technical details regarding a number of other approaches must be supplemented by other means.

With few exceptions, the processes discussed were in the development, engineering design, and scale-up stages rather than the exploratory research stage. Thus, the technology was fairly well understood for many of the hydrogen production and methane reforming cycles; the principles are known for oil shale processing. An examination of the costs of fuel production indicated that solar-derived fuels were currently significantly more expensive than fuels derived from nonrenewable resources. This report briefly discusses the formal contributions, comments regarding those talks, and the general discussion following the talks.

The first talk concerned the GRI experience in solar fuel research and was given by Kevin Krist, the project manager. The GRI funded research was aimed at low-cost conversion of inorganic materials to gaseous fuels. The program ran from 1981 to 1989 and was funded at an annual level of one million dollars or more. The program emphasized photochemical approaches, although thermochemical processes were also evaluated. It is of note that support ended in 1989 and no gas production research is currently underway with GRI support. The key technical factor cited for the lack of success in water-splitting processes was the difficulty in maintaining charge separation in aqueous solutions for sufficient time for fuel-forming reactions to occur. In non-aqueous systems the researchers concluded that effective charge separation was possible, but that the systems work indirectly—generating electricity first, and subsequently producing gaseous fuels via electrolysis. Krist stated that GRI felt that very long-term research was required to develop the type of molecular catalysts that were probably necessary to carry out the reactions efficiently in aqueous phase.

The GRI studies analyzed the cost of solar-produced fuels by assuming a hypothetical nonconcentrating pond system capable of making hydrogen or methane using a membrane separation technique. The costs were based on land area, membrane requirements, and various other costs to arrive at a fuel price. Methane costs for the pond were estimated at $30/MMBtu and for concentrating systems at $82/MMBtu. Methane production costs for mature thermochemical processes were estimated at $40/MMBtu. These costs were contrasted with current wellhead natural gas prices of less than $2/MMBtu. Methane costs were strongly influenced by process efficiency and material costs. The cost of photovoltaic-operated electrolysis to produce hydrogen was estimated at $20/MMBtu for mature plants. By comparison, 1988 prices for hydrogen produced by the reforming of natural gas were about $7/MMBtu. Another conclusion was that high value fuels should be sought to improve the economics of solar fuels processing.

The engineering considerations and costs for thermal water-splitting, thermochemical, and hybrid solar hydrogen production were discussed by Ertugrul Bilgen. He stressed that the plant scale must be very large to supply modest amounts of hydrogen. For example, a typical plant size considered of 500 MW thermal would only supply about 1/3% of Hawaii's energy needs. He concluded that high efficiency and large plant size were necessary for success. Thermal decomposition of water at 2500ºK, for the conditions considered, had a 5% hydrogen composition. The key problem is avoiding the back reaction after dissociation. Separation and recovery may be carried out at low or high temperatures; at low temperatures, rapid quenching is followed by separation by a diffusing membrane; at high temperatures, separation is affected by selectively permeable membranes. However, these means of separating the products have not yet undergone practical testing. Bilgen also concluded that, for better economics, the plants should probably be operated as cogenerators.

Several thermochemical processes were discussed. Most of these processes were based on sulfur cycles involving decomposition of sulfuric acid and sulfur dioxide at temperatures of about 1200ºC. The GA, Mark 16, and Cristina processes were discussed. The Cristina process offered the advantage that sulfuric acid and SO3 decomposition could be operated in reverse at night using air or oxygen as a vector to provide a continuous source of heat for



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