Strategic Issues Facing Transportation, Volume 5: Preparing State Transportation Agencies for an Uncertain Energy Future (2014)

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

161 Hydrogen has been used for various commercial applica- tions for over a century, but serious interest in hydrogen as a transportation fuel began with the first hydrogen fuel-cell bus demonstration by Ballard Power Systems in 1993 (OFE 2000). Soon thereafter, many automakers began to launch their own hydrogen-fueled vehicle research and development programs. While hydrogen can be used in internal combustion engines, either as pure hydrogen (Cho 2004) or combined with natural gas (Burke et al. 2005), most auto automakers have focused on hydrogen fuel-cell vehicle technology, which will likewise be emphasized in this appendix. Over the longer term, the use of hydrogen as a transportation fuel could offer several compelling advantages: • Reduced petroleum use in transportation. Hydrogen can be generated through several processes using a variety of feedstocks, many of which can be produced domestically. A large-scale shift to hydrogen could drastically reduce the use of petroleum in passenger transportation. • Greenhouse gas mitigation. Some of the potential feed- stocks for hydrogen, such as biomass or low-carbon elec- tricity and water, would allow for very low well-to-wheel carbon emissions. And when hydrogen is processed within a fuel cell to produce electricity to power a vehicle, the only tailpipe by-product is water. Depending on how it is pro- duced, then, hydrogen could support a dramatic reduction in GHG emissions from the transportation sector. • Improved air quality. Vehicles powered by hydrogen fuel cells emit no harmful air pollutants from tailpipes and could thus improve air quality within urban areas by reducing mobile-source emissions. Depending on the processes and feedstocks used to generate hydrogen, there could also be reductions in harmful air pollutants at the point of fuel production. • High energy efficiency. In comparison to the use of gaso- line or diesel within an internal combustion engine, fuel cells are able to use the embodied energy within hydrogen much more efficiently. If the cost of producing and delivering hydrogen can be reduced such that it is roughly the same as the cost of gasoline or diesel, and if the cost of fuel-cell vehicles can likewise be reduced, then FCVs could offer significantly lower total per-mile energy costs of travel. Some transit agencies have begun to explore the use of hydrogen fuel-cell buses, with a total of 25 fuel-cell buses being demonstrated in eight locations as of 2012 (Eudy, Chandler, and Gikakis 2012). However, light-duty vehicles are expected to offer a much larger market for hydrogen fuel. This appendix will focus on light-duty FCVs. Potential hydrogen applica- tions in the medium- and heavy-duty fleets are considered in Appendix F. E.1 Production, Distribution, and Refueling While a significant quantity of hydrogen is already pro- duced in the United States each year, most of this is used for applications other than transportation. To support a major shift to hydrogen as a transportation fuel, the total volume of hydrogen production would need to be expanded dramatically. It would also be necessary to develop the infrastructure for distributing hydrogen and refueling vehicles, requiring sig- nificant investment. E.1.1 Current Hydrogen Generation Hydrogen, like electricity, is an energy carrier. It does not occur naturally on earth but rather must be produced from other sources such as natural gas, coal, wood, water, or bio- mass. As of 2011, hydrogen production capacity within the United States exceeded 9 million metric tons per year. Roughly two-thirds of this amount is captive hydrogen—that is, hydro- gen that is used where it is produced—for oil refineries and for making ammonia and methanol. The remaining third is A p p e n d i x e Hydrogen Fuel-Cell Vehicles

162 merchant hydrogen, used mostly in off-site refineries and dis- tributed by pipeline or liquid tankers (Joseck 2012). To put current U.S. hydrogen production in perspective, a kilogram of hydrogen is roughly equivalent, in terms of embodied energy, to a gallon of gasoline, though fuel-cell vehicles achieve much higher fuel economy on an energy- equivalent basis. The total production of 9 million metric tons of hydrogen per year translates to about 25 million kilo- grams of hydrogen per day, capable of powering about 36 to 41 million FCVs (Joseck 2012). The total number of light-duty vehicles in the U.S. fleet, however, is approaching 250 million (ORNL 2012). Even assuming that all of the hydrogen currently produced in the United States could be redirected to trans- portation, the total volume would still need to be increased by almost an order of magnitude in order to replace current gasoline consumption. E.1.2 Hydrogen Production Pathways There are several feedstocks and processes that could be used to produce additional hydrogen for use as a transportation fuel. The most mature production process is thermochemical hydrocarbon conversion, encompassing such methods as steam methane reformation and gasification. Potential feedstocks include coal, oil, natural gas, and biomass. Steam reforming of natural gas, for example, is the most common method of hydro- gen production today, used mainly for industrial and refining applications. Fossil fuels currently provide the cheapest option for producing hydrogen, but they also result in the emission of greenhouse gases as a by-product. Over the longer term, the resulting carbon emissions could be captured at the source and subsequently sequestered deep in the earth to mitigate climate effects, although the technical and financial feasibility of