 5, 16–37.

Rubenstein, James M., The Changing U.S. Auto Industry: A Geographical Analysis, Routledge, 2002.

Rysman, Marc, “Competition Between Networks: A Study of the Market for Yellow Pages,” The Review of Economic Studies, 2004, 71 (2), 483–512.

Sallee, James, “The Surprising Incidence of Tax Credits for the Toyota Prius,” American Economics Journal: Economics Policy, 2011, 3, 189–219.

Sexton, Steven and Alison Sexton, “Conspicuous Conservation: the Prius Halo and Willingness to Pay for Environmental Bona Fides,” Journal of Environmental Economics and Management, 2014, 67, 303–317.

Stoneman, Paul and Paul Diederen, “Technology Diﬀusion and Public Policy,” Eco- nomic Journal, 1994, 104 (425), 918–930.

32 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 34 of 53

Weyl, E. Glen, “A Price Theory of Multi-sided Platforms,” American Economic Review, 2010, 100, 1642–1672.

Zhou, Yiyi, “Failure to Launch in Two-sided Markets: A Study of the U.S. Video Game Market,” 2014. working paper.

Zivin, Joshua, Matthew Kotchen, and Erin Mansur, “Spatial and Temporal Het- erogeneity of Marginal Emissions: Implications for Electric Cars and Other Electricity- Shifting Policies,” Journal of Economic Behavior and Organization, 2014, 107, 248–268.

33 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 35 of 53

Table 1: Summary Statistics

Variable Mean Std.dev Panel (a): Vehicle Demand Equation Sales of a EV model 9.62 40.01 Gasoline prices ($) 3.52 0.26 EV retail price - tax incentives ($) 33161 18569 No. of charging stations 22.13 45.74 Residential charging stations from the EV project 9.21 65.41 Annual personal income ($) 41607 82536 Hybrid vehicle sales in 2007 945 1859 No. of grocery stores 278 624 Average commute (minutes) 22.96 3.37 College graduate share 0.40 0.07 Use public transport to work share 0.02 0.03 Drive-to-work share 0.88 0.05 Share of white residents 0.78 0.11 Number of observations 14563 Panel (b): Charging Station Equation No. of charging stations 9.94 28.13 No. of EV installed base 134 584 Charging station tax credit (%) 4.56 14.7 Public funding or grants 0.33 0.47 No. of grocery stores 186 455 Hybrid vehicle sales in 2007 568 1354 Current gasoline prices ($) 3.49 0.27 Gasoline price last year ($) 3.25 0.39 Gasoline price two years ago ($) 2.78 0.59 Gasoline price three years ago ($) 2.78 0.58 State EV incentives (rebates+tax credits) ($) 1575 3121 Number of observations 4236

34 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 36 of 53

Table 2: Parameters for Simulating Indirect Network Eﬀects

Coeﬃcients Values Variables Values β1 0.8 p 30,000 β2 -1.5 X 16 β3 1 Z 2 γ1 0.4 γ2 1 δ 0.9

35 Table 3: EV Demand Equation

Variable OLS (a) OLS (b) OLS (c) OLS (d) OLS (e) GMM (f) ln(No. of charging stations) 0.378*** 0.389*** 0.502*** 0.352*** 0.177*** 0.844*** (0.017) (0.017) (0.021) (0.037) (0.036) (0.162) ln(gasoline price)*PHEV 1.095*** 1.186*** 1.922*** 0.095 -0.033 -0.083 (0.203) (0.301) (0.353) (0.208) (0.201) (0.206) ln(gasoline price)*BEV 0.611*** 0.704*** 1.908*** 0.560** 0.386* 0.419* (0.203) (0.300) (0.406) (0.229) (0.217) (0.220) ln(retail price - tax incentives) -0.480*** -0.470*** -0.817*** -1.378*** -1.433*** -1.288*** (0.034) (0.034) (0.179) (0.142) (0.140) (0.131) ln(residential charging from EV project) 0.046* 0.035 0.019 0.026** 0.059*** 0.050*** (0.025) (0.024) (0.022) (0.012) (0.011) (0.010) ln(personal income) 0.667*** 0.664*** 0.729*** 2.018** 1.215* 2.056** (0.162) (0.161) (0.179) (0.852) (0.695) (0.855) 36 ln(hybrid vehicle sales in 2007)*time trend 0.031*** 0.011** (0.003) (0.005) ln(personal income)*time trend 0.003 0.001 Direct Exhibit:

