The Bush Administration and Hydrogen

See allHide authors and affiliations

Science  21 Nov 2003:
Vol. 302, Issue 5649, pp. 1331-1333
DOI: 10.1126/science.302.5649.1331

Two recent articles, “Potential environmental impact of a hydrogen economy on the stratosphere” (T. K. Tromp et al., Reports, 13 June, p. 1740) and “Rethinking hydrogen cars” (D. W. Keith, A. E. Farrell, Policy Forum, 18 July, p. 315), may have caused some to question the goals and objectives of President Bush's FreedomCAR and Hydrogen Fuel Initiative. Although we welcome rigorous scrutiny and public discussion, we must, for the record, respond to several errors and mischaracterizations in the articles.

We believe that Tromp et al. grossly overstated the estimates of hydrogen (10 to 20%) that would escape from hydrogen production, distribution, and refueling systems. These losses were based on a paper on liquid hydrogen (1), which has inherently higher release rates because of the need to release “boil off” through pressure valves. Our work, on the other hand, is focused on gaseous hydrogen. Moreover, because of safety and other considerations, hydrogen refueling systems, unlike gasoline refueling, will be designed as “closed” systems with negligible losses to the environment. We have already demonstrated such systems at prototype refueling stations. Remarkably, the authors assumed an annual loss of 60 million metric tons, which ironically is roughly the amount of hydrogen we estimate would be needed to fuel an entire domestic fleet of 230 million light-duty hydrogen fuel cell vehicles. Obviously, if we only need about 60 million metric tons to fuel the entire domestic light-duty fleet, leakage rates would be far less.

Keith and Farrell state that they are taking “a longer-term strategic view of energy policy,” yet they offer only short-term measures that ultimately will not solve the United States' dependence on foreign oil or address air pollution and greenhouse gas concerns. We agree that short-term measures are important, which is why the Bush Administration advocated hybrid vehicle tax credits, raised Corporate Average Fuel Economy (CAFE) standards for the first time since the 1996 model year (the greatest increase in fuel economy standards in the past 20 years), and supports a renewable fuels standard to increase ethanol production and use. But these are interim strategies that can only briefly moderate, and cannot completely eliminate, our increasing demand for foreign oil. Over the long term, alternatives to petroleum fuels are needed, and hydrogen is the only energy carrier that offers the prospect of a domestically based zero-emissions transportation fuel.

The Department of Energy is not rushing to deploy hydrogen cars, as Keith and Farrell seem to suggest and as some in Congress are urging us to do. Instead, we are engaged in a long-term research and development effort focused on key enabling technologies. Only after these technologies progress to the point where they can meet customer expectations, and only when industry can establish a business case for substantial investments in hydrogen infrastructure, will hydrogen fuel cell cars be successfully commercialized in large numbers. Although some fuel cell vehicles are on the road today, we believe it is unlikely that affordable fuel cell vehicles will be produced for mass consumer markets until sometime after 2015.

  • *Assistant Secretary of Energy


  1. 1.


Garman argues, correctly, that if people make very little molecular hydrogen, then very little can leak into the atmosphere. However, he considers only the 53 megatons of annual hydrogen production required for a single program in a single country—one or two orders of magnitude smaller than the scale a future hydrogen economy must take if it is to significantly impact global fossil fuel use.

The Department of Energy (DOE) has estimated that 265 megatons of hydrogen per year would be needed to meet the expected transportation energy needs, alone, of the United States in 2020 (1). More broadly, fossil fuel use in the United States in the year 2000 produced 3.3 TW of energy, and 3.5 gigatons of hydrogen would have been required to completely replace these fossil fuels with fuel cell technologies (1). Globally, fossil fuels produced about 10 TW of energy in 2000, and 10.6 gigatons of hydrogen would have been required to replace them. By the year 2020, these numbers are expected to increase by factors of about one-third. That is, the scale of H2 production Garman suggests amounts to less than 1% of the global energy demand by the year 2020.

While few have suggested that molecular hydrogen will be the medium for most or all energy use within our lifetimes, much of the interest in a hydrogen economy comes from its potential to significantly reduce fossil fuel use. Although people are likely to disagree on what is significant, one definition comes from studies of greenhouse warming, which suggest that 30% reduction in CO2 emissions by 2020 will be needed to stave off the worst consequences (2). If this 30% were achieved through use of hydrogen fuel cell technologies (in the case that H2 comes from sources with low CO2 emissions), global annual production of hydrogen must be about 4.7 gigatons per year. Note that this calculation, based on DOE estimates of the relationship between energy use and H2 production, differs from the Schultz et al. estimate of H2 needed for 50% reduction in fossil fuel use (3, 4); the source of this discrepancy is unclear to us.

Garman also takes exception with our suggestion that economy-wide leakage of up to 10 to 20% should be considered. However, more recent estimates, including one by a broad and highly qualified group, are consistent with our discussion. The recent DOE report cited above (1) states: “Leakage rates much greater than 1% are likely if no action is taken to engineer systems in advance to minimize hydrogen leakage.”

This report does not discuss the possibility that economy-wide leakage rates could be less than 1%. Similarly, Schultz et al. (3) adopt loss rates of 3% (preferred) to 10% (extreme) for their recent model and describe the upper limit that we discussed as “possible but very unlikely [for] safety and economic reasons.” Prather's recent calculations also make use of the 3 to 10% estimate (4). Combination of Schultz et al.'s preferred leakage rate (3%) with the H2 production needed to replace one-third of projected global fossil fuel use in 2020 results in expected emissions of about 140 megatons per year—similar to the current amount in the entire atmosphere, or 1.8±1.5/0.50.5 times current annual production from all sources (5). There are insufficient data to project how such a rise in hydrogen sources would translate into increased steady-state atmospheric concentrations, because the current rate of soil uptake and its dependence on atmospheric concentration are poorly known. Nevertheless, it seems reasonable to consider that factors of several increases in sources could lead to factors of several increases in concentration.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.


Garman's claim that we offer no measures to address air pollution or greenhouse gas emissions is simply not true. To restate, policies to address current high emitters, to improve average vehicle efficiency, and to reduce emissions of CO2 and pollutants in the electric power sector will be highly cost-effective and should be aggressively pursued in the near term, while long-term goals can be addressed by research on biofuels and synthetic petroleum, in addition to hydrogen.

The Bush Administration's minor (7%) increases in fuel economy for the least efficient half of light-duty vehicles and small changes in tax credits are indeed short-term measures. The new light-truck fuel economy standard will only slow, not reverse, the steady decline in average light-duty fuel economy the United States has experienced since 1986. Moreover, the rhetoric of hydrogen cars silently assumes that technological advances will be deployed by industry to achieve public policy goals, despite the pervasive evidence that, absent regulation, firms (correctly) use new technologies to increase returns and market share.

We know of no conclusive evidence that hydrogen is a better long-term transportation fuel than synthetic hydrocarbons, electricity, biofuels, or some combination, let alone proof that it is the “only energy carrier that offers the prospect of a domestically based zero-emissions transportation fuel.” We commend the Department of Energy's hydrogen research program for focusing on onboard storage—a key issue. However, given that the solutions we recommend—efficiency and petroleum substitutes—will be the cheapest ways to reduce petroleum use and CO2 emissions for several decades, a target for the deployment of hydrogen cars in 2015 is an example of the government picking the technological winner and rushing to judgment.

Navigate This Article