Hydrogen-Powered Fuel Cell Electric Vehicles Compared to the Alternatives

At the beginning of the 20th century, American drivers faced a fundamental choice: battery electric vehicles (BEVs) or gasoline-powered internal combustion vehicles (ICVs). In 1989 and 1900 more BEVs were sold than all gasoline ICVs and steam engines combined. Gasoline ICVs won out due to longer range, short refueling time and, eventually (after years of picking up jugs of gasoline at the drug store!), a conenient supply of relatively inexpensive gasoline.

We face a similar choice at the start of the 21st century with one big difference: Most knowledgeable analysts now conclude that we must replace most of the venerable gasoline- and diesel-powered ICVs with electric motors to simultaneously eliminate most local air pollution, to cut greenhouse gas (GHG) emissions by 80% below 1990 levels, and to substantially reduce our dependence on petroleum. But we still face a fundamental choice on how to supply the electricity to those electric vehicles: batteries or hydrogen-powered fuel cells, a choice not available in the early 1900's.

We show by detailed computer simulation modeling that hydrogen-powered fuel cell electric vehicles (FCEVs) are superior to BEVs in terms of GHG reductions, range, refueling time and life cycle cost. Despite these overwhelming advantages of hydrogen-FCEVs, the Obama administration in the US has decided to fund plug-in hybrid electric vehicles (PHEVs) and BEVs while downgrading FCEV developments that were strongly supported by the previous Bush administration; Steven Chu, the new Secretary of Energy, eliminated all Federal funding for hydrogen and FCEVs in his first budget request for Fiscal Year 2009 without consulting with any industrial stakeholders (automobile or energy companies), and without consulting his own Hydrogen & Fuel Cell Technical Advisory Committee (HTAC) set up by Congress to keep the Secretary informed on hydrogen and FCEV issues.

Fortunately Congress restored much of the funding for FY 2009. However, Secretary Chu cut funding for the hydrogen and FCEV budet by 22% in his latest (FY2011) request, from $174 million down to $137 million and cut back funding for the very successful on-road FCEV demonstration project from $13 million to $11 million. Meanwhile, many other nations are vigorously pursuing the hydrogen-FCEV option as a cornerstone of GHG and oil reductions in the transportation sector. Germany, Japan, Denmark and South Korea all have substantial FCEV programs. Germany, for example, has formed a consortium of industries¹ called "H2 Mobility" supported by the German governments with a goal of deploying 600,000 FCEVs and 1,000 hydrogen stations all across Germany by 2020.

Greenhouse Gas Emissions

To compare the impact of various alternative vehicles, we developed a detail computer simulation program to compare:

• Gasoline-powered hybrid electric vehicles (HEVs), the base case
• Gasoline-powered Plug-in hybrid electric vehicles (PHEVs) with a maximum of 75% market share to account for those drivers that do not have access to home charging outlets² or assigned off-street parking that could accommodate a charging outlet.
• Biofuel(Cellulosic ethanol) -powered PHEVs
• Hydrogen-powered fuel cell electric vehicles (FCEVs)
• Battery-powered electric vehicles (BEVs) with energy drawn from the electrical grid

We also analyzed vehicles powered by natural gas, including natural gas HEVs, natural gas PHEVs, but did not consider these to be an option for the long-term, since natural gas, although more plentiful in the US with the discovery of new natural gas reserves in shale formations, is not a long-term sustainable option³.

These simulations look at realistic scenarios for alternative vehicle introduction over time. For example, Figure 1 shows the fraction of new car sales in the FCEV scenario. Rather that assuming 100% FCEVs at some point in time, we assume that FCEVs are gradually added to the existing fleet of ICVs, HEVs, and PHEVs, just as HEVs are now being added to the conventional car fleet, so the vehicles on the road in any year are a mixture of conventional vehicles and alternative vehicles, just as we have conventional vehicles and HEVs on the road today. By the end of the century, 98% of new cars sold are FCEVs in the FCEV scenario.

Figure 1. Fraction of new car light duty vehicle sales in the US for the Fuel Cell Electric Vehicle Scenario. Reprint with permission International Journal of Hydrogen Energy (IJHE)

In the model we assume that hydrogen is made initially from natural gas, which is the dominate source of industrial hydrogen today. Hydrogen made from natural gas used in a FCEV reduces GHGs approximately 45% compared to burning gasoline in an ICV. The carbon footprint of hydrogen is then gradually reduced further over time as indicated by Figure 2. Hydrogen is made from biofuels such as cellulosic ethanol, then from biomass gasification, and (possibly) from coal gasification with carbon capture and storage (CCS), and, eventually, from electrolysis of water using "clean electricity" from renewables and nuclear power when and where available, resulting in zero or even negative net GHG emissions⁴.

Figure 2. Sources of hydrogen over the 21st century used in model.

Similarly, we assume that electricity comes initially from a hypothetical electrical grid which takes the West Coast grid mix scaled to the entire US. The US West Coast grid has a lower percentage of coal generation (32%) compared to 52% for the entire US. So this simulation is more favorable to PHEVs and BEVs than would be the case for the entire country. In addition, we assume that the grid becomes greener over time as shown in Figure 3, with increasing shares of renewable energy and nuclear energy over time. In addition, all coal plants are replaced with new coal generators including CCS, or CCS is added to all existing coal plants.

Figure 3. Electricity grid mix for the entire US (Scaled from the US Western Electricity Coordinating Council grid mix on the West Coast) CCS = carbon capture and storage.

The greenhouse gas emissions for each alternative vehicle scenario are calculated using the Argonne National Laboratory GREET model⁵ that takes into account all GHGs from "well-to-wheels." The results are shown in Figure 4. The horizontal line at the bottom of this graph corresponds to an 80% reduction in GHGs from the 1990 levels, the goal suggested by the Climate Change community to stabilize GHGs in the atmosphere. This figure shows that HEVs help reduce GHGs, gasoline-powered PHEVs help even more, but, to achieve the 80% reduction goal, the nation must transition to electric vehicles, either BEVs or FCEVs.

Figure 4. Well-to-Wheels Greenhouse Gas Emissions for the various alternative vehicle scenarios.

As shown below, the FCEV has many advantages over BEVs that will make them much more likely to be accepted by most drivers. Furthermore, the BEV dashed line in figure 4 assumes that all light duty vehicles (including vans, SUVs and light duty trucks) are powered by batteries, which is unlikely with current or even advanced batteries based on today's technologies or advances thereto.

Even with this assumption, the FCEV scenario will reduce GHGs more than BEVs over most of the century, since it will take many decades to clean up the electric grid. Thus by mid-century there are still many coal-fired plants in the US, some without carbon capture and storage (CCS), so electricity to charge batteries on PHEVs still generates considerable GHGs. Hydrogen made from natural gas, on the other hand, begins cutting GHGs 45% immediately, and GHG reductions continue to grow as lower carbon hydrogen sources are introduced according to Figure 2.

The DOE's Argonne National Laboratory has recently completed a detailed analysis⁶ of GHG emissions from PHEVs, BEVs, and FCEVs for a fixed year (2020). Their conclusions generally support the results of our model: FCEVS using hydrogen made from natural gas cut GHGs more than any PHEV or BEV in most parts of the US, regardless of the PHEV all-electric range⁷.

Even in California, with its lower carbon grid mix, FCEVs will have lower GHGS than any PHEV⁸. Furthermore, for most parts of the US, plugging in a PHEV actually increases GHGs, so we would be better off (lower GHGs) by running a PHEV on gasoline (essentially an HEV) and never charging the batteries with grid electricity⁹. In states like Illinois that depend primarily on coal-based electricity, a series PHEV with 40 miles AER like the proposed Chevy Volt would have higher GHG emissions than even a conventional (non-hybrid) gasoline car¹⁰.

Near-Term Greenhouse Gas Emissions

The proponents of PHEVs and BEVs have claimed that these vehicles running on grid electricity are required to cut GHGs in the near-term before FCEVs could enter the marketplace. In our computer simulations, we do assume that PHEVs enter the marketplace five years ahead of FCEVs as shown in Figure 5 to give industry enough time to install sufficient hydrogen fueling systems.