(0.016) (0.020) Testimony Year-quarter fixed effects No Yes Yes Yes Yes Yes Vehicle model fixed effects No No Yes Yes Yes Yes SC-5; of

MSA-model ﬁxed eﬀects No No No Yes Yes Yes Source: D.

Note. -The number of observations is 14563. The dependent variable is ln(EV sales) by model by quarter in a Jester

MSA. Retail prices are manufacturers’ suggested retail prices. Tax incentives include both federal and state-level Li on tax credits and tax rebates associated with EVs. All standard errors are clustered at the MSA level. ln(residential et. behalf al July charging from the EV project) is the logarithm of quarterly number of residential charging stations that are built in Indirect of a MSA by the EV Project, which is managed by ECOtality North America and supported by the U.S. Department 22, SC-NRDC-ELPC Network of Energy to deploy level-2 residential and commercial charging stations in 22 U.S. cities. ln(hybrid vehicle sales in 2016 2007)*time trend and ln(personal income)*time trend are included to address the concern that the charging stations Page - U-17990 trend could be correlated with potential trend of EV demand. The P-value for under-identiﬁcation test for the last 37 Effects column is 0. of 53 *p<.1. **p<.05. ***p<.01. July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 38 of 53

Table 4: First Stage Results for EV Demand Equation

Variable ln(No. of grocery stores)*ln(lagged national stations) 0.139*** (0.023) ln(retail price-tax incentives) -0.193*** (0.054) ln(gasoline price)*PHEV 0.146 (0.125) ln(gasoline price)*BEV 0.064 (0.135) ln(residential charging from EV project) -0.011 (0.010) ln(personal income) -0.980 (1.035) ln(hybrid vehicle sales in 2007)*time trend 0.010*** (0.003) ln(personal income)*time trend 0.018 (0.022) R2 0.677 Note. -The dependent variable is ln(No. of stations). The number of observations is 14563. Standard errors are clustered at the MSA level. The model includes year-quarter fixed effects and MSA- model (e.g., Nissan Leaf in San Francisco) fixed effects. *p<.1. **p<.05. ***p<.01.

37 Table 5.A Additional Speciﬁcations for Vehicle Demand (a) Variable GMM (a) GMM (b) GMM (c) GMM (d) GMM (e) GMM (f) ln(No. of charging stations) -0.282 0.785*** 0.817*** 0.791*** 0.866*** 0.851*** (0.243) (0.229) (0.156) (0.152) (0.165) (0.157) ln(No. of stations)*commute time 0.038*** (0.010) ln(gasoline price)*PHEV -0.086 -0.090 -0.079 -0.075 0.037 -0.093 (0.193) (0.200) (0.205) (0.204) (0.204) (0.245) ln(gasoline price)*BEV 0.389* 0.411* 0.424* 0.440* 0.292 0.440* (0.208) (0.217) (0.221) (0.218) (0.222) (0.256) ln(retail price - tax incentives) -1.324*** -1.294*** -1.214*** -1.300*** -1.499*** (0.130) (0.130) (0.147) (0.130) (0.150) ln(residential charging from EV project) 0.049*** 0.048*** 0.049*** 0.051*** 0.054*** 0.049*** (0.009) (0.012) (0.010) (0.009) (0.009) (0.010) ln(personal income) 1.896*** 1.710* 2.059** 2.026** 2.167** 2.849*** (0.708) (0.958) (0.868) (0.839) (0.859) (1.040) ln(hybrid sales in 2007)*time trend 0.011*** 0.019 0.011** 0.012** 0.011** 0.009 (0.004) (0.024) (0.005) (0.005) (0.005) (0.006) ln(personal income)*time trend -0.001 0.044 0.002 0.003 -0.004 -0.020

38 (0.015) (0.081) (0.020) (0.019) (0.020) (0.027) ln(hybrid sales in 2007)*time trend2 -0.000

(0.001) Direct ln(personal income)*time trend2 -0.003 Exhibit:

(0.005) Testimony ln(No. of stations)*Battery EV 0.094 (0.059) SC-5; of

College shares*time trend -0.023 Source: D.

(0.081) Jester Drive-to-work share*time trend -0.117 Li on (0.116) et. behalf Use public transit to work*time trend 0.071 al July Indirect (0.167) of 22,

White share*time trend -0.100** SC-NRDC-ELPC Network 2016

(0.040) Page

Note. -The dependent variable is ln(EV sales) by model by quarter. The number of observations for columns (a)-(e) - U-17990 39 Effects

is 14563 and that for the last column is 11614 due to missing values for the added controls in small MSAs. Column of

(d) adds two additional IVs including tax credit for charging stations and a dummy for other state funding for 53 charging stations. All specifications include year-quarter fixed effect, and MSA-model fixed effects. Standard errors are clustered at the MSA level. *p<.1. **p<.05. ***p<.01. Table 5.B Additional Specifications for Vehicle Demand (b) Variable GMM (g) GMM (h) GMM (i) GMM (j) GMM (k) ln(No. of charging stations) 0.842*** 0.953*** 1.005*** 0.725*** 0.683*** (0.162) (0.170) (0.281) (0.224) (0.347) ln(gasoline price)*PHEV -0.083 0.099 0.127 0.176 (0.206) (0.211) (0.204) (0.201) ln(gasoline price)*BEV 0.420* 0.590** 0.024 0.238 0.223 (0.220) (0.231) (0.283) (0.218) (0.220) ln(retail price - tax incentives) -1.288*** -1.283*** -0.927*** -1.297*** -1.307*** (0.131) (0.131) (0.268) (0.135) (0.139) ln(residential charging from EV project) 0.050*** 0.049*** 0.027 0.054*** 0.055*** (0.009) (0.011) (0.020) (0.012) (0.011) ln(personal income) 2.058** 2.184** 0.045 1.764** 2.900*** (0.854) (0.933) (1.362) (0.831) (0.835) 39 ln(hybrid vehicle sales in 2007)*time trend 0.011** 0.006 0.013 0.014** 0.015* (0.005) (0.005) (0.008) (0.006) (0.008)

ln(personal income)*time trend 0.001 -0.001 0.014 0.008 0.007 Direct

(0.020) (0.020) (0.032) (0.020) (0.020) Exhibit: Observations 14563 14563 6720 14328 13990 Testimony

Note. -Column (g) employs the logit model from Berry (1994) and the dependent variable is ln(skit)- SC-5; ln(s0mt) where skmt is the market share of model k in market m and time t and s0mt is the share of Source:

of consumers who are not purchasing an EV. Column (h) includes specific time fixed effects for ZEV D.

states to control for potential competition eﬀect from the introduction of more EV models. Column Jester