this concept at scale remains to be demonstrated. If CCS does not prove successful, the use of biomass as a feedstock, which draws carbon dioxide from the atmosphere as it grows, could serve as a low-carbon source of hydrogen (Ogden et al. 2011). Electrolysis, in which electricity is used to split water into oxygen and hydrogen gas, is another option for producing hydrogen fuel. In terms of greenhouse gas emissions, the ben- efits of electrolysis depend on the feedstocks used to produce the electricity. With electricity generated from coal (or from natural gas, to a lesser extent) without carbon capture and sequestration, the production of hydrogen would still result in significant GHG emissions. If, on the other hand, the elec- tricity is generated from low-carbon sources such as nuclear, solar, wind, hydropower, or thermal power plants with carbon capture and sequestration, then the production and use of hydrogen as a fuel would have very low well-to-wheel GHGs. Unfortunately, the production of hydrogen fuel from water using clean or renewable electricity is expected to remain more costly than hydrogen derived from fossil fuels or biomass over the next decade or more. Looking out even further, clean hydrogen could be generated via thermochemical water split- ting with high-temperature heat from nuclear or concentrated solar power, via biological or photoelectrical water splitting, or via biomass pyrolysis, although these technologies are more economically challenging (Ogden et al. 2011, NPC 2012). In examining alternate hydrogen production pathways, another consideration is whether the fuel is generated at a large central plant and then transported (e.g., via truck or pipeline) to refueling stations or is instead produced locally, at much lower volume, near the point of sale. Because mature distribution networks for natural gas, water, and electric power are already in place, steam reformation and electrolysis are both amenable to distributed generation. The per-unit cost of smaller-scale production is generally higher than for large-scale production at central plants. Absent a national hydrogen distribution network, however, and depending on the location of a refuel- ing station and the volume of hydrogen required, distributed hydrogen production could in some cases be less expensive than producing hydrogen centrally and then distributing via truck. Table E.1, based on a recent NRC study, provides a sense of the potential trade-offs (NRC 2013b). The table shows the Table E.1. Future commercial-scale hydrogen production costs and emissions. Scale Feedstock Process Low Carbon Cost ($2009/gge) Centralized Natural gas Steam methane reformation No 1.40 Coal Gasification No 1.80 Coal Gasification with CCS Yes 2.50 Biomass Gasification Yes 2.10 Water/wind Wind power electrolysis Yes 3.82 Water/nuclear Thermochemical splitting Yes 1.39 Distributed Natural gas Steam methane reformation No 1.60 Water/grid Grid power electrolysis No 3.80 Source: NRC (2013b); costs in 2020 assuming 10 million FCVs on the road.

163 estimated costs (in 2009 dollars per gge) and emissions char- acteristics of several centralized and distributed hydrogen production processes and feedstocks in the 2020 time frame. Note that the estimates assume a very rapid rate of FCV adoption, with 10 million FCVs on the road by 2020, allowing for greater economies of scale in production costs. In other words, the table can be viewed as demonstrating what is pos- sible on the fuel production side, as opposed to what is likely; the cost estimates would certainly be higher at lower levels of FCV adoption. Additionally, the estimated prices in the table focus solely on production and do not include transport costs for the centralized production methods, or station costs. As indicated in the table, options involving distributed generation, electrolysis, or lower-carbon emissions generally entail higher production costs. The study from which these estimates were drawn assumed that an additional cost of $2 per gge would be required for distribution and station costs for the centralized production methods, and that an additional cost of $1.88 per gge for station costs would be required for the distributed production methods. Note that the recent significant decline in natural gas prices has contributed to the lower cost estimates for steam methane reformation in the table. E.1.3 Distribution For hydrogen that is produced on site at fueling stations, distribution is obviously not needed. As indicated in Table E.1, however, the per-unit cost of producing hydrogen in larger volume can be lower than distributed generation. As a result, it may be cheaper to produce hydrogen centrally and then transport it to a refueling station. The most cost-effective means of distributing hydrogen to a fueling station depends on volume of use. Three scales are commonly considered (Washington State DOT 2009): • Gaseous hydrogen tube trailers. At small scales (roughly 20 to 60 refills per day), gaseous hydrogen can be delivered by trucks carrying tube trailers that store compressed gas. • Liquid hydrogen tank trailers. At larger scales (200 to 500 refills per day), liquid hydrogen can be delivered by trucks with cryogenic tanks (cooled to less than minus 253 degrees Celsius). The hydrogen must be liquefied at the production facility, transferred to the tanks, and finally transferred to storage tanks at the fueling station. There is a cost and energy penalty in liquefying hydrogen, but the higher-volume capacity makes this a better option than tube trailers with greater demand. • Gaseous hydrogen pipelines. Finally, gaseous hydrogen can be delivered by pipelines in the same manner as natural gas. However, pipelines for hydrogen are quite expensive, requiring alloy steel