Figure 5. Number of PHEVs and FCEVs on the road in their respective scenarios; The FCEVs lag the PHEV market penetration by five years.

Despite this 5-year head-start for PHEVs, however, hydrogen-powered FCEVs would actually reduce GHGs more than gasoline-powered PHEVs in the decade from 2020 to 2030 as shown in Figure 6. This is because FCEVs powered by hydrogen made initially from natural gas will immediately cut GHGs by approximately 45%, while PHEVs in most parts of the country will not significantly decrease GHGs since they still rely on gasoline for most of their energy, and most electricity is made from coal, the dirtiest fuel (most carbon content).

Figure 6. Near-term (2020 to 2030) greenhouse gas emissions for the alternative vehicle scenarios.

For example, Kromer and Heywood at MIT¹¹ have calculated the GHGs from PHEVs with 10, 30 and 60-mile all electric range. As shown by the solid bars on Figure 7, GHGs are not changed significantly compared to a non-plug-in gasoline HEV....plugging in does not cut GHGS in most parts of the US. Note that the Chevy Volt with 40 miles all-electric range would not cut GHGs in most parts of the country, seen by interpolating between the 30-mile and 60-mile bars on Figure 7. The error bars on these graphs indicate the GHGs from regions dominated by coal plants (right higher bars) and regions dominated by natural gas electrical plants (left lower bars).

Figure 7. Greenhouse Gas Emissions calculated by Kromer and Heywood of MIT [2007] (their figure 38 for Hybrid electric vehicles (HEV), and plug-in hybrid electric vehicles with 10 (PHEV-10), 30 (PHEV-30) and 60 miles all-electric range (PHEV-60) As noted, the upper bars indicate the GHGs for regions with 100% coal-generated electricity, and the lower bars show the emissions for regions that have 100% natural-gas generated electricity; the arrows indicate the GHGs for a future "optimistic, cleaner grid mix" with 50% zero-carbon sources (renewables or nuclear), 20% natural gas and 35% coal (which adds to 105% total, which is not explained in the MIT report), with the fossil generators at operating at higher efficiency (50% for natural gas and 40% for coal, compared to 37% efficiency for natural gas and 33% for coal with current technology.

Oil Consumption

Figure 8 shows the oil consumption calculated from the Argonne GREET model for each alternative vehicle scenario. These curves are in the same order as the GHG emission curves: HEVs help reduce oil, PHEVs help more, biofuel PHEVS help even more, but all-electric vehicles (either BEVs or FCEVs) are required to make substantial reduction in oil consumption. The lower dashed line in Figure 8 corresponds to the oil consumption from the light duty vehicle fleet that would allow the US to supply all its non-transportation petroleum requirements from oil sources in the Americas (North, Central and South with the exception of Venezuela.)

Figure 8. Oil consumption over the 21st century for the various alternative vehicle scenarios. Reprint with permission International Journal of Hydrogen Energy (IJHE)

Near-Term Oil Consumption

As with GHGs, petroleum consumption will be reduced more by FCEVs than by PHEVs in the 2020 to 2030 decade as shown in Figure 9. This is because FCEVs eliminate all oil consumption, while PHEV still rely on gasoline for long distance travel.

Figure 9. Near-term (2020-2030) oil consumption estimates for the various alternative vehicle scenarios (the BEV scenario would have the same oil consumption as the FCEV scenario).

Societal Costs

We have monetized the societal costs of urban air pollution, GHG emissions and oil imports and estimated total societal costs for each alternative vehicle scenario as summarized in Figure 10. Note that by the end of the century, total societal costs for the base case (HEVs only) exceeds $360 billion per year. The alternative vehicles all reduce societal costs to some degree as summarized in the second column of Table 1. The FCEV scenario, for example, reduces total societal costs by $320 billion per year; the second-best option, the BEV scenario, would reduce costs by $300 billion per year (again, assuming that new battery technology is discovered and developed that could economically satisfy customer requirements for all light duty vehicles, including vans, SUVs and light duty trucks).

Figure 10. Total annual societal costs for each scenario.

Table 1. GHG reductions below 1990 level in 2100 and Annual Societal Savings in 2100 for each alternative vehicle scenario

Batteries vs. Fuel Cells

Weight

Batteries are much heavier than fuel cell systems for a given vehicle range as illustrated in Figure 11. This figure also shows the substantial improvement in battery technology over the century, comparing the old lead acid battery technology used in all starter batteries on today's cars, plus the nickel metal hydride (NiMH) battery technology used on many HEVs, and advanced versions of the Lithium ion (Li-ion) battery used in many cell phones and laptop computers. But even with advanced Li-ion batteries, a vehicle with 300 miles range¹² would weigh almost twice as much as a FCEV with 300 miles range.

Figure 11. Vehicle mass for battery electric vehicles (BEVs) and for fuel cell electric vehicles (FCEVs) as a function of vehicle range; these curves include the effects of mass compounding. Reprint with permission International Journal of Hydrogen Energy (IJHE)

Note that the lines are curved and not straight. This results from the effects of mass or weight compounding. To increase the range of a BEV, more batteries must be added; as a result of this extra mass, extra structure must be provided to hold the added mass; then the vehicle motor must be larger to provide the required acceleration for this extra mass¹³; then larger brakes are required to stop the vehicle, etc.

All of these mass increases then require yet more batteries to provide the required range, and the feedback process begins all over again! As a result, the total mass grows non-linearly with range if the calculation is done properly. The same effect increases the mass of the FCEV, but the impact is much smaller, since adding extra hydrogen to increase range is very light. The compressed gas tanks do become slightly heavier, but to a much smaller degree than batteries.

Volume

Most people understand that batteries are very heavy, but many do not realize that batteries also take up considerable space. Some observers who reject FCEVs refer to the large compressed hydrogen tanks required for FCEVs. But the total volume on the vehicle occupied by battery packs is actually larger than the volume required for hydrogen tanks, the fuel cell system and the peak power battery¹⁴ used in most FCEVs as shown in Figure 12.

Figure 12. Storage volume of energy system for BEVs and FCEVS vs vehicle range. Reprint with permission International Journal of Hydrogen Energy (IJHE)

Vehicle Cost

FCEVs are very expensive today, since most are prototype vehicles, often hand-assembled by Ph.D.'s! But Kromer and Heywood at MIT¹¹ have analyzed the likely cost of alternative vehicles once they reach mass production, and have concluded that FCEVS with 350 miles range will cost less than PHEVs with more than 20 miles range, and will cost much less than BEVs with 200 miles range as summarized in Figure 13, which shows the MIT estimates of incremental cost of alternative vehicles over advanced gasoline ICVs.

Figure 13. Mass production incremental cost estimates from Kromer & Heywood (MIT-2007)

Infrastructure Cost

Some commentators have stated that a hydrogen infrastructure might cost hundreds of billions of dollars. These high estimates are usually based on the assumption that hydrogen would be produced at a central plant and shipped by pipeline to fueling stations. This would entail the construction of a national hydrogen pipeline system similar to the existing natural gas pipeline network that would be extraordinarily expensive.

But there are other options for delivering hydrogen to FCEVs. For example, hydrogen can be produced at the fueling station by a process called "steam methane reforming" (SMR). An SMR produces hydrogen from natural gas (methane is the main ingredient in natural gas) and water. In effect, the existing natural gas and water pipelines become the backbone of the "hydrogen infrastructure." In addition, a large quantity of hydrogen is currently supplied to industrial users by truck in both gaseous and liquid form, again avoiding the need for building a national pipeline network. Finally, hydrogen can be generated at the fueling station by electrolyzing water. In this case the water pipelines and electrical grid provides the backbone of the hydrogen infrastructure system. This is only appropriate when and where the electricity is generated by renewables or nuclear or from coal and gas with carbon capture and storage (CCS) to avoid large GHG emissions.

Ironically, the "infrastructure" required to support PHEVs and BEVs could cost much more than the hydrogen infrastructure. For example, a recent survey by Deloitte found that only 39% of drivers have access to a home outlet for charging their vehicles¹⁵, so new charging oulet posts must be installed to charge car batteries in PHEVs and BEVs. In addition, the Electrification Coalition¹⁶, a BEV advocacy group, has recommended that we must install two public charging outlets for every PHEV or BEV initially, decreasing over time to one public outlet for every two BEVs. As shown in Table 2, the costs for a public charging outlets in mass production are estimated at several thousand dollars each. With 4 to 8 hour charging times, one outlet can at best serve one to two BEVs per day, so costs will be high. Furthermore, current costs for Type II charging outlets are much greater than shown in Table 2. For example, Coulomb Technologies¹⁷ is proposing to spend $37 million¹⁸ to install 4,600 public charging outlets, or a cost of $8,043 each.

Table 2. Estimated costs of battery electric vehicle charging outlets

Since at least one charging outlet (either at home or public) must be installed for every BEV the electrical infrastructure costs will vary between $3,000 to $8,000 per vehicle. In contrast, one hydrogen fueling station can support thousands of FCEVs, much like current gasoline stations. Thus a hydrogen fueling station with 1,500 kg/day capacity might cost up to $4 million initially, but it could provide hydrogen for 2,500 FCEVs¹⁹, which corresponds to $1,580 per vehicle. Therefore the BEV infrastructure costs per vehicle may be 2 to 5 times greater than the hydrogen infrastructure costs per vehicle. Over time costs for both infrastructure systems are expected to decrease with mass production.

Refueling Time

One way to visualize the difficulty of rapidly charging car batteries is to consider the power delivered every time we fill up a car tank with gasoline. Pumping 13 gallons of gasoline in 3 minutes is equivalent to 10 megawatts (10 million watts) of power. A Type I 120-volt, 20-amp home electrical circuit is limited to approximately 1.9 kW of power, or 5,000 times slower than a gasoline hose. The National Renewable Energy Laboratory has been monitoring the operation of 140 FCEVs for several years²⁰. The average hydrogen fueling rate was 0.81 kg/minute for 14,000 separate hydrogen fueling events with an average fill time of 3.3 minutes, which is equivalent to 1.61 Megawatts of power, or 840 times faster than the Type I home circuit. In other words, it is much easier and faster to move molecules of gasoline and hydrogen than to move electrons through wires and battery terminals with finite resistance.

Conclusions

1. Electric Vehicles will be required. Internal combustion engines must be replaced by electric motors on most light duty vehicles in the 21st century to achieve our societal goals of an 80% reduction in GHGs below 1990 levels, near-zero local air pollution and a substantial reduction in petroleum consumption.

2. Two Choices: batteries or fuel cells. The electricity used to drive these motors can be supplied by batteries or by hydrogen-powered fuel cells.

3. Fuel cell electric vehicles (FCEVs) are superior to battery electric vehicles (BEVs) in terms of;
      a. Weight
      b. On-board storage volume
      c. Mass production vehicle cost
      d. Fuel infrastructure cost
      e. Refueling time

4. We need all of the above. Despite the overwhelming superiority of FCEVs, we still need a mix of HEVs, PHEVs, BEVs and FCEVs to minimize environmental and energy security risks of transportation. Three major auto companies (GM, Toyota and Daimler) have all published documents advocating the following mixture of alternative vehicles over the next few decades:
      a. Battery electric vehicles for small "city cars" with limited range.
      b. PHEVs for slightly longer range vehicles, and
      c. FCEVs for long-range, full-function vehicles, trucks and buses.

5. Fund a full portfolio of vehicle options. Therefore governments should support the development of all these alternative vehicles. It is too early to pick winners and losers, given the gravity of our environmental and energy security challenges. It is extremely short-sighted and ultimately counterproductive to focus on any one alternative vehicle at the expense of the others; consistent funding and deployment of all the options will maximize our chances of achieving a sustainable transportation system in the decades to come.