(i) includes sample of only BEV models. Column (j) and column(k) increase the lag of national Li on number of charging stations to two and three quarters when constructing the instrumental variable. et. behalf al July All specifications include year-quarter fixed effects, and MSA-model fixed effects. Standard errors are Indirect of

clustered at the MSA level. 22, SC-NRDC-ELPC Network

*p<.1. 2016 **p<.05. Page - U-17990

***p<.01. 40 Effects of 53 Table 6: Charging Station Equation

Variable OLS (a) OLS (b) OLS (c) OLS(d) GMM(e) GMM (f) ln(EV installed base) 0.374*** 0.540*** 0.136*** 0.115*** 0.613*** 0.659*** (0.025) (0.028) (0.029) (0.028) (0.157) (0.157) Charging station tax credit (%) -0.005*** -0.003* 0.003 0.003 0.012 0.012 (0.002) (0.002) (0.013) (0.014) (0.014) (0.014) Public funding or grants 0.099* 0.077 0.007 -0.020 0.078 0.088 (0.060) (0.057) (0.048) (0.047) (0.054) (0.055) ln(No. of grocery stores)*ln(lagged national stations) 0.042*** 0.030*** 0.183*** 0.118*** 0.063** 0.058** (0.005) (0.005) (0.017) (0.020) (0.025) (0.025) ln(hybrid vehicle sales in 2007)*time trend 0.018*** 0.009** 0.009* (0.003) (0.004) (0.005) 40 Year-quarter fixed effects No Yes Yes Yes Yes Yes MSA fixed effects No No Yes Yes Yes Yes Overidentification test (p-value) 0.3435 0.1347 Direct

Under-identiﬁcation test(p-value) 0.0000 0.0001 Exhibit: Testimony Note. -The number of observations is 4236. The dependent variable is ln(total number of publicly accessible charging stations) in a MSA in a quarter. Tax credit for charging stations is the state-level tax credit to cover the cost of installing SC-5;

charging stations and is measured in percentage. Public funding is a dummy variable indicating whether public grants of Source: D.

for building charging stations are available. ln(No. of grocery stores)*ln(lagged national stations) is the instrumental Jester variable used in the EV demand equation. Column (e) is the preferred speciﬁcation with gasoline prices in previous years Li on as the IV for ln(EV installed base). Column (f) adds tax incentives for EV purchases as an additional IV. All standard et. behalf al

errors are clustered at the MSA level. July Indirect *p<.1. of 22, SC-NRDC-ELPC

**p<.05. Network 2016 ***p<.01. Page - U-17990 41 Effects of 53 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 42 of 53

Table 7: First Stage Results for Charging Station Equation

Variable ln(current gasoline price) 0.206 (0.233) ln(gasoline price last year) 3.304*** (0.824) ln(gasoline price two years ago) 3.236*** (0.676) ln(gasoline price three year ago) 3.087*** (0.810) ln(hybrid vehicle sales in 2007)*time trend 0.021*** (0.004) ln(No. of grocery stores)*ln(national charging stations) 0.085*** (0.023) Charging station tax credit (%) -0.017 (0.018) Public funding or grants for stations -0.228*** (0.061) R2 0.922 Note. -The number of observations is 4236. The dependent variable is ln(EV stock). The model includes year-quarter fixed effects and MSA fixed effects. Standard errors are clustered at the MSA level. *p<.1. **p<.05. ***p<.01.

41 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 43 of 53

Table 8: Policy Impacts of Federal Income Tax Credits for EVs

Time Observed EV Sales Counterfactual Sales Sales Reduction Percentage 2011-1 1,105 772 333 30.10% 2011-2 3,241 2,580 661 20.40% 2011-3 2,813 1,887 926 32.91% 2011-4 3,900 2,256 1,644 42.16% 2012-1 4,307 2,015 2,292 53.22% 2012-2 7,030 3,517 3,513 49.97% 2012-3 9,662 5,575 4,087 42.30% 2012-4 12,665 7,838 4,827 38.11% 2013-1 21,140 12,931 8,209 38.83% 2013-2 24,803 15,571 9,232 37.22% 2013-3 25,782 15,679 10,103 39.19% 2013-4 23,747 12,884 10,863 45.74% Total 140,195 83,505 56,690 40.44% Note. -Counterfactual sales are the simulated sales in all 353 MSAs in our data after removing the federal income tax credit for EV buyers (the amount based on diﬀerent models) while holding everything else the same.

