uted generation, the cost of sequestration appears prohibitive (DiPietro, 1997). Release of carbon dioxide from distributed generation plants during the period of a transition to a hydrogen economy may be a necessary consequence unless an alternative such as hydrolysis with electricity from renewable resources becomes sufficiently attractive or R&D significantly improves distributed natural gas production systems. Further information on the technology and the economics of conversion is given in Appendix G.
Distributed generation from natural gas could be the lowest-cost option for hydrogen production during the transition. However, it has never before been achieved in a manner that meets all of the special requirements of this application. The principal challenge is to develop a hydrogen appliance with demonstrated capability to be mass-produced and operated in service stations reliably and safely with only periodic surveillance by relatively unskilled personnel (station attendants and consumers). The capability for mass production is needed in order to meet the demand during the transition, when thousands of these units would be needed, and in order to minimize manufacturing costs. These units need to be designed to maximize operating efficiency and to include the controls, “turndown” capability, and hydrogen storage required to meet the variable demand for hydrogen during a 24-hour period. They must also be designed to meet the hydrogen purity requirements of fuel cells. Steam reforming process technology is available for this application, and companies have already provided one-of-a-kind units in the size range of interest.1 Whether it will be possible to utilize partial oxidation or autothermal reforming for the distributed generation of hydrogen appears to depend on developing new ways of recovering oxygen from air or separating product hydrogen from nitrogen. This is needed because conventional, cryogenic separation of air becomes increasingly expensive as unit size is scaled down. Membrane separations, in contrast, appear amenable to this application and may provide the means for producing small, efficient hydrogen units.
Currently, there is little if any market for mass-produced hydrogen appliances such as those described, and it is clear to the committee that the DOE should stimulate development of these devices. The primary challenges involve the development and demonstration of the following:
A mass-produced hydrogen appliance suitable for distributed generation in fueling stations, and
A complete hydrogen system for fueling stations, capable of meeting variable demand for hydrogen on a 24-hour basis.
Each of these challenges is discussed below.
The committee estimates that, with further research and development, the unit capital cost of a typical distributed hydrogen plant producing 480 kilograms per day (kg/d) of hydrogen could be reduced from $3,847/kg/d to $2,000/kg/d, and the unit cost of hydrogen reduced from $3.51/kg to $2.33/ kg. These hydrogen unit costs are based on a natural gas price of $6.50 per million British thermal units (Btu); a change in natural gas price of plus or minus $2.00 per million Btu would change hydrogen cost by about 12 percent with current technology. Improved plants could reduce CO2 emissions from an estimated 12.1 to 10.3 kg per kilogram of hydrogen, and overall thermal efficiencies could improve from 55.5 to 65.2 percent, in each case without sequestration. Additional information on these estimates as well as estimates for central station (i.e., large, centralized) hydrogen generators using natural gas is included in Appendixes E and G.
The DOE program publications indicate that the program on distributed generation will include demonstration of a “low-cost, small-footprint plant” (DOE, 2003a, b). However, it is not clear whether the program gives priority to distributed generation or includes an effort to demonstrate the benefits of and specific designs for mass production in the specified time frame of the program. The latter would involve concomitant engineering, including design for manufacturing engineering to guide research and prepare for mass production of the appliance. It would also include development of a system design for a typical fueling facility, including the generation appliance, compression, high-pressure storage incorporating the latest storage technology, and dispensers. With today’s technology, the ancillary systems cost about 30 percent as much as the reformer. The committee believes that these costs can be reduced by over 50 percent and that efficiency can be improved through system integration and incorporation of the latest technology. Compression and high-pressure storage are examples of areas in which significant improvements are expected.
The DOE program is positioned to stimulate the development of newer concepts such as membrane separation coupled with chemical conversion, and this seems appropriate to the committee. However, most of the effort appears to be directed toward partial oxidation or autothermal reforming. The committee believes that steam reforming could be the preferred process for this application and that it should also be pursued in parallel with the effort on partial oxidation.
Finally, the committee notes that the DOE program places significant emphasis on centralized hydrogen plants using natural gas and believes that this effort should be limited, given the increasing importation of natural gas, to those developments that would be applicable to distributed generation.
Recommendation 8-1. The Department of Energy should focus its natural gas conversion program on the develop-