Potential to Electrify Miles with Different Plug-in Vehicle Innovation Paths. D. J. Santini, and Y. Zhou

Transcription

1 1 Potential to Electrify Miles with Different Plug-in Vehicle Innovation Paths D. J. Santini, and Y. Zhou Danilo J. Santini Phone: (703) Center for Transportation Research Argonne National Laboratory 9700 South Cass Avenue, Bldg 362 Argonne, Illinois Yan Zhou* Phone: (630) ; Fax: (630) Center for Transportation Research Argonne National Laboratory 9700 South Cass Avenue, Bldg 362 Argonne, Illinois *Corresponding author Presented at the 94th Transportation Research Board Annual Meeting Words: (4 Tables + 2 Figures) =7492 The submitted manuscript has been created by UChicago Argonne, LLC, Operator of Argonne National Laboratory ( Argonne ). Argonne, a U.S. Department of Energy Office of Science laboratory, is operated under Contract No. DE-AC02-06CH The U.S. Government retains for itself, and others acting on its behalf, a paid-up nonexclusive, irrevocable worldwide license in said article to reproduce, prepare derivative works, distribute copies to the public, and perform publicly and display publicly, by or on behalf of the Government.

2 2 ABSTRACT With their high energy and power capabilities, lithium ion batteries allow many powertrain combinations and permutations. California regulators credit two plug-in technology innovation paths: (1) PHEVs spunoff from HEVs; (2) battery electric vehicles (BEV) and REX/BEVx spun off from BEVs, where x signifies limited gasoline engine range extension. A third path is the Super EV with 200+ mile range and >302 hp, far more than otherwise available BEVs. Data on 2014 commercial plug-in vehicles is presented. Past cost of ownership studies focusing on path 1 PHEVs and path 2 BEVs are discussed. Price and marketability implications are examined. A thought experiment is derived, informed by marketability and financial payback considerations, assuming regular use of % of a BEV battery pack and 100+% of a PHEV or REX/BEVx pack. The BEV is chosen by customers driving less miles than the hypothetical maximum electric range; PHEVs and REX/BEVx vehicles by those driving further. Promising potential to electrify miles nationwide is estimated with modified option 2 BEVx designs. Financially, the best path 2 market segment is new, low density single family construction where the vehicle makes long commutes at relatively high average speed. The REX/BEVx range extension feature can be effective because a significant proportion of total national miles (42%) is caused by a relatively few vehicles (12%) driving beyond the range of most currently available BEVs (~ 70 miles). Some 45% of those miles could be electrified by a BEVx with 70 miles of electric range with one charge per day. Adaptation of BEVx range extension would enhance REX/BEVx marketability, increasing national fleetwide GHG and oil use reduction.

3 3 BACKGROUND Lithium ion (li-ion) battery technology can support a wide variation in degrees of electrification of vehicles. A property of lithium ion battery packs is that power (kw) and energy storage (kwh) are positively linked (1). This property has been used to advantage in what some have called the Super EV, the Model S Tesla, which has the largest battery pack of any commercial electric car (Table 1), with performance car acceleration. It takes advantage of a lightweight aluminum body and very low aerodynamic drag. Compared to conventional (steel bodies with higher aerodynamic drag) gasoline cars with similar acceleration performance, GHG benefits of the Model S have been estimated to be quite dramatic (2). Unfortunately, the high performance premium vehicle market is a limited niche market. One question is whether the technology success of the Model S will trickle down to the mass market with steel vehicle bodies. Most electric vehicles have battery packs with far less kwh than the Tesla, and less kw. They generally use steel bodies and are predominantly variations of conventional gasoline vehicle models. However, a recently introduced EV, the BMW i3 model, uses a combination of carbon fiber body panels and aluminum frame, resulting in light weight other than the battery pack, achieving pack weight reduction (3). Among 2014 EVs it is the most efficient and has lowest GHGs (Table 1) (4, 5). Unlike any other plug-in electric vehicle, this version also has a range extender (REX) option, in which a small engine and generator are added. Range in the REX mode is approximately equal to the electric range (5), and is limited by gasoline tank size, chosen to meet California Air Resources Board (CARB) BEVx regulations (6). These regulations also define technical requirements for plug-in hybrids (PHEVs) to obtain credit for all-electric range. PHEV in this paper refers only to qualifying vehicle designs. California regulations adopted in a Memorandum of Understanding by seven other states (12) presently allow half of the zero emissions vehicle category (EVs and fuel cell vehicles) credit to BEVx vehicles. The performance and range in x mode (hybrid mode) is intentionally restricted to cause a strong incentive for drivers to recharge as soon as possible (6). Further, the control strategy is required to delay engine starts until the battery pack is nearly depleted, so that all-electric operation may be maximized. In (7) it is suggested that this configuration could easily be modified by changing control strategy to turn the engine on sooner and operate in hybrid (HEV) mode (a suggestion made to CARB by GM in (6)). A larger gasoline tank could be incorporated, allowing range like that provided by PHEVs and conventional vehicles. Such REX configurations should increase marketability in the rest of the U.S. compared to those meeting CARB BEVx requirements. A recently completed international study of PHEVs used vehicle simulation and cost projection to evaluate a large number of alternative plug-in electric powertrains in comparison to hybrids (HEVs) and to one conventional powertrain (7). It estimated that an input-split PHEV with approximately 30 mile range would be the most financially desirable powertrain in the U.S. for consumers that average miles per day, should gasoline prices rise to $5.00/gallon (7, 8). It also estimated that a simulated steel body all-electric vehicle could be the best alternative if driven miles per day locally (no inter-city driving). Only two of 27 intensively commuting vehicles examined over a year of use in the Commute Atlanta study (9) were exclusively used in intra-metro driving (10). That study, augmented by analysis of the National Household Travel Survey (NHTS) (11) (Table 2) implies that a significant percentage of inter-city driving is necessary (~20% on average). When shares of inter-city travel are this high (8), the international study estimated that range and refueling time drawbacks consistently make the BEV less financially attractive than a PHEV.