42 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 44 of 53

Table 9: Comparison of EV Income Tax Credit and Charging Station Subsidy Policies

EV Sales Increase EV Sales Increase EV Sales Increase from Policy 1 from Policy 2 from Policy 2 Low Cost High Cost 2011-1 333 2,532 1,439 2011-2 661 3,839 2,214 2011-3 926 4,116 2,422 2011-4 1,644 5,937 3,505 2012-1 2,292 6,500 3,891 2012-2 3,513 8,599 5,185 2012-3 4,087 9,074 5,482 2012-4 4,827 9,856 5,966 2013-1 8,209 17,832 10,762 2013-2 9,232 18,340 11,088 2013-3 10,103 18,880 11,443 2013-4 10,863 19,397 11,801

Sales increase in 3 years 56,690 124,904 75,199 Total increase long-term 184,049 403,558 267,741 Total increase in 10 years 168,131 373,748 245,687

Government spending per EV $5,022 $2,290 $ 3,452 Note. -Both policies use $924.2 million. Policy 1 provides rebate in the form of tax credits for EV buyers with the amount as the same as the current income tax credit policy for each EV model. Policy 2 uses the same budget to build charging stations evenly in each quarter in all metro areas weighted by population. We assume investing a charging station has a lower-bound cost of $27,000 and an upper-bound cost of $50,244. Total increase includes vehicle sales increase during the policy eﬀective period and the increase in future years (till 2055) discounted to year 2011 by a 5% discount rate.

43 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 45 of 53

Figure 1: National Quarterly EV Sales and Public Charging Stations

Source. -Author’s calculations using Hybridcars.com monthly sales dashboard data and electric charging station location data by Alternative Fuel Data Center of the Department of Energy. Note. -The quarterly EV sales plotted include both BEV and PHEV sales.

44 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 46 of 53

Figure 2: Spatial Distribution of EVs and Public Charging Stations

Panel (a) Installed Base of EVs Per Million People

Panel (b) Public Charging Stations Per Million People

Note. -Map boundaries deﬁne metropolitan statistical areas. Both graphs are shown for the fourth quarter of 2013).

45 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 47 of 53

Figure 3: Steady State Solution and Stability

Note. -qt is the number of new EV sales in each period. qt = G(qt−1) is the implicit function deﬁned in equation (3). q∗ is the steady state solution.

46 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 48 of 53

Figure 4: Impacts of Income Tax Credits (5 periods) under Feedback Loops

Panel (a) EV Sales Increase Due to Feedback Loops

90 EV sales without subsidy EV sales with subsidy 80

EV sales 50

20 10 20 30 40 50 60 70 80 Time

Panel (b) Charging Stations Increase Due to Feedback Loops

120 Charging stations without subsidy 110 Charging stations with subsidy

100

60 Cumulative stations 50

20 10 20 30 40 50 60 70 80 Time

Note. -EV sales and charging stations without subsidy are solutions given by the system defined in Section 3. The simulated subsidy effects are due to a policy design that gives EV buyers a tax credit of $7,500 for the first five periods. The parameters for simulations are reported in Table 2 and the intuitive findings remain with different assumptions of the simulation parameters.

47 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 49 of 53

Figure 5: Policy Comparison and Relative Strengthen of Indirect Network Eﬀects

3.5 Price elasticity −1.4 Price elasticity −1.5 3 Price elasticity −1.6

2.5

1.5

0.5 Ratio of policy effects (policy2/policy1)

0 0.8 1 1.2 1.4 1.6 1.8 2 Ratio of indirect network effects parameters (β1/γ1)

Note. -This figure depicts the relationship between the relative policy effects of two subsidy designs and the relative strengthen of the indirect network effects on both sides of the EV market. Policy 1 subsides the EV purchase by tax credits and policy 2 uses the same amount of funding subsidizing charging stations. Given a price elasticity of EVs, Policy 2 becomes more and more effective than Policy 1 when the effect of charging stations on EV demand (denoted by β1) becomes larger relative to the effect of EV stock on charging stations (denoted by γ1). When the magnitude of the price elasticity increases (decreases) and consumers are more (less) sensitive to prices, policy 2 becomes less (more) effective relative to policy 1 for a given ratio of β1/ γ1.