to avoid hydrogen embrittlement of low-carbon steel and special seals due to the small molecule size. Estimated costs range from $1 million to $1.5 million per mile in urban settings. The delivery of hydrogen via pipeline might therefore become cost-effective only when FCVs have achieved significant market share, on the order of 25% or more of the on-road fleet (Washington State DOT 2009). E.1.4 Hydrogen Fueling Stations Should hydrogen emerge as a successful transportation fuel, it is expected that most vehicles will refuel at stations similar to current gasoline filling stations. Building out a network of hydrogen refueling stations will require significant investment. As of early 2012, there were fewer than 80 hydrogen stations in the United States, with a significant share in California, and more than 20 additional stations in the planning stages (Fuel Cells 2000, 2012b). Most of these are for demonstra- tion programs, however, and are not intended for public use. Existing stations tend to be small, and a number have been decommissioned. To provide a sense of the scale of the investment that could be required to develop a network of hydrogen fueling stations, the API reports that there are more than 150,000 locations across the country that sell gasoline, including service stations, truck stops, convenience stores, and marinas (API 2013). To support early-stage market penetration for FCVs, one study estimated that the total number of hydrogen fueling stations needed would be about 3% to 7% of the current number of gasoline stations (Nicholas and Ogden 2006). This roughly translates to between 4,500 and 11,000 stations, up to two orders of magnitude above the existing number of stations. If hydrogen fuel-cell technology were to become so success- ful that FCVs eventually replaced fossil-fueled internal com- bustion engines within the light-duty fleet, then the number of hydrogen fuel stations would continue to expand, ultimately approaching the current number of gasoline stations. Yet stations have high capital requirements, with cost estimates falling in the range of $1 million to $7 million, depending on size and configuration (NPC 2012). Hydrogen stations are costly, in large part because they are more complex than existing fueling stations. Gasoline is stored as a liquid at ambient temperature both at the station and in the vehicle. As a result, the two main components of a gasoline fueling station are a storage tank and a dispenser. In contrast, hydrogen is likely to be produced or stored in different forms or pressures at the station and on the vehicle. Stations might store hydrogen as a cryogenic liquid, store it as a gas at mod- erate pressure (less than 2,500 psi), produce it on-site at low to moderate pressures, or receive it from a pipeline at low pressures (hundreds of psi). In contrast, most current FCV prototypes are equipped with tanks that store hydrogen at 5,000 or 10,000 psi. (While other forms of onboard hydrogen

164 and $10 to $13 per kilogram for on-site electrolysis. Another study (NRC 2013b) projected the cost of hydrogen fuel at roughly $10 per kilogram in the near term. Along similar lines, the NPC (2012) estimated the near-term cost for centrally produced and delivered hydrogen to fall in the range of $9 to $12 per kilogram and the cost for distributed hydrogen to fall in the range of $15 to $25 per kilogram. In a high market penetration scenario considered in the NRC study, infrastructure begins to scale up after 2015, and hydrogen costs decline to about $4 per kilogram by 2050 (NRC 2013b). Ultimately, though, the cost of hydrogen fuel will depend on such factors as aggregate demand, advances in technology, type of production and feedstock, production scale, and mode of transport for any centrally produced hydrogen. E.2 Vehicle Technology Auto manufacturers initiated fuel-cell vehicle develop- ment programs in the late 1990s. The effort to commercialize hydrogen FCVs has received considerable public and private support and interest in the intervening years, and California has played an important role in this history. The formation of the California Fuel Cell Partnership (CFCP) in 1999—a col- laboration involving public agencies, auto manufacturers, fuel providers, and other interested parties—marked an important early milestone. The CFCP works to demonstrate and promote the potential for fuel-cell vehicles and has engaged in many activities aimed at bringing FCVs closer to market. Members conduct outreach to the public, fleet owners, and public offi- cials, and they also train first responders on hydrogen-related issues. Above all, the CFCP seeks to stimulate collaboration among public and private groups working toward the develop- ment and adoption of fuel cells and hydrogen (CFCP, undated). The California Air Resources Board’s ZEV program has also been influential in the development of FCVs (CARB 2013). Launched in the 1990s and subject to several revisions in the years that followed, the main aim of the program has been to require auto manufacturers to produce specified quantities of ZEVs for sale in California. As part of the program, CARB has mandated the demonstration of limited numbers of FCVs. In the most recent ZEV revisions, auto manufacturers have the opportunity to meet credit requirements through some combination of FCVs along with battery electric vehicles. This allows automakers to choose which of the technologies best meet their strategic plans. At the federal level, the U.S. Department of Energy funds fuel-cell research and development through the U.S. DRIVE Partnership (previously known as FreedomCAR and the Partnership for a New Generation of Vehicles), which has existed since 1993 (NRC 2013a). The U.S. Department of Energy’s Alternative Fuel Data