4 4 Recent evaluations of advanced electric powertrains tacitly recognized that standard all electric vehicles with ranges of miles and normal aerodynamic properties won t be driven as many annual miles as the typical vehicle (13, 14). Ref (13) assumed 15% lower annual BEV miles (9000 mi.), while the study overseen by Ward and Joseck (14) assumed 31% less. Ref (7) evaluated three cases: (1) identical local driving by BEVs and other vehicles (2) 8.5% more driving (inter-city) by non-bevs and (3) 19% more driving by non-bevs (inter-city). In cases 2 and 3, cost of use of rental cars for inter-city driving was used to achieve parity in annual miles for the total cost of ownership estimates. In 2012 CARB suggested delaying use of assumptions about an unknown second vehicle, pending field data for BEVs (6, p. 54). Recently it was reported that EV Project Leafs averaged 808 miles per vehicle month for cases with good daily records while Volts averaged 1020 miles per vehicle month, 759 (74%) electric (15). On the basis of driving patterns when good daily records were available, Volts drove about 21% more total miles than Leafs and 6% less electric miles. However, for the 15 month reporting period, there was an average of 7.0 vehicle months of use per year for the participating Nissan Leaf BEVs, against 8.8 for Volts. If vehicle months is a direct translation of days of actual use into month equivalents, the Volt rate of generation of good daily records would be about 268 days per year, the Leafs 213. Data reporting problems caused an unknown degree of understatement of rates of use. PRIOR INVESTIGATION OF FINANCIAL VIABILITY OF ELECTRIC DRIVE A feature of recent financial viability evaluations of EVs (7,8,16) is that they imply that unsubsidized BEVs will need to be driven longer local daily distances than PHEVs (input-split technology with < 60 kw electric drive) to be financially viable, due to estimates of higher BEV cost when 70+ miles of range is available. All else equal, as incremental cost increases, more driving is necessary for financial payback. Implications are opposite of deductions often made via engineering assumptions simply considering technical feasibility to cover distance (17). Such ad hoc judgments assume that electric drive vehicles will be owned and used by those who do not drive intensively. Ref (7, 8 and 16) estimate a relatively high cost of ownership for unsubsidized extended range electric vehicles (EREVs such as the Chevrolet Volt) with high electric drive power (> 100 kw) combined with sophisticated hybrid powertrains. Lower electric drive power levels just sufficient to assure consistent all-electric operation in everyday PHEV driving with smaller (8-9 kwh) and less powerful (60kW) battery pack have been estimated to have lowest unsubsidized total cost of ownership when gasoline prices are $5/gallon and daily driving is miles (8). The Ford Fusion Energi is most similar to this technology, but has less range, due in part to a pack with less energy (7.5 kwh), and to more weight. Ref (18) addresses efficiency enhancing regenerative braking benefits of battery pack power levels. Benefits end after a power level far lower than used in EVs has been reached (but higher than 35 kw). A 2011 financial viability analysis (19) examined the daily distance categories 0-50 and 50+ miles, estimating that electric drive would not be financially attractive for the average distance in the 0-50 distance. More recent 2013 analysis (8) disaggregated the 0-50 mile category into 0-30 and mile s, with very favorable implications for HEVs relative to conventional drive within miles. The study also benefitted from improved battery pack cost modeling (1). Although HEV vs. conventional vehicle results did not differ in the sense of statistical significance, the 2013 revision resulted in estimates that HEVs had lower total life cycle cost than conventional vehicles for the 30-50