48 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 50 of 53

Figure 6: Sales Impacts from Two Subsidy Policies 20000 15000 10000 5000 EV sales increase due to policy 0

2011 1st 2011 2nd 2011 3rd 2011 4th 2012 1st 2012 2nd 2012 3rd 2012 4th 2013 1st 2013 2nd 2013 3rd 2013 4th Year−Quarters

policy 1 policy 2 (lower station cost) policy 2 (higher station cost)

Note. -Each data point represents EV sales increase by quarter due to the policy. Policy 1 gives new EV buyers a tax credit of $2,500-$7,500 based on diﬀerent models as the current income tax credit policy for EVs. Policy 2 builds charging stations in all MSAs with the same total spending as policy 1 by assuming a charging station investment cost with a lower bound cost of $27,000 and an upper bound of $50,244.

49 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 51 of 53

Figure 7: Heterogeneous Policy Eﬀectiveness (Policy2/Policy1)

Note. -Policy 1 gives new EV buyers a tax credit of of $2,500-$7,500 based on different models as the current income tax credit policy for EVs. Policy 2 builds charging stations in all MSAs with the same budgetary cost as policy 1 assuming the investment cost per station being $27,000. The figure plots the ratio of the EV increases due to the two subsidy policies. The policy effectiveness of policy 2 varies across locations due to the heterogeneous impacts of public charging stations on the EV demand. The regions with the dots are locations where policy 1 is more effective.

50 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 52 of 53

Appendices

Entry Model of Charging Stations

The entry model is developed based on Gandal et al. (2000). Denote EV owners’ demand for

charging station j by D(p1, ..., pN ), where N is the number of charging stations available in a

given market, and pj is the price at charging station j, j = 1, ..., N. We assume that demands are symmetric in terms of prices at different charging stations. Furthermore, we assume that the marginal cost is constant, denoted by c, and that the profit function from each EV owner (pj − c)D(p1, ..., pN ) is quasi-concave in pj. Under these assumptions, there exists an equilibrium in which all stations charge the same price which depends on N, denoted by D(p) 0 p(N). Denote the equilibrium markup by ϕ(N)(≡ − ∂D(p) ) and assume ϕ (N) < 0, which [ ∂p ] is consistent with most common competition models. Let f(N) = ϕ(N)D(p(N))/N. Then EV the per-period profit of a station in market m at time t is πmt = Qmt f(Nmt). Now consider a charging station’s entry decision. If a charging station enters market m

at time t, it pays the entry cost Fmt and earns the proﬁt streams (πmt, π(mt+1)...), generating

a discounted profit of −Fmt + πmt + δπmt+1 + ..., where δ is a discount factor common to all stations. If a station enters market m at time t + 1, it generates a discounted profit of 2 −δFmt+1 + δπmt+1 + δ πmt+2.... In a free-entry equilibrium firms must be indifferent between these two options. This implies

EV Fmt − δFmt+1 = πmt = Qmt f(Nmt).

Taking the natural logarithms of both sides, we can get

EV ln(Fmt − δFmt+1) = ln(f(Nmt)) + ln(Qmt ) (6)

We specify the entry cost as

ln(Fmt − δFmt+1) = ρ0 + ρ1Zmt + τmt, (7)

where τmt is the unobserved entry cost, and the vector of covariates Zmt includes the state- level tax credit given to charging station investors measured as the percentage of the building cost, a dummy variable indicating whether there exists public grants or funding to building charging infrastructure, the interaction term of number of grocery stores in a MSA in 2012 with the lagged number of charging stations in all MSAs other than own (the instrument in the EV demand equation), and other control variables.

