A misunderstanding being vigorously promoted by the solar lobby (for example
Al Gore in his book, Earth in the Balance, 1992) is that with government
backing and subsidies to promote development, the price of photovoltaics would
plummet and open up the same kind of vast market for solar cells as has happened
with computer chips in recent decades. Consideration of the different roles
played by the semiconductors in these two applications shows, however, that
such expectations are unfounded.
It has remained a consistent trend for a surprisingly long time that the speed
of commercially available semiconductors and the amount of memory on a chip
has doubled, while at the same time the price per transistor has halved, about
every eighteen months. These benefits all follow from improving technology that
enables transistors to be made smaller, thus fitting more of them onto a chip.
In other words, performance per unit cost increases because more and more of
what computers do--computing--gets done on smaller and smaller areas of silicon.
But the function of solar cells is to collect sunlight; to collect a
given amount, you need a given amount of area. The technology that enabled computer
circuits to shrink to such tiny sizes gains you nothing because you'd need correspondingly
more of them to do the same job.
The only area with room for improvement would be raising the conversion efficiency
of sunlight to electricity. The progress made in semiconductor computing circuits
is irrelevant to this function, and there are reasons for supposing that the
20% or so achieved by the best systems currently is pushing the limit. But even
in the wildest flights of fancy, with absolute perfection and no losses at all,
you can't go beyond 100%.
More information on this appears in the October 1997 issue of The Energy
Advocate newsletter. (For details of EA, see the March 26 posting in the
ENERGY section of the Bulletin Board