Center resource provides a comprehensive listing of additional federal and state programs storage, such as cryogenic tanks for liquid hydrogen, have been explored, it appears that FCVs will rely on high-pressure gas- eous storage for at least the near future.) As such, hydrogen stations are likely to require a compressor in addition to a storage tank and dispenser, and may require a vaporizer as well if the hydrogen is stored cryogenically at the station. The high cost of hydrogen fueling stations poses a chicken- and-egg type of problem for the successful rollout of hydrogen fueling stations. Specifically, absent a sufficient number of FCV owners, fuel providers will be reluctant to build hydrogen fueling stations given the considerable cost involved. At the same time, prospective FCV buyers will hesitate to make the transition to hydrogen unless they know that it will be con- venient to access fueling stations. Ideally a small number of stations could be built before the initial rollout, with additional stations to be added as FCV ownership increases. In practice, though, this may prove difficult to coordinate. One concept for overcoming the challenge of providing hydrogen fuel to early FCV adopters is the mobile refueler (CEC 2004). A mobile refueler consists of a trailer that contains all the components necessary to fuel vehicles—storage tank, compressor, and dispenser—and can be moved from one place to another. For example, a mobile refueler might serve one neighborhood on Mondays and Thursdays, another on Tuesdays and Fridays, and so forth. Although mobile refuelers would be less than ideal in terms of convenience, they might still play a valuable role in supporting initial adoption until a broader network of permanent hydrogen fueling stations emerges. As with electric and natural gas vehicles, home refueling with hydrogen is a realistic possibility at some point. Options might include a tri-generation appliance that uses natural gas to produce some combination of heat, electric power, and hydrogen fuel, or a home electrolysis device. Over the foresee- able future, however, such technology is likely to be priced out of the reach of most households, necessitating the consider- ation of publicly accessible hydrogen refueling infrastructure to enable a large-scale shift to hydrogen as a transportation fuel (Ogden et al. 2011). E.1.5 Current and Future Hydrogen Costs The present cost of hydrogen as a transportation fuel for the end user would be high given the nascent state of the market. While a few hydrogen refueling stations exist, most of these are for demonstration projects and are oversized relative to demand. This has the effect of inflating the apparent cost by allocating capital investments over a smaller quantity of dis- pensed fuel. Drawing on data from existing stations, however, Wipke et al. (2010b) estimated that early hydrogen costs for a station operating at higher volume would fall in the range of $8 to $10 per kilogram for on-site natural gas reformation

165 target cost set by DOE is $30 per kW, which developers hope to achieve via improved materials, reduced use of platinum, and increased power density (Ogden et al. 2011). At $30 per kW, fuel cells would still be more expensive than ICEs, but the differential would be more modest. Onboard hydrogen storage technology. The other cur- rent challenge for hydrogen vehicles is determining the most cost-effective way to store enough hydrogen on board to accommodate a 300-mile range. The main options include high-pressure cylinder tanks, super-cooled liquid hydrogen, or the use of metal hydrides or other special materials able to absorb hydrogen under pressure. Absent breakthroughs in the latter two approaches, most manufacturers to date have opted for compressed hydrogen in their demonstration vehicles, and this is expected to be the technology of choice for at least the near term. Demonstration FCVs produced by Daimler Chrysler, GM, Honda, Hyundai, and Toyota have been able to achieve ranges of 270 to 400 miles using tanks with compressed hydrogen stored at 5,000 to 10,000 psi (Ogden et al. 2011). Estimates for the cost of mass-produced compression tanks based on current technology fall in the range of $12 to $19 per kilowatt hour. This translates to about $2,800 for a tank able to hold enough hydrogen to provide a compact FCV with a range of 300 miles (NRC 2013a). E.3 Cost and Performance Demonstration FCVs offer generally strong performance, acceptable range, strong fuel economy, and—depending on how the hydrogen is produced—the potential for significant reductions in greenhouse gas and criteria pollutant emissions. The main limitation, other than lack of refueling stations, is high vehicle premiums. E.3.1 Vehicle Cost The cost of fuel-cell vehicles at present is hard to determine, given that fuel-cell vehicles are not widely available and so far only small numbers have been produced. Therefore, most studies have taken the approach of projecting future vehicle prices based on assumptions about the rate of market adop- tion and the reduction of costs with greater economies of scale and learning. On commercial introduction around 2015, one estimate is that FCVs are expected to have a price premium of about 1.4 times that of gasoline ICE vehicles (NPC 2012). Other studies provide more conservative esti- mates. Ogden et al. (2011) present a series of future cost estimates linked to year and cumulative FCV sales: 5,000 FCVs sold by 2014 with a cost of $140,000 per vehicle, over 50,000 sold by 2015 with a cost of $75,000, over 300,000 sold by 2017 with a cost of $50,000, and two million sold by 2020 with a related to hydrogen (EERE 2013). While many involve cash or tax incentives to promote adoption of hydrogen fuel and FCVs, regulatory mandates such as the ZEV program have also been employed. Without public subsidies and regulations, it is unclear that FCVs would have progressed as far as they have. Indeed, the costs for both vehicles and fueling infrastructure are prohibitive. Over the past 10 years, though, most automakers have demonstrated fuel-cell vehicles. The organization Fuel Cells 2000 has compiled a list of such vehicles around the world (Fuel Cells 2000, 2012a). Some of the models referenced are prototype or concept vehicles, but many are based on commer- cial platforms. Honda, Toyota, Daimler, GM, and Hyundai have all announced plans to commercialize FCVs by around 2015 (NRC 2013b). To achieve successful market penetration, continued technological advances and cost reductions for fuel-cell vehicle technologies will be crucial. The two main vehicle development challenges involve the fuel cell and onboard hydrogen storage. Other key components of hydro- gen fuel-cell vehicles are electric motors, power controllers, and batteries for hybrid operation and cold-start support. The following discussion draws upon an overview from Ogden et al. (2011) on the challenges and prospects for FCV technology. Fuel-cell technology. Fuel cells can be thought of as highly efficient electrochemical engines that combine hydrogen and oxygen to produce electricity, which is then used to power an electric motor to propel the vehicle. There is no combus- tion, however, or emissions of greenhouse gases or local air pollutants from the vehicle tailpipe. Rather, water is the sole by-product of FCV operation (Ogden et al. 2011, NRC 2013b). One of the most important fuel-cell technologies for trans- portation applications is the proton exchange membrane. Vehicle manufacturers have made significant progress with PEM fuel-cell systems—the size and weight, for instance, have been reduced such that a fuel cell now fits easily within a compact vehicle, and FCVs provide strong driving perfor- mance and tolerance for low-temperature operations. There are, though, some remaining challenges—most notably with respect to durability and cost. Current PEM fuel-cell systems have a lifetime of about 2,500 hours in on-road tests, well short of the 5,000-hour life viewed as a target for commer- cialization. PEM fuel-cell durability is continuing to increase, however, and laboratory tests have demonstrated lifetimes of more than 7,000 hours (NRC 2013a). The cost of fuel-cell technology, often measured in dollars per kilowatt, is currently quite high, owing in part to low pro- duction volumes. The U.S. Department of Energy, however, has estimated that if current fuel-cell configurations were produced at greater scale—on the order of 500,000 units per year—the cost would fall to about $49 per kW (NRC 2013a). This would translate to about $4,000 for an 80-kW system, which is roughly twice the cost of a comparable internal combustion engine. The

166 in comparison to conventionally fueled vehicles. Unlike the emissions of greenhouse gases from the combustion of gaso- line or diesel in an ICE, the onboard conversion of hydrogen to electricity releases no GHG emissions. Rather, all of the emissions associated with hydrogen stem from its production and transport—that is, from the well-to-tank portion of the fuel cycle. Thus the magnitude of the GHG emissions benefits depends largely on the method of producing and transporting hydrogen. Using data from the U.S. Department of Energy’s dem- onstration program, researchers at the National Renewable Energy Laboratory compared the well-to-wheels greenhouse gas emissions from present fuel-cell vehicles based on different hydrogen production pathways. As a point of reference, the researchers considered a mid-size petroleum-fueled pas- senger vehicle that was estimated to produce 484 grams of CO2e emissions per mile. In comparison, emissions for an FCV fueled via on-site reformation of natural gas would be reduced by 25% on average (at 362 grams of CO2e per mile), while emissions from an FCV fueled via electrolysis using electricity from the grid would be reduced by 22% on average (at 378 grams of CO2e per mile). The study also noted that fueling an FCV based on electrolysis with renewable electricity would produce no GHG emissions (Wipke et al. 2010c). As part of this study, with the aim of facilitating a con- sistently framed examination of the GHG emissions reduc- tion potential for the different fuels and vehicle technologies considered, the research team conducted an exercise relying on emissions factors from ANL’s GREET model (ANL 2012). For each of the fuels and vehicle technologies—gasoline, diesel, hybrid electric, natural gas, biofuels, plug-in and battery electric, and hydrogen—the team used GREET’s specifications for a mid-size passenger vehicle in 2010 along with two similarly configured vehicles in the 2050 time frame. The two hypothetical future cases are intended to be broadly consistent with expectations in the research literature regard- ing possible advances in fuels and vehicle technologies, but one embeds modest assumptions while the other is more optimistic. Figure E.1 presents the GHG emissions modeling results for conventional vehicles and FCVs. The GREET model for 2010 FCV efficiency and well-to-wheels emissions is used, and for future fuel economy and well-to-wheels emissions, fuel economy assumptions from NRC (2013b) are used, reduced by 17% to adjust for on-road fuel economy. The intent of including current as well as future conventional vehicles is to help clarify not only how FCVs might perform in relation to the current light-duty fleet, but also how they might per- form against conventional vehicles with much-improved fuel economy in the future. Table E.2 lists the salient assumptions for the