1 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-5; Source: Li et. al Indirect Network Effects Page 53 of 53

We specify the proﬁtability of charging station f(N) as follows

EV ln(f(Nmt)) = λ0 + λ1ln(Qmt ) + ϑmt, (8) where Nmt is installed base of charging stations by time t which captures the competition among charging stations, and ϑmt is an error term that captures the unobserved local demand shocks. Furthermore, we decompose the local shock as

(τmt − ϑmt)/λ1 = Tt + ϕm + ςmt, (9) where Tt is year-quarter dummies that control for time eﬀects common to all the MSAs,

ϕm is market fixed effects that control for time-invariant and MSA-specific preferences for charging stations, and ςmt is an term capturing those idiosyncratic local demand shocks. From equations (6), (7), (8), and (9), we can obtain the charging station equation (5) in Section 4.2:

EV ln(Nmt) = γ0 + γ1ln(Qmt ) + γ2Zmt + Tt + ϕm + ςmt, where γ0 = (ρ0 − λ0)/λ1, γ1 = −1/λ1, γ2 = ρ1/λ1.

2 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-6; Source: 17990 ELPC-CE-98 Page 1 of 1

17990-ELPC-CE-98

Question:

11. Wbat analysis has the Company relied on regarding the placement of the 750 240 V charging stations?

a. What will be the strategies for reaching ont to potential businesses for possible placement? b. What criteria will be used for deciding between different locations? c. Will there be an effmi to locate stations at apartment buildings and condominiums? d. Will the company consider additional 110 V charging outlets at locations with longer dwell times to increase the number of charging options (eg. workplaces)?

Response: a. Consumers will utilize various sources to solicit interest from potential site hosts. Where applicable, Consumers may utilize local publications I Chambers of Commerce, and email communications to business customers. We will also utilize information collected through Customer Account Managers and from customers who have previously expressed interest. b. The selection process will consider factors such as: level of accessibility to the general public, proximity to existing stations, public safety, and areas of higher traffic frequency. c. The cunent infrastmcture program does not include a segment dedicated to multi-family applications. The current program is sttuctured to expand PEV infrastructure to the broad base of customers, whereas multifamily residence applications could minimize the accessibility of those stations to a select few. In most cases, customers residing in multifamily residences will be eligible for the residential reimbursement component of the program. d. Yes, utilization of 110 V charging outlets will be incorporated into the Work Place Charging segment of the program.

Customer Services Department

1 99000137 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-6; Source: 17990 ELPC-CE-98 Page 1 of 1

17990-ELPC-CE-98

Question:

11. Wbat analysis has the Company relied on regarding the placement of the 750 240 V charging stations?

Customer Services Department

1 99000137 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-7; Source: 17990-ELPC-CE-93 Page 1 of 1

17990-ELPC-CE-93

Question:

6. What analysis has the Company relied on regarding the placement of the charging stations along major highways across the Lower Peninsula?

Response:

6. There was no one specific document used to draft the preliminary target points for charging stations. In general, the Company analyzed high traffic areas and/or areas where there was limited charging station infrastructure in place.

uho H. Morales May18,2016

Customer Services Department

6 99000132 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 1 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 2 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 3 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 4 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 5 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 6 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 7 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 8 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-8; Source: Partial Lefkowitz Testimony from Maryland PSC Case 9418 Page 9 of 9 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 1 of 9

17990-MEC-CE-49

Question:

5. Refer to the bullet point items listed on page 44 of Andrew Bordine’s testimony, lines 3- 12:

a. Has Consumers quantified, projected, or estimated the energy loss benefits of any of these items? b. If yes, please provide. c. Has Consumers incorporated or otherwise considered the inclusion of any of these items in any projection or estimate of future line or system loss projections? d. If yes, please provide. e. Has Consumers begun or completed any updated line or system loss study? f. If yes, please provide.

