In an October 2003 paper for the Commission, we examined the difficulties facing hydrogen as an AFV. We will not repeat that analysis here, but will briefly report on some of the key analyses published subsequently, which tend to reinforce our main conclusions The central challenge for any AFV seeking government support beyond R&D is that the deployment of the AFVs in the infrastructure to support them must cost effectively address some energy or environmental problems facing the nation.

Yet in the spring issue of Issues and Science and Technology, two hydrogen advocates, Dan Sperling and Joan Ogden of U.C. Davis, wrote, "Hydrogen is neither the easiest nor the cheapest way to gain large near- and medium-term air pollution, greenhouse gas, or oil reduction benefits."34 A 2004 analysis by Jae Edmonds et al. of Pacific 10

Northwest National Laboratory concluded in that even "in the advanced technology case with a carbon constraint . hydrogen doesn't penetrate the transportation sector in a major way until after 2035." A push to constrain carbon dioxide emissions actually delays the introduction of hydrogen cars because sources of zero-carbon hydrogen such as renewable power can achieve emissions reductions far more cost-effectively simply replacing planned or existing coal plants. As noted above, our efforts to reduce greenhouse gas emissions in the vehicle sector must not come at the expense of our efforts to reduce greenhouse gas emissions in the electric utility sector.

In fact, Well-to-Wheels Analysis of Future Automotive Fuels and Powertrains in the European Context, a January 2004 study by the European Commission Center for Joint Research, the European Council for Automotive R&D, and an association of European oil companies, concluded that using hydrogen as a transport fuel might well increase Europe's greenhouse gas emissions rather than reduce them. That is because many pathways for making hydrogen, such as grid electrolysis, can be quite carbon-intensive and because hydrogen fuel cells are so expensive that hydrogen internal combustion engine vehicles may be deployed instead (which is already happening in California). Using fuel cell vehicles and hydrogen from zero-carbon sources such as renewable power or nuclear energy has a cost of avoided carbon dioxide of more than $700 a metric ton, which is considerably higher than most other strategies being considered today.35

A number of major studies and articles have recently come out on the technological challenges facing hydrogen. A DOE study noted that transportation fuel cells currently cost about $5,000/kw, some 100 times greater than the cost of internal combustion engines.36 A 2004 article for the Society of Automotive Engineers noted, "Even with the most optimistic assumptions, the fuel cell powered vehicle offers only a marginal efficiency improvement over the advanced [diesel]-hybrid and with no anticipation yet of future developments of IC engines. At $100/kW, the fuel cell does not offer a short term advantage even in a European market."37

A prestigious National Academy of Sciences panel concluded a major report in February with a variety of important technical conclusions.38 For instance, the panel said, "The DOE should halt efforts on high-pressure tanks and cryogenic liquid storage.. They have little promise of long-term practicality for light-duty vehicles." A March study by the American Physical Society concluded that "a new material must be discovered" to solve the storage problem.39 An analysis in the May 2004 issue of Scientific American stated, "Fuel-cell cars, in contrast [to hybrids], are expected on about the same schedule as NASA's manned trip to Mars and have about the same level of likelihood."40
Hydrogen fuel cell vehicles face major challenges to overcome each and every one of the barriers discussed above. Of all AFVs and alternative fuels, fuel cell vehicles running on hydrogen are probably the least likely to be a cost-effective solution to global warming, which is why the other pathways deserve at least equal policy attention and funding.