5 5 mile daily distance at current gasoline prices. This implied considerably greater viability for electric drive than the 2011 study. By helping achieve economies of scale, financial viability for HEVs can be complementary to financial viability of PHEVs using essentially the same powertrain electric machinery and power electronics, such as the Ford Energi system used in both the Fusion and C-Max, and the Toyota hybrid system used in the Prius, Plug-in Prius and Prius V. The 2011 study estimated that the highest benefit to cost ratio would be found for electric vehicles traveling beyond their range of 76 miles and charged twice per day. The benefits relative to conventional vehicles were dramatic. Benefits exceeded those for HEVs and PHEVs. The 2011 and 2013 financial viability analyses relied on vehicle simulations, accounting for real world driving. Many of the simulated vehicles with electric drive now have approximate counterparts in the marketplace, along with official estimates of real world electric and fuel consumption developed by the Department of Energy and U.S. Environmental Protection Agency (Table 1 (4, 5)). FALLACY OF USING VEHICLE COUNTS LESS THAN A HYPOTHETICAL RANGE It is conventional wisdom that a count of the share of vehicles driven less than a given number of miles daily represents a measure of feasibility of implementing electric miles. The following recent quote illustrates the fallacy: The powertrain chief says that the customer with up to a 48km (30 mile) commute to work represents 40-50% of commuters across the U.S.A. (17, p. 41) The key words are up to. An examination of Table 3 shows an estimate that the percentage of vehicles with up to 30 mile daily driving distance from the residence and back (home-to-home) is 62%. Unfortunately, it also illustrates that the percentage of sample miles represented is only 27%. Further, the prior financial analysis estimates that HEVs, PHEVs, and BEVs driven up to 30 miles every day will have a higher total cost of ownership (TCO) than gasoline vehicles. Fig. 1 also shows the difference in vehicle counts and mile counts as a function of daily distance driven for the NHTS one-day sample. Work and non-work vehicle use is presented separately. Comparing the two frames illustrates the shift of shape of the daily distance curve when average distances over many months are used rather than a sample including all single day of use values.. CAUSES OF RELATIVE ADVANTAGE OF HEV, PHEV, AND BEV POWERTRAINS Driving Pattern as a Determinant of Relative Advantage In view of the high costs of electric drive powertrains, financial viability requires that vehicles with such powertrains be driven more intensively than competing conventional powertrains (7). This means that it is desirable for these vehicles to be driven as many days per year as possible (20), as well as many miles per day. In the case of plug-in vehicles that primarily charge overnight at the residence, the more days with morning departure and evening return, the better. Further, ideally the day s travel uses as large a fraction of battery pack capacity as possible. Travel departing from a residential plug and returning at night is home-to-home travel (see Table 2). Additionally, home-to-other and other-to-home categories are broken out. The former category is also one allowing fully charged morning departure.

6 6 Compared to Vans and SUVs, cars more frequently return to the residence at night (Table 2). Consistently, 90% or more of miles are accumulated on a day when charging at the residence would be possible (i.e. home-to-home and home-to-other). Home-to-other, or other-to-home have a much lower fraction of daily miles than home-to-home travel (Table 2). Nevertheless, many work trips occur when the vehicle is used for this travel. Although the fraction of work trips is lower, these work trips are longer, so the average work trip distance per day is about as much in home-to-other and other-to-home as for home-to-home trips. The fraction representing overnight shift work is unknown. The NHTS daily sample used to compile Table 2 undercounts long distance trips involving other-to-other travel. In (10) travel of sample vehicles within Atlanta was separated into intra-metro travel and out of metro area tours. Given differences in definitions in that study and this, estimated shares of miles driven on the out-of-atlanta tours (10) look about the same as here (Table 2). However, the miles per day for out of metro area travel in Atlanta were considerably higher than estimated here. Both studies imply that Vans and SUVs are driven further than cars when going out of metro area. The Atlanta study implies that there are more out of metro area trips for these hauling (passengers, luggage) vehicles. The two studies imply car driving patterns could be met with less charging infrastructure than needed to serve plug-in SUVs and Vans. A large fraction of customers in the U.S. will occasionally drive their vehicles on Interstate highways between cities, often with multiple passengers and luggage, if not towing a boat or camper. Providing for inter-city driving needs is challenging for electric drive vehicles for multiple reasons. In PHEVs there is generally loss of luggage space (Table 1) and reduction in engine power relative to a conventional vehicle, causing performance penalties under heavy load. Most HEVs also have this disadvantage, due to use of less powerful engines to increase everyday urban efficiency. On Interstate highways, PHEVs will be driving in HEV mode. Compared to HEVs the PHEV (or EREV or REX/BEVx) often have an additional fuel economy penalty when in HEV mode (Table 1) due to the battery pack weight. BEVs are recognized as simply being inadequate for towing (21). However, BEVs should be as suitable as more capable conventional vehicles when used for lightly loaded leisurely recreation trips outside of metro areas. To date, all forms of electric drive have been predominantly used in cars. Over the last decade, the market share of HEV SUVs rose and then fell. The HEV share steadily increased in cars (22). The car oriented patterns of production and sales of electric drive vehicles are considered marketplace recognition of relative advantage (and disadvantage). Plug-in Hybrid, Extended Range Electric, and Range Extended Electric For a PHEV, extended range electric (EREV), and electric range extended vehicle (REX or BEVx) it is financially desirable that the vehicle go beyond its electric range every day. In Table 3 for home-to-home trips, the share of miles electrifiable if every trip beyond the hypothetical EV range of a PHEV/EREV/REX is estimated ( PHEV or REX only column). The estimate accounts for the fact that the first n miles for all days of use will be all-electric. The estimate, based on an assumption of one charge per day, implies maximum miles electrifiable (if all PHEV/EREV/REX/BEVx vehicles had the same all-electric range) would be 30% for 30 mile range. These vehicles also realize a benefit relative to a gasoline powertrain when driven in HEV mode, after completing operation in EV mode. Net dollar fuel saving benefits of operation of a Ford PHEV with 20 miles of all-electric range is illustrated in Fig. 2, when driven 1.5 hours/day at speeds from approximately 20 to 60 miles per hour. 1.5 hours per day was chosen because this approximately represents average hours of driving for cars, regardless of population

7 7 density where the vehicles are driven (Table 4). The target market is for 1.5 hours per day of driving and above. In view of the fact that PHEV/EREV/REX/BEVx vehicles can run on gasoline, it is easy for the consumer to drive past the electric range in these vehicles, without range anxiety. Official range ratings are provided on the fuel economy.gov website (Table 1). Due to variation in daily driving (i.e. many days with few miles, Fig. 1), estimates in Table 3 must be used as relative, rather than absolute. No vehicle is consistently driven the daily distance of any distance. Thus, computations probably overstate the miles electrifiable by PHEV/EREV/REX/BEVx vehicles with a given range. It is certain that some days of travel will be shorter than the maximum all-electric range. Results imply a relatively small share of miles electrifiable by such vehicles with miles of all-electric range. The Table 3 column HEV share of PHEV or REX shows estimated share of miles operated in HEV mode. For EV range of up to 40 miles, over half of miles would be in HEV mode. This illustrates the importance of maintaining high efficiency in HEV mode for highest net fuel savings. Electric and Super Electric Vehicles For BEVs, one cannot drive beyond the vehicle s range without a second charge. Nevertheless, the range capabilities of BEVs (Table 1) when driven in relatively dense areas nearly as slowly (~ mph) (20) as rated city driving (2, ~ 20 mph) are adequate for typical intra-metro driving. Unfortunately, with higher densities small fractions of vehicles are driven fewer miles per day, so fuel savings/day is smaller (Table 4). The challenge for BEV owners is to maximize use of battery pack capacity, incentivizing driving at higher speeds and longer daily distances (Fig. 1 and Fig. 2). This contrasts with the Ford PHEV and Chevrolet EREV, where fuel savings benefits per day decrease steadily as the share of HEV miles increases after EV range has been exceeded (Fig. 2). The most common highway cycle BEV range in Table 1 is between 70 and 80 miles. Home-toother and other-to-home car distances in Table 2 indicate the average car will have an out-of town travel day distance around 80 miles. However, cold and hot temperatures and/or snowstorms will occasionally reduce the range below 70 miles. The next generation of standard EVs could well include a REX/BEVx option and have a rated EV range of about 75 miles (engine heat may reduce cold weather range penalties relative to the BEV version) to meet the minimum requirements of CARB regulations for BEVx credit. The BMW i3 REX is already designed in this manner. Fig. 2 illustrates the fuel savings as a function of average speed, if driven 1.5 hours, for the REX in comparison to the gasoline 328i. Table 4 illustrates those cars in the lowest density areas average about 36 mph when driven to work. This results in an estimated savings per day of $5.00/day (Fig. 2). The BMW 328i base list price is $37,400; the i3 BEV $41,300, and the i3 REX $45,200. With 272 charges per year (20) at $5.00/day and ten years of use with a discount rate of 4-8%, the fuel savings would be worth $11,000 to $9000. On this basis the net ten-year cost of the i3 would be competitive with the reference 3 series model (least expensive is the 320i at $32,750). A potential starting market for BMW consumers would be new houses in outskirts of metro areas where long work commutes make the i3 models competitive. Charging infrastructure, when installed in new single family homes during construction, is least expensive and would possibly be available for under $1000. In the case of Ford customers, if the smaller Focus was acceptable, net financial benefits of selecting the BEV Focus instead of the Fusion Energi (Table 1, Fig. 2) would be several thousand dollars. Tesla Super EV cars (Table 1) have a range of over 200 miles. Nearly all daily driving requirements can be met with this daily driving distance capability, though the Commute Atlanta

8 8 investigation implies that Vans and SUVs would considerably more often find such a range limitation problematic for inter-city travel (10). The requirement for intra-metro charging of Super EVs should be minimal, since nearly all intra-metro travel should be possible with a full overnight charge. However, aggressive driving and extreme temperatures could cause some Super EV owners to appreciate the ability to have an occasional intra-metro charge. While the Super EV has the capability to serve nearly all intrametro driving needs, it is expensive, largely limiting its market to those otherwise purchasing high performance premium cars. The list price of the Model S is competitive with many such cars imported to the U.S. Thus, even ignoring ten thousand and more dollars of lifetime fuel cost benefits the Tesla can compete in the luxury market, thanks largely to the power accompanying battery packs of 60 kwh and larger (see 1). The Value of Long-distance Driving, Given Heterogeneity of Distances for Each Vehicle Daily driving range greater than the range obtained by an overnight charge can be accomplished by (1) public charging infrastructure or (2) gasoline HEV mode capability in a PHEV, EREV, or REX/BEVx. We consider benefits of the latter strategy to obtain national coverage for electric drive. Our analysis of all travel in the daily travel sample of the NHTS implies that the ability of a vehicle to travel more than 70 miles per day is quite important. When long distance home-to-other, and other-to-home travel is included, the estimated share of daily travel between 70 and 1000 miles is 42% of the total, arising from operation of only 12% of vehicles. In Table 3 and Fig. 1, the numbers are lower (26% and 7%), since these are based only on home-to-home travel less than 200 miles. When all national miles are considered, long range capability of BEVx vehicles will be quite valuable if the greatest share of national miles is to be electrified. When one averages travel for individual vehicles over several months of time, the impression of need for long-distance travel incorrectly diminishes (Fig. 1 (23)). With respect to the need for range, annual vehicle averages are particularly deceiving. In the Commute Atlanta study (10) only four of 91 sample vehicles averaged more than 70 miles per day. Three of the four were used intensively for commuting. For suburban consumers traveling more than 1.5 hours per day, the time costs for a second fill at a public location could be problematic and lead to a preference for an REX/BEVx. However, if workplace charging were available in an assigned spot, time costs for a second fill would be small, increasing probability of choosing a BEV. However, if full costs of installation of workplace charging (24) were imposed on the owner, the REX might still be preferred. Plug-in Technology Innovation Systems Strategies (Table 3) The shorter the electric range of PHEVs, EREVs, and/or REX/BEVx, the fewer miles can be electrified. Within Table 1, there are four PHEVs for which there is an HEV counterpart available in the same vehicle body. Call HEV/PHEV pairs technology innovation system (TIS) number one (TIS1). There is one 2014 BEV and REX/BEVx pair (BMW i3). Call this TIS2. California regulators keep these systems separate. There are separate credit systems for the two. TIS1 was allowed all-electric operations credit up to 50 miles of range. TIS2 was allowed credit only after achieving 75 or more miles of all electric range. This intentional split was made so that the BEVx strategy would support development of pure electric vehicles.

9 9 Regardless of what California did, only eight of fifty states (12) are presently aggressively planning to follow California regulations. Could the rest of the country benefit from either PHEVs or BEVx vehicles with range between 50 and 75 miles? The share electrifiable columns of Table 3 presume no public charging infrastructure and one overnight residential charge per day. Table 3 estimates (EV + REX column) imply small gains by opening up credits between 50 and 75 miles, but this comparison ignores potential vehicle cost savings. For the BEVx configuration, Table 4 implies that 50+ miles of electric drive range would allow many of those driving more than average daily miles to drive all-electrically consistently. If such a BEVx proved cost competitive with PHEVs with less all-electric range, then, via greater fuel savings and enhanced sales, the proportion of EV miles vs. HEV miles would rise. Consider the relative benefits of the existence of the BEVx option within TIS2. In the first case, assume only TIS1. This is represented by the Table 3 column labeled PHEV or REX only. Due to a need for financial return, it is assumed that PHEV/EREV/REX/BEVx vehicles will only be sold to those who travel over 100% of all-electric range. Electric miles replace gasoline miles during the first part of the day s travel, afterwards gasoline is used. The maximum benefit is estimated as if every customer with a long enough daily distance pattern owned a PHEV/EREV/REX/BEVx with range equal to the top end of the. Presumed cost pressures of larger packs, as well as packaging (space) and HEV mode efficiency issues are assumed to cause consumers with range needs less than available to choose another vehicle. Under these assumptions, the potential of the PHEV vehicle class is maximized with a range of about 30 miles, which is consistent with prior financial analysis of the best range for the technology (8). The TIS2 offers an ability to serve more customers. For those who do not need long range, the BEV version, which is less costly than the BEVx (see i3 costs in Table 1) is assumed to be chosen by consumers able to drive 50% - 100% of rated range. The range of the BEV evaluated is equal to the top of the distance. Estimated potential electric mileage share ( EV only column) is maximum at 35% for BEVs with range from miles (the range of most present BEVs, see Table 1). For consumers that drive further than the range of the BEV, the BEVx option with a HEV mode and gasoline tank large enough to allow 200 miles of range or more is assumed to be chosen. Results are the same as for the PHEV/EREV/REX/BEVx in the EV only. The TIS2 system, having the option for both BEV and BEVx configurations, has more hypothetical electrified miles potential than the HEV/PHEV TIS1 combination. The sum of potentials is shown in the EV + REX column. The sum peaks at about miles. Thus, the TIS2 design result for the BMW REX, allowing for a reserve to deal with cold temperature and perhaps battery pack deterioration, looks approximately correct, though conservative (i.e. with an EV range cushion ). As public charging infrastructure is installed, allowing multiple charges per day, the share of BEVs in TIS2 would rise. The EV share of EV/REX column implies that at ~ miles the ultimate share of BEVs within TIS2 would be about two thirds. Nevertheless, the computations do not consider the number or value of occasional long inter-city trips beyond 200 miles. Proper consideration of this consumer preference could well increase predictions of REX share. Vehicle design to assure this capability (larger gasoline tanks in particular) satisfactory to more consumers may be critical. Finally, consider the pure EV solution (TIS3). Suppose that regulators chose to have a philosophical opposition to the BEVx alternative and demand that only all-electric vehicles of 150 miles of range or more would be acceptable (REX total range), in order to assure all-electric operation. In this case, given the ad hoc cost-related marketability assumptions, less plug-in vehicles would be sold and less miles would be electrifiable (EV only column, 26%) than for either best case in TIS1 (30%) or TIS2

10 10 (55%). Further, the hypothetical regulatory demand would have the market shrinking effect of pushing up vehicle cost, possibly invalidating the assumption that the vehicles could be sold to customers driving only 50% of vehicle range. This thought experiment implies that the detailed study of alternative strategies within the TIS2 system (BEV/BEVx combinations) is desirable. Relative to TIS3 (Super EVs), the relative cost drops sharply (Table 1), while potential to electrify miles rises sharply. A caveat is that the BMW i3 has four seats, while the Tesla Model S has five plus two. Smaller children can fit in the two rear facing seats, but luggage space is sacrificed. The TIS3 Tesla is a large car, the TIS2 i3 a subcompact car. Relative to TIS1 the promise of electrifying more miles is encouraging, particularly if future BEVx cost increments relative to PHEVs can be held below those in retail price terms in Table 1. The investigation of potential target markets taking into consideration driving speed, days of use and hours per day of use implied that the TIS2 system competes best with conventional gasoline vehicles when implemented in new houses at the outskirts of metro areas by owners with long commutes. It is ironic that the technology sought for reduction of urban emissions and for reduction of fossil fuel use (GHGs, see Table 1) might in the short-run contribute to urban sprawl by making travel from distant suburbs more financially viable. Nevertheless, the long-term must be considered. As battery technology improves and if petroleum fuels, though abundant, become even more costly to produce more widespread application of the technology will ensue. In closing, observe that each of the three TIS options serves different market niches. The implication of this investigation is that they are generally complementary rather than directly competitive. The potential for success in all three niches may be enhanced by reducing battery costs via a large, flexible battery manufacturing facility as planned by Tesla (25). CONCLUSIONS Lithium ion batteries have created the possibility for many combinations and permutations in powertrain design. Given the possibilities California regulators provide credits for two distinct plug-in technology paths: (1) PHEVs spun-off from HEVs and (2) REX/BEVx vehicles spun off from BEVs. California s regulators kept the categories separate because their ultimate goal is an affordable long-range BEV to assure zero emissions. The market has provided a third option, a costly high performance long range Super EV competitive with imported luxury performance vehicles. The thought-experiment marketsegments analysis done here suggests that the second option is most promising for increasing electrified miles. However, each technology path addresses a different market niche. The sum is greater than the parts by establishing complementary scale in electric drive manufacturing. Within the second option, the BEVx range extension feature significantly enhances the value of the package. It is argued to be particularly effective due to a high proportion of total national miles (42%) being caused by a relatively few vehicles (12%) driving beyond the range of most currently available BEVs (> 70 miles). Nationwide automaker implementation of the BEVx HEV mode technological constraints imposed by California regulators may shackle that technology option in most of the nation. BEVx vehicles with long gasoline range and good HEV mode performance could go anywhere that gasoline infrastructure existed. By enabling such vehicles to make trips where BEVs would be infeasible, greater BEVx sales would result, additional home base ZEV operation would be facilitated in many areas that could use it, and national GHG and oil savings benefits would be increased.

