Page 775
Appendix H
A Solar Hydrogen System
To convert solar radiation to electricity, one makes use of photovoltaic materials akin to the solar cells used to energize battery-free pocket calculators. However, as an energy source, solar radiation is relatively dilute. Impressive amounts of (desert) land area would have to be devoted to this use in order to replace fossil fuel supplies. For example, it is estimated that the complete replacement of such supplies in the United States would require total collector fields on the order of 50,000 square miles, about 1 percent of the total U.S. land area (Ogden and Williams, 1989). On the other hand, obtaining the same power from biomass grown on energy farms would require more than 10 times that area. Even obtaining synfuels from coal would, in 14 years, use up the 24,000 square miles of land thought to be available for strip mining.
Once the solar energy system generates electricity, the electricity can be used to generate hydrogen. Hydrogen is a transportable, clean-burning fuel that can be used as energy for vehicles, planes, and many other devices. This appendix describes the cost-effectiveness of one such system.
Photovoltaic Materials
Prior to 1980 the only commercially available solar cells were those made of high-grade single-crystal silicon. Fabrication of these crystals requires large amounts of time, material, and energy. Much more promising for application to solar power is the later technology of thin-film amorphous (i.e., noncrystalline) silicon cells. The films, typically 1 micron (0.0001 cm) thick, are prepared by deposition from silicon vapor onto a substrate such as glass, plastic, or stainless steel, a process that lends itself easily to mass production. A square meter of cell area would require only 3 g
Page 776
of silicon, a very abundant element. The efficiency of conversion of the power in solar radiation to electricity has increased from 1 percent, for the first cells produced in 1976, to almost 12 percent for modest-area laboratory modules and almost 14 percent in 1987 for small-area laboratory cells. Higher efficiencies, estimated at 18 to 20 percent, may be attained in a few years with multilayer cells, each layer tuned to a different part of the solar spectrum.
Commercial Photovoltaics
The Alabama Power Company has a 100-kW amorphous-silicon generating field in operation at present. Efficiencies of currently available commercial photovoltaics range from 5 to 7 percent. Present-day manufacturing facilities are typically of modest capacity, on the order of 1 MW/yr, at a cost of $1.50 to $1.60 per peak watt. Within a few years, plants of 10-MW capacity per year may be on line. These plants are expected to produce cells of 6 percent efficiency for about $1.00 per peak watt. A 50-MW power plant to sell electricity to the Southern California Edison Company (Chronar Corporation, anticipating a photovoltaic cost of $1.25 per peak watt) and a 70-MW/yr production plant (ARCO Solar, Inc.) are in the planning stage. Looking to the end of the 1990s and the possibility of production levels of many hundreds of megawatts per year, Ogden and Williams (1989) project that costs could drop to the range of $0.20 to $0.40 per peak watt, based on reduced outlays for specialty glass, labor, and depreciation, together with commercial efficiencies increasing to 12 to 18 percent. Allowing for electrical wiring losses and for dirt and dust on the modules would reduce their overall efficiencies by an estimated 15 percent, that is, to 10.2 to 15.3 percent. Land costs, site preparation, array wiring, support structures, and other construction represent additional area-related costs that would come to about $50/m2 with present technology, but economies of scale might bring these down to $33/m2.
On the other hand, these figures are pertinent for the U.S. Southwest, and supplying power to other parts of the country means finding means other than electric power lines for energy transport. Also note that these costs are much lower than that used in the Mitigation Panel's analysis as described in Appendix J. Rather than using projections of cost, the panel made a deliberate decision to use only current cost in estimating the cost-effectiveness of different energy options.
Hydrogen Costs
The cost for the electrolytic production of hydrogen depends on the capital cost of the electrolyzer and the cost of the DC electricity to run it. There
Page 777
is little economy of scale beyond a hydrogen production rate of 2 MW. Similarly, the scale economies for photovoltaic power disappear beyond levels of 5 to 10 MW. The hydrogen production units could then be highly modularized, with typical unit capacities of 5 to 10 MW and per-unit capital costs of $4 million to $12 million, depending on photovoltaic module costs. Projected costs for solar hydrogen produced in the Southwest would range from $31.80/GJ (equivalent to $3.88 per gallon of gasoline) with 6 percent efficiency for the photovoltaic module producing DC electricity at $0.089/kWh (1990) to approximately half those costs by 1995 and to $9.10/GJ ($1.11 per gallon of gasoline equivalent) based on 18 percent module efficiency for the year 2000. Compression to 70 atmospheres, for transport through a 1000-mile pipeline, would add another $0.16 to $0.20 to the cost per gallon of gasoline equivalent.
Phasing In
One of the very attractive features about a solar hydrogen power economy is that it lends itself to a gradual phase-in. Even today, photovoltaic power is very economical for specialized purposes including corrosion protection, spacecraft, navigation buoys, and small remote water pumps or electric power sources. Installations for supplying peak-load daytime power to utilities are marginally economic at the present time. Daytime power for residential use would be economic at solar module costs of $0.70 to $1.50 per peak watt.
Hydrogen-powered transport, although feasible today for lightweight vehicles with modest ranges, would benefit greatly from improvements in the technology for hydrogen storage. One anticipates that hydrogen-powered transportation would be economical first for fleet vehicles and, as such, could be tested initially in major cities in the Southwest without recourse to pipelines for hydrogen transmission.
Summary
Photovoltaic hydrogen power offers a number of advantages. The energy source is radiation from the sun, the materials involved are abundantly available, and the burning of hydrogen fuel iswith the exception of nitrogen oxidesfree of polluting or greenhouse gas emissions, including CO, CO2, volatile organics, SO2, and particulate matter. The basic technology exists today, and some small-scale applications of solar power are economical even at the present time. Implementation of solar hydrogen power on a larger scale would lend itself to gradual phase-in, and one can expect to see increasingly important applications become economical as improvements are made in solar module efficiency and in hydrogen storage technology.
Page 778
Reference
Ogden, J. M., and R. H. Williams. 1989. Solar Hydrogen: Moving Beyond Fossil Fuels. Washington, D.C.: World Resources Institute.