Not long ago, proposals for using the

ENERGY Sun's energy were apt

to be received with
considerable skepti-

cism. Within a few agencies of the feder-
al government and at an increasing num-
ber of university and industrial labora-
tories, that is no longer the case. Indeed,
perhaps the most impressive testimony
to the prospects for this type of energy
is the score of prestigious scientists and
engineers who have begun working on
methods for converting the sun's radia-
tion into forms more useful to man-
heat, electricity, or chemical fuels.

Within 5 years, many of these scientists believe, solar-powered systems for heating and cooling homes could be commercially available at prices competitive with gas or oil furnaces and electric air conditioners. Still more significant, but farther in the future, may be means of using heat from the sun to generate electricity; experimental solarthermal units have been constructed in several countries, and several groups in the United States are designing systems to take advantage of improved materials and manufacturing techniques. Eventually the direct conversion of solar radiation to electricity by means of photovoltaic cells or its bioconversion to wood, methane, or other fuels on a large scale may become economically feasible.

Solar radiation is the most abundant form of energy available to man, and is so plentiful that the energy arriving on 0.5 percent of the land area of the United States is more than the total energy needs of the country projected to the year 2000. Sunlight is diffuse and intermittent, however, and its use on earth requires large areas to collect sufficient amounts of energy and, for most applications, the means to store energy. Despite its abundance, solar energy has not been exploited except in a limited way in water heaters, furnaces, and space applications; nor are the technologies that would allow widespread use commercially available. Systems for heating and cooling houses or for generating electricity with sunlight could be built now, but they would cost more than comparable systems that burn fossil fuels. For some applications, how

ever, the disparity in cost may rapidly disappear as solar technology improves and as the costs of fossil fuels rise.

Whether or not solar energy becomes generally available in the near future, there is growing agreement that this source of energy will be important in the long run. That being the case, proponents believe that it is the most underfunded area of research in the energy field, accounting for less than 1 percent of federal research expenditures related to energy.

Of the proposed uses of solar energy, heating and cooling for homes and lowrise commercial buildings are the most developed and will almost certainly constitute the first significant use of solar energy in this country. Solar water heaters are already in commercial use in Florida and in several countries overseas. Experimental houses have been equipped with solar heating systems and preliminary development of cooling systems has begun.

Solar Heating in the Home For space heating, the solar collector is typically a black metal surface that readily absorbs sunlight and is covered with one to three panes of glass to reduce the heat loss. The glass is transparent to the incoming sunlight, but absorbs the longer wavelength radiation emitted by the hot metal, so that a "greenhouse" effect is created and the effectiveness of the collector is increased. The heat is collected in water or air that is circulated through the collector during the day, and part of it is stored for release at night or in bad weather. Hot water, hot rock, and chemical (change of phase) storage systems have been experimentally tested, depending on the type of heating system envisioned (1).

For air conditioning, most investigators believe, refrigeration systems that depend on absorption of the coolant fluid appear to offer the best choice. Experimental cooling units are being developed by several university and industrial research groups. At the University of Delaware, for example, a group headed by K. W. Boer is designing complete household energy systems that would utilize heat pumps for space conditioning. In other prototype sys

tems, such as that developed by Erich Farber at the University of Florida, heat from the sun is used to drive ammonia from an ammonia-water solution, and the ammonia is collected and condensed. When cooling is needed, the liquid ammonia is allowed to evaporate and expand as in a conventional cooling system, and the spent vapor is reabsorbed in water.

For absorption refrigerating systems to work smoothly, temperatures around 120°C or higher will be needed, and thus solar collectors that are more efficient than those for heating purposes alone will be required. One possibility may be surface coatings of the type developed in recent years for space applications, which emit very little of the solar radiation that they absorb and which consequently attain higher temperatures than uncoated metal collectors. If such coatings can be produced on a large scale, their use might help to reduce the cost of solar heating and cooling, since collectors are the most expensive item of a solar energy system. Combined cooling and heating systems, which have not yet been built, are also expected to improve the economic prospects for both because of the joint use of the collector.

