Semi-OT:The trendiest development in solar energy of late has been to incorporate photovoltaic collectors into the skin of buildings, integrating them into the structures overall energy system and the structure itself. The advantages are self-evident, but there also are drawbacks—the flat panels are large; they have to be unobstructed; and so on.
An alternative to passive collectors and currently experiencing a renaissance is the solar concentrator. In this kind of system, sunlight is magnified and focused by a variety of means onto a small photovoltaic panel to produce intense heat and light, generating electricity for distribution to a community, a building complex, or over the grid.
Following the 1973 energy crisis, much work was done on thermal systems. Typically, a concentrator was combined with a Stirling engine to produce electricity. There also have been government supported experiments with land hungry systems, in which light from mirrors scattered over acreage the size of a small farm is focused on a single target.
In the I 980s and ‘90s, with the drop in oil prices and consequent decline in government support for far out energy research, such approaches fell from favor. Now concentrators are making a comeback, largely due to steady and impressive advances in the production of high efficiency photovoltaic material for space applications. The materials are too expensive to use in large flat panels, but are relatively economical in concentrator systems, which need smaller quantities.
Causing the most excitement at present is a three junction, gallium-indium-phosphide gallium-arsenide germanium cell grown by Spectrolab Inc. of Sylmar, Calif., a subsidiary of Hughes Electronics Corp., El Segundo, Calif., and a leading supplier of photovoltaic material for highpower satellites. The National Renewable Energy Laboratory (NREL), in Golden, Cob., put a GaInP/GaAs/Ge cell through a battery of tests and, after adding metallic contacts and an anti-reflection coating, found it to be 32.3 percent efficient. In contrast, the highest efficiencies for single crystal silicon cells are in the mid 20s, and for thinfilm amorphous silicon or cadmium-telluride cells,they are in the mid-teens. Both are among the materials popular for flat panels.
The NREL officials who have tested the Spectrolab material expect still higher efficiencies to be attained. Moreover, ,the CamP/GaAs/Ge cell is considered all but ready for production using standard semiconductor techniques, said Sarah Kurtz, head of NRFLs Ill-V development team.
Interest in the gallium compound is not confined to NREL. Concentrating Technologies LLC, Owens Cross Roads, Ala., is busy incorporating the material into the design of a 2—2.5kW system, which it plans to start deploying this fall at the Arizona Public Service Co.’s STAR test facility in Tempe. The design uses a 14-in2 array of mirrors to concentrate sunlight by a factor of 200—400 (200-400 suns) on a 20—22.5-cmxcm square of the PV material.
(The STAR facility has been testing solar systems since 1988 and by now has looked at or is looking at just about every system type there is, said Herb Hayden, solar program coordinator for the Arizona company. STAR’s test activities got a boost from new Arizona legislation requiring utilities to obtain certain fractions of their electricity from renewables.)
Another company looking at the Spec-trolab material is Solar Systems Pry. The company, in Hawthorn, Australia, has had a lot of experience producing concentrator systems using an optimized silicon cell
made by SunPower Corp., Sunnyvale, Calif. Its current SS-20 system [see photograph-sorry] is a large dish that tracks the sun and concentrates light by up to 480 suns.
Engineering challenges associated with such complex collectors are not trivial. They are hard to install in remote areas are vulnerable to weather extremes, and require some ongoing maintenance. For good performance, sunlight has to be evenly concentrated over the entire sur-face of the cells, because the efficiency of the whole panel is limited by the least efficient of the series connected cells. Provision must be made for cooling the dish or array, as the heat buildup will be intense even in systems with relatively low concentrating factors.
In light of those difficulties, Kurtz sees concentrator systems as more for the future than the present. All the same, she is enthusiastic. In her view, the deciding factor is that these systems, requiring as they do rather little high grade photovoltaic material, can be scaled up rapidly to large production volumes. Suddenly producing far larger quantities of material for collectors, using semiconductor industry materials and techniques, would be far tougher.
One company that has had considerable success designing and building concentrators is Anomix Inc., in Torrance, Calif. It boasts a fully integrated, modular system in which the cooling structure, consisting of tubes and fins, holds the array of photovoltaic cells. Using SunPower’s silicon and concentrating the sun by a factor of 300, it produces 20 kW
Amonix already is deploying 300 kW of its arrays at the Tempe test facility and at a nearby airport, and it hopes next year to install a solar farm with a capacity on the order of 1 MW Like Solar Systems, it sees its technology as near commercial.
Said John Lasich, technical director at Solar Systems: "Combining the simplicity of PV with the leverage of concentration delivers an ideal mix of economy and reliability. This technology is simpler than solar thermal and, at a system level, cheaper than solar panels. With cell efficiencies rivaling heat engines, the day is dawning for PV concentrators."
WILLIAM SWEET, Editor
IEEE SPECTRUM AUGUST 2000
In this 20-kW SS-20 system from Solar Systems, mirrors concentrate light Onto a slab of photovoltaic material supported above them.
(Photo missing here)
Jim |