analysis, including on-road (as opposed to EPA-rated) cost of $30,000. In this same projection, the cost premium of FCVs ultimately levels out in the 2025 time frame to about $3,600 more than the cost of a conventional vehicle. E.3.2 Energy Cost of Travel The energy content in a gallon of gasoline is comparable to that in a kilogram of hydrogen. On an energy-equivalent basis, the fuel economy for today’s FCVs is about twice that of comparable gasoline ICEs, and about 35% to 65% higher than that of a gasoline hybrid electric. In a review by the National Renewable Energy Laboratory of second-generation FCVs participating in the U.S. Department of Energy’s hydrogen demonstration program, the adjusted city/highway fuel economy ranged from 42 to 56.5 miles per kilogram (mpkg) (Wipke et al. 2010a). A study by Sun, Ogden, and Delucchi (2010) estimated fuel economies of 57 mpkg in 2012 and 67.7 mpkg in 2025. The GREET model (ANL 2012) estimated current fuel economy for a mid-sized passenger FCV to be about 52 mpgge (mpgge is roughly the same as mpkg), while a NRC study estimated that the on-road fuel economy for a similar FCV in 2050 could range from around 140 to 170 mpgge (NRC 2013b). As indicated by these estimates, the fuel economy of FCVs is generally expected to exceed that of conventional vehicles. Additionally, as described earlier, analysis in the NRC report suggests that the retail cost of hydrogen fuel could fall to about $4 per kilogram by 2050 assuming significant adoption (NRC 2013b), which would provide the ability to compete with gasoline in the range of $2 to $3 per gallon. Gasoline has been above that range in recent years, and many projections (e.g., EIA 2013) anticipate further increases in gasoline prices in the coming decades. Thus it appears likely that the energy cost of travel for an FCV could be either comparable to or less than, and possibly much less than, the energy cost of travel for a gasoline-fueled vehicle. E.3.3 Operational Performance Beyond the question of fuel-cell durability, current dem- onstration FCVs do not appear to suffer any particular per- formance challenges. Speed and power are acceptable, and vehicle range—at least with compressed hydrogen storage at 10,000 psi—appears to rival that of conventional vehicles. Refueling time is likewise not a problem. Rather, the main near- term challenges relate to vehicle cost and sufficient availability of refueling infrastructure. E.3.4 Greenhouse Gas Emissions FCVs can be expected to offer at least moderate, and poten- tially significant, reductions in greenhouse gas emissions

167 15 grams per mile, owing to the higher vehicle efficiencies and lower GHGs associated with future hydrogen production assumed by NRC (2013b). E.3.5 Local Air Pollutant Emissions As with greenhouse gas emissions, the implications of a shift to FCVs for criteria pollutants—the local air pollutants sub- ject to EPA regulations—depend to some degree on how the hydrogen is produced. Regardless of the production pathway, however, a shift to FCVs should result, on a well-to-wheels basis, in at least moderate and possibly significant reductions in VOCs, CO, NOx, and SOx. Emissions of particulate matter (PM10 and PM2.5), in contrast, appear to be more sensitive to the method of hydrogen production (Sun, Ogden, and Delucchi 2010; ANL 2012). Several production pathways would lead to increased well-to-wheels particulate matter emissions in comparison to gasoline-fueled vehicles. The increase would be most significant if the hydrogen were produced via electrol- ysis using the current U.S. grid mix (Ogden et al. 2011), given the degree to which it relies on coal combustion, although increased substitution of natural gas or other generation for coal would lessen this effect. Additionally, many power plants fuel economy and the assumed carbon intensity, in kilograms of CO2e per gallon of gasoline equivalent (kg CO2e/gge), of hydrogen production and transport. NRC provided several estimates of GHG emissions for dif- ferent methods of hydrogen production and distribution (NRC 2013b). For the 2050 estimates, two GHG values are used from their scenarios. As shown in the table, the first is referred to as a partial use of CCS case, which involves a mix of central natural gas and coal with CCS, central biomass without CCS, and distributed natural gas reforming without CCS. The second, more optimistic, low-GHG case uses more central natural gas with CCS, some central biomass without CCS, electrolysis from a low-GHG grid, and a small amount of distributed natural gas reforming without CCS. Based on these assumed values and GREET’s emissions factors, Figure E.1 graphs the well-to-wheels greenhouse gas emissions perfor- mance, in grams of CO2e per mile of travel, for current and future conventional models and FCVs. Note that the GREET model estimates that current FCV emissions are 272 grams per mile, which is about 40% less than current ICE emissions. This is due to the GHGs associated with making hydrogen from natural gas without CCS. The emissions for the two future FCV cases are much lower, at 37 grams per mile and Source: Computations by authors based on data from ANL (2012) and NRC (2013b). 0 100 200 300 400 500 FCV Gasoline ICE Grams CO2-Equivalent per Mile of Travel (well-to-wheels) 2010 2050 Base 2050 Op mistic Figure E.1. GHG reduction prospects for hydrogen FCVs in 2050. Time Frame Scenario Gasoline ICE (mpg) FCV (mpgge) Hydrogen Production Assumptions Carbon Intensity (kg CO2e/gge) 2010 24.8 52 Central natural gas reforming 14.08 2050 base 72 138 Partial use of CCS 5.1 2050 optimistic 91 171 Low-GHG production mix 2.6 Source: Computations by authors based on data from ANL (2012) and NRC (2013b). Table E.2. Assumptions in emissions comparisons for hydrogen FCVs.