6. US loosing our lead. We are concerned that the US is falling behind other nations in the development of FCEVs; if we follow the current path of the Obama administration in supporting BEVs and PHEVs at the expense of FCEVs, then we risk repeating the mistakes of the 1990's, when US automakers failed to vigorously develop hybrid electric vehicles, so that we had to import most HEVs from Japan. Ten or twenty years from now, we may be faced with the same situation with respect to FCEVs with one added twist: if we do not proceed with the build-out of hydrogen fueling systems to continue the deployment and gain hand-on experience with FCEVs, then the US may not even be able to import foreign FCEVs, and we will not be able to meet our GHG reduction goals without an accelerated and inefficient deployment of hydrogen fueling systems. It would be much less costly in the long run (and much more fruitful in terms of lessons learned) to continue the current roll-out of hydrogen stations such as those in the California "Hydrogen Highway" rather than have to rush later to make up for lost time.

References

1. Available at: http://www.hydrogen-planet.com/en/hsub2-sub-mobility-initiative-receives-iphe-award.html
2. A recent survey by Deloitte revealed that only 39% of car drivers have home garages or equivalent that could accommodate a battery charging outlet; thus this model effectively assumes that 36% of drivers(to bring the total fraction of PHEV owners up to 75%) have assigned off-street parking or driveways where a controlled charging port could be installed. Reference: "Gaining Traction: A Customer View of Electric Vehicle Mass adoption in the US automotive market," Deloitte consulting LLC, 2010.
3. In addition, a natural gas-powered PHEV (the best natural gas vehicle option) could not achieve the goal of reducing GHGs by 80% below 1990 levels.
4. Hydrogen made from biomass could result in negative GHG emissions if the CO2 generated at the biomass gasification plant was captured and stored underground. New biomass growing would then take CO2 out of the atmosphere, resulting in a net decrease in GHG emissions for the entire fuel cycle.
5. The Greenhouse Gases, Regulated Emissions and Energy Uses in Transportation (GREET) model, the Argonne National Laboratory
6. Amgad Elgowainy, J. Han. L. Poch, M Wang, A. Vyas, M Mahalik, A. Rousseau "Well-to-Wheels Analysis of Energy Use and Greenhouse Gas Emissions of Plug-In Hybrid Electric Vehicles" Argonne Report # ANL/ESD/10-1; June, 2010
7. Ibid, Elgowainy et al., Figure 6.10
8. Ibid. Elgowainy et al., Figure 6.12
9. Ibid, Elgowainy et al., Figure 6.10
10. Ibid. Elgowainy et al., Figure ES.1
11. Kromer M, Heywood J. Electric powertrains: opportunities and challenges in the U.S. light-duty vehicle fleet. Sloan Automotive Laboratory, Massachusetts Institute of Technology, Publication No. LFEE 2007-03 RP, May 2007.
12. The 2010 Deloitte survey of drivers revealed that over 70% of drivers would require a 300-mile range before they would consider purchasing a BEV.
13. All vehicles in this simulation are designed to have the same characteristics including acceleration 10 seconds for 0 to 60 mpg, hill climbing capability, etc. The base vehicle or glider is a 5-passenger car with a drag coefficient of 0.33, rolling resistance of 0.0092 and cross sectional area of 2.127 m².
14. While hydrogen and fuel cells provide most of the electricity for FCEVs, most FCEVs also include a battery bank to provide extra power during acceleration, and to store energy from regenerative breaking which improves efficiency.
15. "Gaining Traction: A Customer View of Electric Vehicle Mass adoption in the US automotive market," Deloitte consulting LLC, 2010.
16. The Electrification Coalition's "Electrification Roadmap".
17. "Coulomb Technologies to provide 4,600 free charging stations," reported in ConSENSEus: energy for transportation in the US, June 7, 2010
18. $15 million of this $37 million cost was provided by a grant under the US Recovery & reinvestment Act, another indication of the Obama Administration's tilt toward PHEVs and BEVs over FCEVs.
19. Assuming 60 miles/kg FCEV fuel economy which has already been certified by two DOE national laboratories for the Toyota FCEV version of their Highlander SUV, and13,000 miles travel per year.
20. Keith WIpke et al., "Controlled Hydrogen Fleet and Infrastructure Demonstration and Validation Project, Report NREL/TR=560-7451, March 2009, The National Renewable Energy Laboratory