11 11 ACKNOWLEDGEMENT This work was supported by the Vehicle Technology Office of the Office of Energy Efficiency and Renewable Energy of the United State Department of Energy. REFERENCES 1. Nelson, P.A., et al Modeling the Performance and Cost of Lithium-Ion Batteries for Electric Drive Vehicles, Argonne National Laboratory report ANL-11/32, September. 2. Santini, D.J., and A.J. Burnham. Reducing Light Duty Vehicle Fuel Consumption and Greenhouse Gas Emissions: The Combined Potential of Hybrid Technology and Behavioral Adaptation. SAE Morris, C. The Launch of Many Firsts. Charged Electric Vehicle Magazine. No. 9, Aug pp ChargedEVs.com The official government source for fuel economy information. Compare sude-by-side. U.S. Department of Energy EERE and U.S. Environmental Protection Agency OTAQ The official government source for fuel economy information. Model Year 2014 Fuel Economy Guide. U.S. Department of Energy EERE and U.S. Environmental Protection Agency OTAQ. 6. California Air Resources Board. Final Statement of Reasons for Rulemaking Including Summary of Comments and Agency Responses 2012 Amendments to the Zero Emission Vehicle Regulations. Jan , Santini, D. J. Plug-in Hybrid Electric Vehicles. IEA-HEV 2014 Task 15 Report, Feb Santini, D.J., et al. Cost Effective Annual Use and Charging Frequency for Four Different Plug-in Powertrains. SAE SAE World Congress, Detroit, MI, April 16 18, Elango, V., R. Guensler and J. Ogle. Day-to-day travel variability in the Commute Atlanta, Georgia, study. Transportation Research Record, No , No. -1, 2007, pp D. J. Santini, et al. Daytime Charging What is the Hierarchy of Opportunities and Customer Needs? A Case Study Based on Atlanta Commute Data. Transportation Research Board 93rd Annual Meeting. No Federal Highway Administration, 2009 National Household Travel Survey, U.S. Department of Transportation, Washington, D.C California Air Resources Board. Governors Announce Bold Initiative to Put 3.3 Million Zero- Emission Vehicles on the Road by News Release. Oct. 24, Sears, J. and K. Glitman. Transportation Technical Reference Manual: Guide to Characterize the Savings, Benefits, and Costs of Transportation Efficiency Measures. June National Association of State Energy Officials, Arlington, VA Joseck, F. and J. Ward. Cradle to Grave Lifecycle Analysis of Vehicle and Fuel Pathways. U.S. Department of Energy Hydrogen and Fuel Cells Program Record. March 24, The EV Project. How many electric miles do Nissan Leafs and Chevrolet Volts in the EV Project Travel? May

12 Santini, D.J., et al. (2013). Plug-in Electric Cars for Work Travel: Evaluation of Four Electric Powertrains. Transportation Research Record: Alternative Fuels and Technologies. 2385: 53 60, McCraw, J. Green Oval. Electric and Hybrid Vehicle Technology International. July Pp Rask, E.M., H. Lohse-Busch, and D.J. Santini (2013). Analysis of Input Power, Energy Availability, and Efficiency during Deceleration for X-EV Vehicles. SAE SAE Int. J. Alt. Power. 2(2): , Santini, D. J., Vyas, A. D., Saucedo, D. and Jungers, B. Where Are the Market Niches for Electric Drive Passenger Cars? Paper , presented at the 90 th Annual Meeting of the Transportation Research Board. Washington D.C., Jan Zhou, Y., A. Vyas, and D. Santini. Tracking National Household Vehicle Usage by Type, Age, and Area in Support of Market Assessments for Plug-in Hybrid Electric Vehicles. Journal of Transportation Technologies, Union of Concerned Scientists. New Survey Finds Many Households Can Use Electric Vehicles. Dec. 11, 2013http:// 22. Argonne, Light Duty Electric Drive Vehicles Monthly Sales Updates, Argonne National Laboratory, Argonne, IL. Accessed on July 27, Lin, Z. et al. PHEV Energy Use Estimation: Validating the Gamma Distribution for 3 Representing the Random Daily Driving Distance. Paper , presented at the 91 st Annual Meeting of the Transportation Research Board. Washington D.C., Jan Smart, J. PEV Infrastructure Deployment Costs and Drivers Charging Preferences in the EV Project. Presented at: SAE 2014 Hybrid and Electric Vehicle Technologies Symposium February 11, 2014 La Jolla, CA Wile, R.. Tesla Signs Deal With Panasonic On Gigafactory. Business Insider. July 28,

13 TABLES 13 Table 1. Selected technical attributes of 2014 plug-in vehicles via fueleconomy.gov

14 14 Table 2 Vehicle use data by vehicle type for Home-to-Home daily travel and other travel (2009 NHTS) Vehicle Type Day s Travel Category Miles per Day Driven Hours per Day Driven Nonwork mph Overall mph Work trip mph Miles per work trip Work trips per vehicle Car Home to Home % Car Home to Non-home % Car Non-home to Home % Share of miles Van Home to Home % Van Home to Non-home % Van Non-home to Home % SUV Home to Home % SUV Home to Non-home % SUV Non-home to Home % Pickup Home to Home % Pickup Home to Non-home % Pickup Non-home to Home %