Substantial technical problems remain to be solved in the design of cooling systems, in the manufacture of surface coatings for improved solar collectors, and in the optimization of combined solar heating and cooling systems. In most regions of the country backup systems based on conventional fuels will be needed for extended periods of bad weather. Nonetheless, one estimate indicates that if systems were commercially available now, solar heating would be cheaper than electric heating in nearly all of the United States and would be competitive with gas and oil beating when these fuels double in cost (2). Proponents believe that solar heating and cooling systems could ultimately supply as much as half of the nearly 20 percent of total U.S. energy consumption that is now used for residential and commercial space conditioning and could reduce the peak use of electricity in summer.

For implementation of this technology, however, some means to overcome

12

溝酒

3

what are essentially social problems is likely to be necessary. As Jerry Weingart of the California Institute of Technology put it, "developing the technology is not enough," because the fragmented building industry is traditionally slow to adopt new techniques. Solar heating systems, despite their lower fuel costs, will entail higher initial costs, thus discouraging consumer acceptance; some observers have suggested that governmental encouragement in the form of tax incentives or energy performance construction codes should be part of a national energy policy. The slow rate of replacement of housing, in any case, guarantees that several decades will pass before a new heating system could have a significant impact on total energy use. Given the growing shortage of fossil fuels, however, it seems clearly advantageous to move in that direction.

The generation of electricity with heat from solar energy is a more difficult challenge, and there are conflicting ideas about the best approach to the problem. Some engineers believe that small generating units located where the electricity is to be consumed are the ideal way to utilize a resource that is inherently diffuse and well distributed. This group favors the use of power turbines that would operate at temperatures considerably lower than those common in nuclear or fossil-fuel power plants, despite the low thermal efficiency, between 10 and 15 percent, that these units would have. Others have proposed large solar-thermal facilities modeled on existing central power stations. The two concepts differ both philosophically and technically.

Small vapor turbines that used heat from solar collectors to generate electricity were demonstrated by Harry Tabor of Israel's National Physical Laboratory in Jerusalem at the United Nations conference on new sources of energy, held in Italy in 1961. A miniature solar power plant in Senegal is already in operation, and experimental solar engines have been developed by several investigators in the United States. Typically, these units operate at temperatures below 200°C. Their economic advantages relative to other sources of electricity have not been demonstrated, and the concept has attracted only limited interest, in part because of the difficulty of decentralizing the present electrical generation and distribution system.

Preliminary efforts to develop large central power plants are under way.

This concept has attracted considerable interest, although substantial problems remain to be solved before such plants could be economically competitive. Still higher temperatures, between 300° and 600°C, are required to operate modern steam turbines, complicating both collection of solar radiation and the storage of thermal energy. To capture enough energy at these temperatures, mirrors or lenses larger than any yet built will in all probability be needed to concentrate sunlight. Because large areas will be required-in most designs, about 30 square kilometers for a 1000megawatt power station-the transfer of heat from the far-flung solar collectors to the generating facility is also a complicated process. The cost and endurance of the collecting apparatus under operating conditions is a critical but undetermined factor.

Central Power Station

One design proposed by a group that is headed by Aden Meinel of the University of Arizona would use Fresnel lenses to focus sunlight onto a stainless steel or glass ceramic pipe, thus concentrating the solar flux ten times above its normal value. The pipe is covered with one of several types of selective coatings that emit only a small proportion, between 5 and 10 percent, of the energy they absorb and is enclosed in an evacuated glass chamber to reduce conductive and convective heat losses. Nitrogen gas is pumped through the pipe at velocities of about 4 meters per second to transfer the heat from the collectors to a central storage unit. The Arizona team plans to use a eutectic mixture of salts, mostly sodium nitrate, as a heat storage medium; the heat would be used to produce steam for a turbine as needed. Liquid metal or the molten salt mixture itself, despite the greater difficulty in handling these substances, might also be used to transport heat from the collectors to the storage unit.

A second group, headed by Ernst Eckert of the University of Minnesota and Roger Schmidt of MinneapolisHoneywell, Inc., has also begun work on the central power station. Their design includes a self-contained, decentralized system for collecting and storing solar heat. A parabolic reflector would concentrate sunlight onto a heat pipe, a device that can transport heat along its length efficiently by convective processes and that does not require a fluid to be pumped through it. The

pipe's outer surface would be a selective coating, and the pipe would be enclosed in an evacuated chamber. A small heat storage tank attached to each heat pipe and reflector would complete the unit; no centralized heat storage facility would be used. Underground pipes would bring water to each stor age tank and return it as steam directly to a turbine-thus reducing the pumping costs, the Minnesota team claims, compared to the nitrogen system. In addition, they believe, the self-contained system would be easier to construct and maintain.

The effectiveness of the selective coating with which the collecting surface is covered largely controls the temperatures that can be achieved. Two types of selective surfaces are known, both of which absorb much of the incoming radiation-in the visible region of the spectrum but which emit only a small portion of the infrared heat radiation. Surfaces such as one developed by Minneapolis-Honeywell for the Air Force rely on optical interference between two reflective layers separated by a transparent layer of the correct thickness; thin films of this type have been routinely produced by vacuum coating techniques in the commercial manufacture of tinted glass for the exteriors of new office buildings. A second type of surface, developed by B. Seraphin at Arizona, is composed of silicon or similar materials that naturally have selective properties. Layers of silicon and nonreflecting materials are laid down on a highly reflective substrate by chemical vapor deposition techniques; the silicon absorbs sunlight, but transmits infrared radiation, so that the composite surface has a high reflectanceand hence a low emittance-in the infrared.

These selective coatings are particularly important for solar collectors that are built without mirrors or lenses. Simple planar collectors have several advantages over the concentrating systems in that the concentrating collector must focus sunlight on the absorber and hence must follow the sun's motion in the sky; machinery to allow daily tracking complicates the collector design. In addition, focusing collectors operate only on direct sunlight, whereas planar collectors can utilize diffuse sunlight as well-and thus can function in cloudy or hazy weather. Because the performance of some of the most selective coatings decreases markedly at high temperatures, however, power plants using

them would have to operate at temperatures below 350°C, with correspondingly reduced efficiency in the steam turbines. Improved selective coatings may allow planar collectors-which Meinel and his co-workers believe, in principle, to be the most effective in areas of the United States other than the cloudless Southwest-to be used. But most initial designs are based on the assumption that concentration of the sunlight will be necessary, and in these systems the fabrication, cost, and durability of the concentrators are the major concern.

The trade-offs between different types of collectors are not the only feature of the design of solar thermal plants still open to debate. Even with concentrating collectors, it may prove advantageous to operate the system at a reduced temperature, according to the Minnesota team. Their analysis shows increasing efficiency of the collectors, but decreasing efficiency of the thermodynamic cycle of the turbines as the operating temperatures are reduced, with the optimum temperature dependent on detailed design of the system and on the heat storage medium chosen. Heat pipes of the size envisioned have never been built, and other hardware details remain to be considered.

Both groups of investigators believe that the cost of solar-thermal plants will be not more than two or three times what fossil-fueled or nucleargenerating plants cost now, and that rising fuel costs will eventually tip the

balance in favor of solar-thermal plants whose fuel is "free." Before accurate estimates of costs can be made, they agree, more detailed engineering studies and some additional research are necessary. But Meinel, at least, believes that full-scale solar-thermal power plants could be built as early as 1985 with an adequate research effort. Other estimates are somewhat less optimistic, but a group of western utility companies is considering the development of a small solar-powered facility that could serve as a prototype for peak load applications.

Although solar energy has probably the fewest potential environmental problems associated with its use of any of the major sources of energy, some problems, none of which appear to be insuperable, do exist. Collecting surfaces absorb more sunlight than the earth does, and while this is not likely to alter the local thermal balance in household or other small-scale use, the larger expanse of collecting surface in a central power plant might. Thermal pollution will also be a problem if watercooled turbines are used-indeed, more so than with nuclear power plants because solar installations are expected to have even lower thermal efficiencies. If waste heat is returned to the atmosphere, it could help to restore the local thermal balance. The effects of small changes in the thermal balance would depend on the local meteorological conditions, but are expected to be small. The lack of particulate emissions

or radiation hazards might allow solarthermal power plants to be built close enough to towns or industrial sites so that their waste heat could be put to use. Finally, like other industrial facilities, large-scale plants would also carry some risk of accidents, with the attendant possibility of leaking heat transfer or storage media into the environment.

Yet another option for generating electricity with sunlight is direct conversion by means of photovoltaic cells. But the cells available now-which were developed for space applicationsare relatively inefficient and very expensive to manufacture. As a longterm prospect, however, both cadmium sulfide and silicon cells are attracting considerable attention. This option, and the bioconversion of sunlight to fuels, will be discussed in future articles.

Space heating and cooling with solar energy are not available today. Solarthermal power plants have yet to be built on any but the smallest scale, and key elements of the necessary technol ogy have not been adequately demonstrated. But both options appear to be close enough to practical tests of their economic feasibility to warrant increased efforts. The ancient dream of power from the sun may not, after all, turn out to be impossible.

-ALLEN L HAMMOND
References

1. Proceedings of the United Nations Conference on New Sources of Energy, Rome, Italy (United Nations, New York, 1961).

2. R. Tybout and G. Lof, Natur. Resour. I. 10, 268 (1970).

Solar cells that generate electricity by

ENERGY means of photovoltaic

processes are the predominant source of power for space satellites. Because such cells convert sunlight directly into electricity without an intermediate thermodynamic cycle, and because sunlight is a large and inexhaustible resource, this technology is an inherently attractive source of power. At present, solar cells are not competitive with other means of generating electricity for terrestrial use, but their long-run potential has attracted increasing attention. Recent improvements in cell fabrication and manufacturing methods have stimulated novel proposals for utilizing solar energy on a large scale. Among the applications being studied are the solarpowered house-which some proponents believe could be a reality within 10 years-centralized generating station, and, as a more distant possibility, large orbiting power stations that would transmit energy back to earth.

Of the difficulties that stand in the way of terrestrial applications, the key problems are those of reducing the cost of solar cell arrays more than a hundredfold, increasing their useful lifetimes, and developing methods for the storage of energy. The relatively small numbers of cells produced for spacecraft are manufactured by costly batch processes and assembled by hand. Radiation damage in space or corrosion by humidity and other environmental agents on earth typically degrades the performance of unprotected cells well before the 20 to 30 years expected of power plants, and there is general agreement that encapsulation of cells in glass or plastic will be necessary. For terrestrial applications, the rotation of the earth and the uncertain

ties of cloud cover make solar energy an intermittent resource, and, as a result, some must be stored for use when the sun doesn't shine. Space power stations do not suffer from the intermittent availability of sunlight, but for them to become feasible, the cost of transporting the components into orbit will have to be greatly reduced.

Nonetheless, there is considerable optimism among those working on direct conversion of solar energy, which has the advantage, they point out, of avoiding virtually all of the environmental contamination problems associated with other sources of power. New and unfamiliar technologies are not required-photovoltaic cells were among the first semiconductor devices to be developed. But in recent years, according to a report by the National Academy of Sciences (1), there has been very little support for research aimed at improving solar cells and also no financial incentive for major industrial development efforts. The report concludes that the efficiency with which existing silicon solar cells convert sunlight to electricity, about 13 percent, might well be increased to 20 percent.

would have to be inverted to alternating current before distribution to the

consumer.

Silicon solar cells are made from single crystals of the material, and production of these crystals has always been a costly and awkward process. In a promising development, however, continuous ribbons of silicon have been grown at Tyco Laboratories in Waltham, Massachusetts. In the Tyco process, which was originally developed for continuously growing single crystals of sapphire, the ribbon is formed from a die that controls the shape of the emerging crystal. According to A. I. Mlavsky of Tyco, the process can already produce crystals at the rate of more than 2 centimeters per minute and could easily be automated to produce 100 crystals simultaneously. Although the process works well for sapphire and has produced silicon crystals with very few dislocations, the silicon is not yet of high enough quality for solar cells. The problem is that molten silicon is a highly reactive substance, very nearly a universal solvent, and it appears to be dissolving part of the die and thus introducing impurities into the silicon. A search for better die materials is continuing, and, if successful, could effectively lower the cost of silicon solar cells.

Silicon cells have been the mainstay of space power systems, but cadmium sulfide and gallium arsenide cells have also been developed and tested. In operation, positive and negative charges A second significant factor in the are generated within the cells by the cost of silicon solar cells is that of the absorption of solar photons. The raw material. Although silicon is an charges diffuse across the cell until they abundant element and is available in either recombine or are separated and metallurgical grade at $600 per ton, collected by an electrical inhomogene- the cost of the extremely pure material ity, typically a p-n junction between two needed for cell manufacture is 100 semiconductor regions. Existing silicon times higher, and its production concells develop about 0.5 volt, so that sumes large amounts of energy. Aslarge numbers of cells must be ar- sembling of solar cells into large arrays ranged in series to achieve high volt- is also expensive and not easily autoages. The output is direct current, mated. Although some have speculated which in a practical power system that costs of silicon solar cell assemblies