168 If the cost of renewable hydrogen can be reduced to the same level, the motivations for investing in a shift to FCVs will be more compelling. Fuel-cell cost. The cost of fuel cells is much higher than the cost of an internal combustion engine. Most projections, however, assume that the cost differential between fuel cells and ICEs can be pared substantially. Indeed, in some ways fuel cells are simpler and easier to produce than engines. For example, they have no moving parts, and they oper- ate at much lower temperatures. Until low-cost fuel cells can be manufactured, though, the increased vehicle cost will be a barrier to consumers. Fuel-cell lifetime. From a marketability perspective, fuel cells must be able to operate roughly the lifetime of the vehicle; otherwise, owners would be required to replace a very expen- sive component. A decade ago, early versions of fuel cells were not expected to last even a couple of years in normal vehicle operation. While this challenge has yet to be entirely resolved, much progress has been made. The fifth-generation fuel-cell stack from General Motors, for example, is expected to last 120,000 miles (CFCP, undated). Onboard hydrogen storage cost. High-pressure compressed- hydrogen storage tanks, which require expensive materials and processes to manufacture, are another factor in the significant premium associated with current hydrogen vehicles. Research- ers are working on other technologies, such as metal hydrides, to reduce the cost of onboard hydrogen storage for FCVs. Reducing the cost of storing sufficient hydrogen to enable a reasonable driving range (e.g., 300 miles or greater) could play an important role in reducing the overall cost of FCVs and, in turn, enhancing their market prospects. Refueling infrastructure. As described earlier, vehicle manufacturers and energy firms will need to coordinate with one another in rolling out FCVs and refueling stations since a successful transition to hydrogen as a transportation fuel will depend on both. A network of hydrogen fueling stations and supportive distribution infrastructure will require significant investment, so fuel providers are unlikely to develop such a network absent significant uptake of FCVs. At the same time, consumers will be reluctant to purchase FCVs unless refueling infrastructure exists. Therefore, some public-sector involve- ment in subsidizing or mandating early refueling infrastruc- ture would prove extremely helpful and perhaps would even be necessary. Competing fuels and vehicle technologies. To achieve significant market share, FCVs will need to compete well not only with conventional petroleum-fueled vehicles, but also with other alternative-fuel options such as biofuels, natural gas, battery electric vehicles, and plug-in hybrids. Automakers must choose where to focus their funding for research, develop- ment, and marketing. If they come to believe, for example, that battery electric vehicles or plug-in hybrids could lead to are located away from dense settlements, mitigating the human health impacts to some extent. E.4 Market Prospects While FCVs offer considerable promise, the barriers that would need to be overcome to support significant market adoption are also significant. Perhaps unsurprisingly, then, future projections for the uptake of FCVs vary widely. E.4.1 Future Market Projections The U.S. Energy Information Administration’s projections from the most recent Annual Energy Outlook are rather pes- simistic regarding the prospects for FCVs over the next several decades (EIA 2013). In the reference-case scenario, EIA projects that the total number of FCVs in the light-duty fleet will only reach 70,000 by 2040, with 4,700 FCVs projected to be sold that year. Even in the high economic growth scenario, total FCVs are projected at just 80,000 in 2040. Other studies, however, present much more optimistic fore- casts. Greene et al. (2008), for example, in a study for the U.S. Department of Energy, developed three scenarios for future FCV sales drawing upon historical growth in hybrid-electric sales as an analog. Within the three scenarios, the number of light-duty FCV sales in 2025 ranges from 500,000 at the low end to 2.5 million at the high end. CARB has laid out a vision in which both battery electric vehicles and FCVs begin to enter the market in significant numbers by 2020 and together account for virtually all light-duty vehicle sales by 2040 (CARB 2009). E.4.2 Factors Affecting Market Prospects Over the past decade there has been much technical progress in fuel cells and fuel-cell vehicles. Fuel-cell stacks operate for many more hours before they fail, and their power density is much higher. Vehicle range has increased almost four-fold. Vehicles can start and operate in cold temperatures. Neverthe- less, there are several barriers and areas of uncertainty that must be addressed in order for fuel-cell vehicles to achieve significant market penetration. Hydrogen cost and feedstocks. The present cost for deliv- ered hydrogen is too high to compete with gasoline, and the cost for renewably generated hydrogen is higher still. Yet future projections suggest that hydrogen produced from natural gas or coal should ultimately provide for lower per- mile energy costs than gasoline-fueled ICE vehicles. Beyond offering several production pathways supportive of energy independence, however, one of the main attractions for hydro- gen as a transportation fuel is the potential for very low well-to- wheels emissions of greenhouse gases and criteria pollutants.

