as the gasoline vehicle minus the annual cost of the hydrogen vehicle. This starts out negative, but becomes positive in 2023, when the cost of hydrogen vehicles becomes less than that of a similar number of gasoline vehicles. While the HFCVs first cost remains higher than for the gasoline car, the net fuel cost savings make the annual cash flow positive at the breakeven year.

  • “CMLTV Diff” (dollars) is the cumulative cost difference, the sum of the cost difference over time (starting in 2012), providing a yearly tally of the total funds that would have to be invested to make HFCVs competitive. At first, there is a negative cash flow (early HFCVs cost more than gasoline cars), but eventually as HFCV and hydrogen fuel costs come down, the negative cash flow “bottoms out” in 2023 at a minimum of about $22 billion, when about 5.6 million fuel cell vehicles have been produced. This minimum is the “buydown” investment that must be supplied to bring the HFCV to cost competitiveness.

Most of the negative cash flow is due to the high price of the first few million fuel cell vehicles. This is not surprising, since, initially, fuel cell vehicles cost a lot more than gasoline vehicles (see Figure 6.6). The subsidy that might be needed by automakers or buyers is the sum of the difference in costs between HFCVs and gasoline cars, each year between vehicle introduction in 2012 and life-cycle cost (LCC) breakeven in 2023. This cumulative difference in vehicle first cost for HFCVs (as compared to a reference gasoline vehicle) is about $40 billion (averaged over the 2012-2023 buydown period, this is about $7,000 per car, or an average of $3.3 billion per year for 12 years). Transition dates and costs are summarized in Table 6.6, relative to a reference gasoline vehicle.

The buydown cost is quite sensitive to assumptions for key factors. For example, if fuel cell vehicles could be introduced at their “learned-out cost” (e.g., the cost of HFCVs once they have become technically mature and are manufactured at large scale), buydown cost requirements for vehicles would be greatly reduced, and fuel cell vehicles would become competitive almost immediately. In this case, the primary transition cost would be building a hydrogen infrastructure to the point at which hydrogen is competitive as a fuel (fuel cost per kilometer), on the order of a billion dollars. Note that this would happen much sooner than the vehicles reaching cash flow breakeven (see bottom row in Table 6.6). Box 6.1 explores the sensitivity of the results to assumptions on HFCV fuel economy and incremental costs, and the cost of hydrogen and gasoline.

TABLE 6.6 Transition Costs and Timing for Hydrogen Cases

Breakeven Year (annual cash flow > 0)


Cumulative life-cycle cost difference (between HFCV and gasoline reference car) to breakeven year

$22 billion

Cumulative vehicles first-cost difference (between HFCV and gasoline reference car) to breakeven year

$40 billion (~$3.3 billion/yr)

Number of HFCVs at breakeven year (millions)

5.6 (1.9% of fleet)

Hydrogen cost at breakeven year


Hydrogen demand; number of hydrogen stations at LCC breakeven year

4,200 tonnes/d; 3,600 stations

Total cost to build infrastructure for demand at LCC breakeven year

$8.2 billion

Year when hydrogen fuel cost per kilometer = gasoline price per kilometer


Hydrogen cost ($/kg)


Gasoline price ($/gal)


Total cost to build infrastructure to meet demand in 2023 (LCC breakeven year)

$0.5 billion (1,000 small on-site SMR stations)


Assumed Greenhouse Gas Emissions for Fuels

Until 2020, all hydrogen comes from on-site SMRs with a CO2 release of 100 g CO2 equivalent per megajoule of fuel. After that time, low-carbon sources such as biomass hydrogen and hydrogen from coal with carbon capture and storage are phased in. By 2050, roughly 31 percent of hydrogen is produced via on-site SMRs, the remainder via low-carbon sources (44 percent coal with CCS, 25 percent biomass H2). Thus, the overall emissions for hydrogen supply in 2050 are 37 g CO2/MJ fuel, based on the CO2 values given in Table 6.7, which shows the assumptions regarding the well-to-wheels emissions associated with different fuel supply pathways. In all cases, the carbon emissions from the hydrogen supply are assumed to follow the curve shown in Figure 6.14, where CO2 emissions are shown as declining linearly between 2020 and 2050. The average CO2 emissions might fall faster than this, because most new capacity after

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement