Mrs. Malaprop

Mrs. Malaprop raised the inappropriate use of vocabulary to a fine art, and Archie Bunker (TV's All in the Family) brought the art into millions of homes. Malapropisms (or more recently, Archie Bunkerisms) aren't limited to humorists, of course. For example, students are forever trying to use memorized free-fall equations when they're dealing with a mass oscillating on a spring. (That passes for humor only among physicists.) 

For a surprisingly long time, it has been the case that the speed of computer chips doubles every eighteen months. This trend, called Moore's Law, has several ramifications, among them that every eighteen months the price of each transistor on an integrated circuit will be cut in half. But, to apply Moore's Law to photovoltaics is to engage in Mrs. Malaprop physics. Moore's Law does not apply and will not apply to photovoltaics. 

Moore's Law, unlike laws of physics, merely represents a trend --- though a surprisingly consistent one. Every eighteen months the speed of calculations doubles, the amount of memory available in a commercial chip doubles, and the price of a transistor halves, to name a few consequences. 

There is no underlying physical cause relating to the eighteen-month time frame, but the relationship between the other quantities is easy to understand. Manufacturers, through fierce competition and uncanny creativity, learn how to pack more and more transistors onto silicon wafers by making the transistors smaller. 

The tight packing of transistors offers a quadruple benefit: signals don't have to travel far, so calculation time is reduced; the cost per transistor is smaller because of the smaller size; the amount of memory (using transistors, of course) that can be placed on a small chip of silicon is increased; and power requirements are reduced. (Yes, "powerful" computers require less energy to accomplish a given task than earlier, less efficient ones.) But, it is a mere accident that all of this doubling and halving takes place in a given constant time and further accident that the characteristic time is eighteen months. 

Al Gore (Earth in the Balance, 1992) tells us: "But this technology [photovoltaics] is in its infancy, and what is required _ as part of the proposed SEI [Strategic Environmental Initiative] _ is a global effort to accelerate the development of cost-effective photocells," implying a precipitous drop in price. If he had ever heard of Moore's Law, he would surely have cited it. 

The Vice President assures us that, "The technical obstacles to developing them are becoming less important than the political and institutional barriers." Hmm. Who, pray, offers political barriers to photocells? Nobody. What are the institutional barriers to photovoltaics? None. Gore lists none, of course, for there are no such barriers at all. Most likely, he simply wants huge subsidies for such piddle power. We can only hope there are barriers against the utter waste of tax money for Gore's nonsense projects. 

Harvard Professors Stobaugh and Yergin (Energy Future, see The Energy Advocate, Aug. 1996) have done a tiny bit more homework than Al Gore, just enough to make their nonsense sound credible. To wit: "But what happened to the semiconductor industry is now happening to photovoltaics. ... The continued reduction of costs could open up a vast market ... As one semiconductor executive explained, `Eighty percent of the cost of photovoltaic systems is in the materials, and eighty percent of the material cost is in the silicon.'" 

The entire system must also include hurricane-resistant structures to hold the photovoltaic collectors, but in the fog-bound world of Harvard's prognosticators, this issue does not rear its ugly head. 

But that's just a minor issue. The Archie Bunkerism lies in applying "what happened to the semiconductor industry" (Moore's Law) to photocells. The reason for the dizzying decline of transistor prices is the equally dizzying reduction of their size. But in photovoltaics, the object of the game is to collect sunlight. That takes real estate. It makes no difference whether there is but one photocell or a billion on the wafer. It is the collection area of the wafer that matters. In short, Moore's Law doesn't apply here. 

Al Gore does get something right, though unwittingly, when he refers to photovoltaics as "small flat panels of silicon or similar materials that are designed to produce currents of electricity." [emphasis added] He merely forgot to explain how small panels collect a lot of sunlight. 

Rumor has it that Gore is planning to remove the solar cell from his calculator to power Air Force One.