Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site ucbvax.BERKELEY.EDU Path: utzoo!watmath!clyde!burl!ulysses!ucbvax!BROWNVM.BITNET!ST401385 From: ST401385@BROWNVM.BITNET Newsgroups: net.space Subject: Solar energy Message-ID: <8603130758.AA00386@s1-b.arpa> Date: Wed, 12-Mar-86 15:01:47 EST Article-I.D.: s1-b.8603130758.AA00386 Posted: Wed Mar 12 15:01:47 1986 Date-Received: Fri, 14-Mar-86 06:08:04 EST Sender: daemon@ucbvax.BERKELEY.EDU Organization: The ARPA Internet Lines: 91 A recent posting from Kenneth Ng quoted some obsolete figures for solar power efficiencies. This happens to be something I know something about, so I present a more complete (and up to date) set of figures here. > from "The Health Hazards of NOT Going Nuclear", by Petr Beckman. > At best, the concentration of solar energy is 1 kilowatt per > square meter. Correct. This is at the earth's surface, at noon, on a panel oriented perpendicular to the incident sunlight, on a clear day. (In space, the energy density is about 1350 kw/m2) Panels that do not track the sun average somewhat less, by a factor of cos theta averaged from -90 to 90 degrees, or 2/pi. If you then add in night, the actual amount of incoming sunlight on a (non-tracking) panel is 1/pi times this, or about 320 watts/meter2. (If you track the sun, the cos correction goes away, but the panels have to be separated from each other to avoid shadowing, so the land area needed increases) > theoretical maximum efficiency is 22%. I don't have any idea where he gets this idea. This is flat-out WRONG. In fact, existing solar cells beat this figure. Maybe he means theoretical max efficiency for a single crystal, single junction, unconcentrated silicon solar cell made using 1970 technology. That IS about 22%. The best figure I have for single crystal, single junction, unconcentrated silicon using current technology is 27-29%. (M.B. Spitzer, PhD thesis, Brown University, 1981; Tiedje et. al., IEEE Trans on Electron Dev., May 1984). Theoretical limit for multibandgap cells is 40-75 percent, depending on how many layers you want to make. > Current solar technology is about 10% efficient. This book must be using data from 1962. Correct data (1985) is: Technology Best efficiency Lab: on sale: Single Cry. Silicon 19% 14% Amorph Si. 11% 5% GaAs 21% Conc. GaAs 26% For reference, single crystal Si is what's mostly currently used in terrestrial arrays. Amorphous Si is used mostly in consumer goods, like calculators; most people think some sort of amorphous technology (not necessarily a-Si) will replace single crystal for terrestrial arrays in a few years, when the efficiency reaches 15 percent or so. GaAs is not currently used for terrestrial applications; I think some recent satellites have used it. > Allowing a 50% spacing between solar collectors to allow for repair > and shadows Seems a little high to me (unless you're talking tracking, in which case it seems a little low), but I'll accept it. >you have a solar power plant 50 square miles big to generate 1000 >megawatts of power. 50 square miles/(0.6**2 sq.km per sq.mi)*1000000 m2/km2 = 1.4E8 m2 Even at your assumed efficiency of 10% * 1Kw/m2 *.67 packing fraction, this is 9000 MW. You made an arithmetic error. More to the point, the average home needs about 2 kW (ave) of electricity. Assume 15% efficiency, 1kW/m2, and include the 1/pi correction, this means 21 square meters of area per home, or an area 15 feet by 15 feet. This will fit on the roof. The big problem for photovoltaics currently is price. Cost for large purchases is currently (1986) $6.00 per peak watt. Installation, BOS, etc. pretty much doubles this cost to $12/watt. Include the 1/pi correction to get average watts from peak watts, and you get $38/watt. Nuclear reactors are currently being built for about $6 per watt, and run about 90% uptime, so the (capital cost) is about $7/watt. (on the other hand, there are fuel and operating costs for nuclear plants which don't exist for photovoltaic arrays. But I think (I don't know) that these costs are small compared to the capital cost.) Many people think that photovoltaics will get down to about $0.80/watt within a decade; in this case it looks very competitive. I'm not sure I believe all of the claims people make, however. On the other hand, photovoltaics is competitive NOW in third world countries. This is because these countries have not made the huge investments needed to set up an electric grid--eg, a set of copper wires connecting any point in the country to any other. Current costs (in the US) of extending the grid are about $10,000 per mile. This is just out of the reach of third world countries, and it is thus economically worthwhile to generate electrically directly on site. I'm not sure what conclusion you can draw from these facts, except that it is, at the current time, not at all clear which of many technologies will win out in the next couple of decades (except to say that oil will definitely lose out, sooner or later.) --Geoffrey A. Landis, Brown University Reply to: ST401385%BROWNVM.BITNET@WISCVM.ARPA