Efficient food production
Mar. 9th, 2005 02:11 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png)
I was noodling around on soc.history.what-if and made a calculation I'd never bothered with before: if a human needs enough food to produce N Watts, how many square meters are required to intercept that much sunlight? OF course I was too lazy to actually look up insolation for various latitudes but the BOTEC I committed seemed to show that it should be a few square meters.
Even Fairbanks, Alaska, gets from 90 to 350 watts/m^2. Say your mark 1 human needs at least 100 watts worth of food to keep functioning [1]: They'd need about one square meter dedicated to collecting solar powers, asssuming no losses. The entire population of North America should require a few hundred to a thousand square kilometers of converters to power themselves. Even a factor of ten losses should mean that we'd need about 300 square kilometers to feed all of Canada, assuming the lowest insolation in Alaska is what we have to work with, and about 3000 square kilometers to feed all of the USA. That's a square less than 20 kilometers on an edge for Canada and a bit over 50 kilometers on an edge for the USA. Feeding the entire planet should require about 60,000 square kilometers or a square about 250 km on an edge (or less, if we pick someplace sunnier than Fairbanks to grow food).
Clearly modern methods of coverting solar (and fossil) energy into human energy are criminally inefficient.
1: Googling says "at least 2500 kilocalories" per day so call it 4000 to be safe. That works out to about 50 watts, which I will double just because.
Even Fairbanks, Alaska, gets from 90 to 350 watts/m^2. Say your mark 1 human needs at least 100 watts worth of food to keep functioning [1]: They'd need about one square meter dedicated to collecting solar powers, asssuming no losses. The entire population of North America should require a few hundred to a thousand square kilometers of converters to power themselves. Even a factor of ten losses should mean that we'd need about 300 square kilometers to feed all of Canada, assuming the lowest insolation in Alaska is what we have to work with, and about 3000 square kilometers to feed all of the USA. That's a square less than 20 kilometers on an edge for Canada and a bit over 50 kilometers on an edge for the USA. Feeding the entire planet should require about 60,000 square kilometers or a square about 250 km on an edge (or less, if we pick someplace sunnier than Fairbanks to grow food).
Clearly modern methods of coverting solar (and fossil) energy into human energy are criminally inefficient.
1: Googling says "at least 2500 kilocalories" per day so call it 4000 to be safe. That works out to about 50 watts, which I will double just because.
no subject
Date: 2005-03-09 08:50 pm (UTC)Googling for "photosynthetic efficiency" will give you various tables, but be careful with the definitions they use. The 8% figure for sugarcane can also be found in Lovelock in his discussion of C3/C4/CAM plants, if memory serves. I can dig up more from Hall & Rao later, if you'd like.
Charlie, I've never seen a Drexlerian design that didn't look like a high explosive about to go off. (Merkle's are a little better.) Poor guy just isn't a chemist, that's all.
Carlos
Hasn't Drexler been more or less marginalized?
Date: 2005-03-09 09:02 pm (UTC)It's a shame the term "nanotech" has been irredeemably tainted by the more extreme proponents. I don't really want to use the term "biochemistry" because what if I'm talking about something designed with real biochem in mind but not using anything that is actually used on Earth?
Re: Hasn't Drexler been more or less marginalized?
Date: 2005-03-10 12:18 am (UTC)drexler is largely marginalized by serious chemists, but afaik, he still has a large following among nano-ninnies and random lay people.
Re: Hasn't Drexler been more or less marginalized?
Date: 2005-03-10 12:44 am (UTC)(There's a whole strange subclass of MIT-affiliated whoopsies like that. Norbert Wiener's wife claiming McCulloch seduced their daughter, setting back cybernetics a decade. Minsky and Papert stomping on perceptrons, setting back neural net research a decade. Chomsky and linguistics, setting back linguistics two, three decades. Various Media Lab things.)
For a neologism, might I suggest "artificial biochemistry"?
Carlos
no subject
Date: 2005-03-09 11:19 pm (UTC)Re: Hall and Rao's Photosynthesis, p.67 of the fifth edition shows efficiencies of up to 12% in certain wavelengths in Chlorella, but the operative word is "in certain wavelengths". Hall and Rao are all about the photosystems and the chloroplasts, but that's not a measure of the final productivity of even an ideal farm.
no subject
Date: 2005-03-10 03:32 am (UTC)and page 4:
These are cumulative, multiplicative losses. About half is simply because the photons are not energetic enough to make the reaction go (the energy = h * frequency thing I alluded to before). After that, the coupling from the reaction center to carbon fixation. The confusion RuBisCO makes between CO2 and O2 is due to the similar charge and size of the two molecules, and is largely insuperable.
8% efficiency, incidentally, is the maximum rate for sugarcane under cultivation. For its full life-cycle, it's more like 4%.
Carlos