169 Joseck, F. 2012. “Current U.S. Hydrogen Production.” DOE Hydrogen and Fuel Cells Program Record. Nicholas, M. and J. Ogden. 2006. “Detailed Analysis of Urban Station Siting for the California Hydrogen Highway Network.” Transporta- tion Research Record: Journal of the Transportation Research Board, No. 1983: 129–139. Transportation Research Board of the National Academies, Washington, D.C. NPC. 2012. Future Transportation Fuels Study: Advancing Technology for America’s Transportation Future. NRC. 2013a. Review of the Research Program of the U.S. DRIVE Partner- ship: Fourth Report. The National Academies Press, Washington, D.C. NRC. 2013b. Transitions to Alternative Vehicles and Fuels. The National Academies Press, Washington, D.C. OFE. 2000. Fuel Cell Handbook. Ogden, J., C. Yang, J. Cunningham, N. Johnson, X. Li, M. Nicholas, N. Parker, and Y. Sun. 2011. Chapter 3: The Hydrogen Fuel Pathway. In J. Ogden and L. Anderson (eds.), Sustainable Transportation Energy Pathways. Institute of Transportation Studies, University of California, Davis. ORNL. 2012. Transportation Energy Data Book, Edition 31. Sun, Y., J. Ogden, and M. Delucchi. 2010. Societal Lifetime Cost of Hydrogen Fuel Cell Vehicles. Institute of Transportation Studies, University of California, Davis. Wipke, K., S. Sprik, J. Kurtz, and T. Ramsden. 2010a. Controlled Hydrogen Fleet and Infrastructure Demonstration Project. National Renewable Energy Laboratory. Wipke, K., S. Sprik, J. Kurtz, T. Ramsden, and J. Garbak. 2010b. Entering a New Stage of Learning from the U.S. Fuel Cell Electric Vehicle Demonstration Project. National Renewable Energy Laboratory. Wipke, K., S. Sprik, J. Kurtz, T. Ramsden, and J. Garbak. 2010c. “DOE’s National Fuel Cell Vehicle Learning Demonstration Project—NREL’s Data Analysis Results.” In G. Pistoia (ed.), Electric and Hybrid Vehi- cles: Power Sources, Models, Sustainability, Infrastructure and the Mar- ket. Elsevier Press. Washington State DOT. 2009. Alternative Fuels Corridor Economic Feasibility Study. more profit, then efforts to develop and commercialize FCVs could receive less emphasis. References ANL. 2012. The Greenhouse Gases, Regulated Emissions, and Energy Use in Transportation (GREET), Fuel Cycle Model version 1 2012. U.S. Department of Energy. API. 2013. Oil & Natural Gas Overview: Service Station FAQs. http:// www.api.org/oil-and-natural-gas-overview/consumer-information/ service-station-faqs (accessed July 20, 2013). Burke, A., Z. McCaffrey, M. Miller, K. Collier, and N. Mulligan. 2005. Hydrogen Bus Technology Validation Program. UC Davis Institute of Transportation Studies. CARB. 2009. Summary of Staff ’s Preliminary Assessment of the Need for Revisions to the Zero Emission Vehicle Regulation. CARB. 2013. Zero Emission Vehicle (ZEV) Program. http://www.arb. ca.gov/msprog/zevprog/zevprog.htm (accessed October 2, 2013). CEC. 2004. California Hydrogen Fueling Station Guidelines. CFCP. Undated. Recent Announcements Show Progress. http:// www.cafcp.org/recent-announcements-show-progress (accessed November 8, 2010). Cho, A. 2004. “Fire and ICE: Revving up for H2.” Science, 305 (5686): 964–965. EERE. 2013. Hydrogen Laws and Incentives. Alternative Fuels Data Cen- ter. http://www.afdc.energy.gov/fuels/laws/3255 (accessed October 2, 2013). EIA. 2013. Annual Energy Outlook 2013. Eudy, L., K. Chandler, and C. Gikakis. 2012. Fuel Cell Buses in U.S. Transit Fleets: Current Status 2012. National Renewable Energy Laboratory. Fuel Cells 2000. 2012a. Fuel Cell Vehicles (from Auto Manufacturers). Fuel Cells 2000. 2012b. U.S. Hydrogen Fueling Stations. Greene, D., P. Leiby, B. James, J. Perez, M. Melendez, A. Milbrandt, S. Unnasch, and M. Hooks. 2008. Analysis of the Transition to Hydro- gen Fuel Cell Vehicles and the Potential Hydrogen Energy Infrastructure Requirements. Oak Ridge National Laboratory.