Response:

a. Yes, the Company has estimated the energy loss benefits for these items. b. The Company has estimated a1% reduction in system losses from these items. c. No. The System Loss Studies conducted by the Company are not forward looking as suggested in this interrogatory. The System Loss Study is done based upon historical information. d. See part c, above. e. The Company has completed the 2014 System Loss Study. f. Please see the attached document labeled 17990-MEC-CE-49 Attachment.

/05&"UUBDIFEBSFOVNCFSFEEPDVNFOUTUISPVHI

______Andrew J. Bordine May 27, 2016 Customer Management and Grid Infrastructure Department

99000160 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 2 of 9

99000161 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 3 of 9

99000162 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 4 of 9

99000163 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 5 of 9

99000164 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 6 of 9

99000165 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 7 of 9

99000166 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 8 of 9

99000167 July 22, 2016 - U-17990 Direct Testimony of D. Jester on behalf of SC-NRDC-ELPC Exhibit: SC-9; Source: 17990-MEC-CE-49 Page 9 of 9

99000168 STATE OF MICHIGAN

MICHIGAN PUBLIC SERVICE COMMISSION

In the matter of the Application of Case No. U-17990 CONSUMERS ENERGY COMPANY for Authority to increase its rates for the ALJ Dennis W. Mack generation and distribution of electricity and for other relief.

PROOF OF SERVICE

On the date below, an electronic copy of Testimony of Douglas Jester on behalf of Sierra Club, Natural Resources Defense Council, Environmental Law and Policy Center, and Michigan Environmental Council, along with Exhibits SC-1 through SC-9 was served on the following:

Name/Party E-mail Address Administrative Law Judge Dennis W. Mack [email protected] Counsel for Consumers Energy Company [email protected] Robert W. Beach [email protected] H. Richard Chambers [email protected] Kelly M. Hall [email protected] Gary A. Gensch [email protected] Anne M. Uitvlugt [email protected] Bret A. Totoraitis [email protected] Counsel for MPSC Staff Spencer A. Sattler [email protected] Heather M.S. Durian durianh@ michigan.gov Meredith R. Beidler [email protected] Counsel for Michigan Cable Telecommunications Association David E. S. Marvin [email protected] Charles E. Dunn [email protected] Counsel for Charge Point and Energy Michigan, Inc. Laura A. Chappelle [email protected] Timothy J. Lundgren [email protected] John W. Sturgis [email protected] Counsel for Hemlock Semiconductor Corp Jennifer Utter Heston [email protected] Counsel for Michigan Utility Workers Council, Utility Workers Union of America, AFL-CIO John R. Canzano [email protected] Lilya n N. Talia [email protected] Counsel for Environmental Law & Policy Center Margrethe Kearney [email protected] Counsel for The Kroger Company Kurt J. Boehm [email protected] Jody Kyler Cohn [email protected] Counsel for Attorney General John A. Janiszewski [email protected] Celeste R. Gill [email protected] Counsel for ABATE Robert A. W. Strong [email protected] Michael J. Pattwell [email protected] Leland Rosier [email protected] Midland Cogeneration Venture Limited Partnership David R. Whitfie ld [email protected] Residential Customer Group/Michele Rison Don L. Keskey [email protected] Brian W. Coyer [email protected]

The statements above are true to the best of my knowledge, information and belief.

OLSON, BZDOK & HOWARD, P.C. Counsel for Sierra Club, Natural Resources and Defense Counsel, Environmental Law and Policy Center and Michigan Environmental Council

Date: July 22, 2016 By: ______Karla Gerds, Legal Assistant Kimberly Flynn, Legal Assistant 420 E. Front St. Traverse City, MI 49686 Phone: 231/946-0044 Email: karla@envla w.co m and [email protected]