15 15 Table 3 Max. share of miles electrifiable under marketability assumptions (home-to-home travel c ) Share of vehicles vs. share of miles (home-to-home d ) Distance (miles) Total Miles (10^7) Share vehicles < or = max. Share miles < or = max. Share vehicles > max. Share miles > max. Max. share electrifiable EV only a PHEV or REX EV + only b REX EV share of EV/REX HEV share of PHEV or REX % 4% 77% 96% 2 (nf) 20% (nf) 20% na 79% % 14% 54% 86% 10 (nf) 28% 28% 0% 67% % 27% 38% 73% 18 (nf) 30% 30% 0% 59% % 39% 27% 61% 25% 28% 53% 46% 53% % 50% 19% 50% 29% 25% 55% 54% 49% % 60% 14% 40% 33% 21% 54% 61% 47% % 68% 10% 32% 35% 18% 53% 66% 44% % 74% 7% 26% 35% 15% 50% 69% 41% % 79% 5% 21% 35% 13% 48% 73% 38% % 83% 4% 17% 34% 11% 44% 76% 35% % 87% 3% 13% 32% 9% 41% 79% 33% % 90% 2% 10% 30% 7% 37% 80% 29% % 92% 2% 8% 28% 6% 34% 83% 27% % 94% 1% 6% 26% 4% 31% 86% 25% % 96% 1% 4% 25% 3% 28% 88% 23% % 96% 1% 4% 17% 0% 17% 100% na nf = not technically feasible due to inadequate battery kw na = not applicable a Sales only to customers driving from % of range b Sales only to customers driving > 100% of EV only range c Maximum of one charge per day of use dneeds for inter-city capability ( miles) ignored

16 16 Table 4 Sample vehicle driving statistics as a function of home base population density (2009 NHTS) Vehicle Type Population density Miles per Vehicle Hours per vehicle Nonwork mph Overall mph Work trip mph Share of miles in density Share of vehicles in density Share of vehicle type by density Car <= % 47% 35% Van <= % 9% 40% SUV <= % 22% 41% Pickup <= % 21% 58% Car % 57% 35% Van % 9% 34% SUV % 23% 35% Pickup % 11% 25% Car % 61% 28% Van % 9% 24% SUV % 21% 23% Pickup % 9% 16% Car > % 64% 1% Van > % 9% 1% SUV > % 24% 1% Pickup > % 4% 0%

17 $ Saved per 1.5 Hour Day of Use $ Saved per 1.5 hr Day of Use $ Saved per 1.5 Hour Day of Use $ Saved per 1.5 Hour Day of Use Beyond BEV Range Santini and Zhou 17 FIGURES Figure 1. Daily driving distance patterns arising from one day vs. multi-month averaging (Estimated gasoline share is calculated assuming driving a Camry 3.5L gasoline car) $8.00 $7.00 $6.00 $5.00 $4.00 $3.00 $2.00 $1.00 Beyond BEV Range $ ($1.00) Averge Miles per Hour Chevrolet Cruze, Volt and Spark Spark EV Volt 1.5 charge Volt 1 charge Cruze Diesel Cruze ECO $8.00 $7.00 $6.00 $5.00 $4.00 $3.00 $2.00 $1.00 Electricity $0.10/kWh Fuel price $3.60/gal 1 or 1.5 charges/day $ ($1.00) Average Miles Per Hour Toyota - Camry HEV, Prius and RAV4 EV vs. Camry 2.5 RAV4 BEV Prius PHEV 1.5 charge Prius PHEV 1 charge Prius HEV Camry HEV $8.00 $7.00 $6.00 $5.00 $4.00 $3.00 $2.00 $1.00 $0.00 Electricity $0.10/kWh Fuel price $3.60/gal 1 or 1.5 charges/day Beyond BEV Range ($1.00) Average Miles per Hour Ford Fusion (& Focus) Options (vs. 1.5 Turbo) Focus EV Energi PHEV 1.5 charge Energi PHEV 1 charge Full HEV Stop/start 1.5 Turbo $8.00 $7.00 $6.00 $5.00 $4.00 $3.00 $2.00 $1.00 $0.00 Beyond BEV Range ($1.00) Average Miles per Hour BMW 3 Series Options (vs. 328i) i3 BEV i3 REX 3 Diesel Active E 3 (HEV) Figure 2 Daily Fuel $ Savings: Chevrolet (Volt, Spark, Cruze), Toyota (Prius, RAV4, Camry), Ford Fusion (& Focus) and BMW 3 Electric Drive (& Diesel) Alternatives