To me, it raises the question, Where does the power for the cells come from?
After all, it's not like we can violate the laws of physics and, mirabile dictu
, simply have power out of nowhere (unless God is providing the energy, in which case I apologize).



Put differently: These cells are neat, but they'll still need power to infuse them (which they then transfer to the rest of the vehicle). That means generating stations, which means coal, oil, or nuclear, unless someone miraculously makes solar power efficient.



So -- why the hype?

The people proposing them claim that

the hydrogen will come from electrolysis

of water from, eg, wind turbines.



A fantasy.

Is there any doubt that in time, nuclear power will be our principal source of energy? It is bound to be both cheapest and cleanest.

cold fusion nuclear power no less

In the very long run, we have to go solar. You

cannot beat the 2nd Law, and even nukes

create waste heat that cannot be absorbed

indefinitely.



This can be put off for several generations

although there seems no good reason to put

it off as long as possible.



But we'll never get solar until we silence all

its nitwit enthusiasts now and replace them

with -- close your eyes now, Orrin -- rationalists.

Actually, the fuel source is going to be natural gas. The infrastructure largely exists and is significantly safer than pipelines of hydrogen snaking through our cities.



But fuel cell powered cars are going to take decades to catch on, if they ever do. The few being sold now result from a combination of pr and insane regulation although you might see some hybrid cars catching on next decade.



What is more likely to happen with fuel cells, indeed, what might already be happening if the industry hadn't been distracted by the idea of two fuel cells in every garage, is the spread of large fuel cells in factories for industrial self-generation. This will then spread to commercial space, office space and finally homes. But all this requires much higher fuel prices than we now have and there's no reason to believe evergy costs are going to skyrocket any time soon.

We've been trying to make solar power doable for decades, with minimal improvements. I know that inductive reasoning isn't necessarily the right way to go, but put it this way: My father's Ph.D. thesis, in the late 60s, was on moving solar power from its then abysmal 15% energy transfer rate. As I understand it, we have advanced 3% since then.



As for natural gas, even if it's cleaner, when do you think the greenheads are gonna catch on that the power doesn't come from the sky?



And I don't see why we can't just skip nuclear and go right to impulse.

Real solar power will come from orbit, where you can build very large structures cheap enough that the conversion effeciency isn't an issue. Also, there won't be any solar cells - electricity will be generated with a turbine system, using water or nitrogen as the working fluid, dumping waste heat into the dark sky. Power transfer to the ground will be by microwave, which have a very high conversion effeciency (or maybe we'll have orbital towers by then).

I'm skeptical of orbit to surface energy transmission, for a whole host of reasons (defraction, dissipation across the atmosphere, and the energy cost of making that jump anyway). I'm really not sure how efficient (or practical) that idea will be for another several decades.

Wow, a technology debate!



Harry: I don't see why you think the 2nd law is a barrier to nuclear energy - it applies to every energy source. Nuclear doesn't create any more waste heat than other sources - less in fact, because (a) the heat is physically concentrated and so easier to convert to electricity, and (b) hotter (sun only 6000 degrees) which means it can be converted thermodynamically more efficiently. And if it did, the earth would warm and radiate it into space. Energy usage would have to rise dramatically to be a noticeable contributor to global warming.



Against solar, the solar constant at the Earth's radius is only 1366 W per sq meter. This means you need a vast collection area to meet our power needs. It will hardly be cost-effective against nuclear.

Just a note on AOG's scheme - he's generating electricity in a heat engine between the solar-heated gas and a reservoir that cools by radiation. The solar-heated gas cannot be hotter than 6000 degrees, the temperature of the Sun, and will probably be substantially colder, hurting efficiency; but suppose it is 6000 degrees. Suppose the cold reservoir is about Earth temperature, 300 degrees (hard to get it substantially colder). Since the rate of cooling scales with the 4th power of the temperature, the cold reservoir would need a radiating surface area approximately (6000/300)^4 greater than the solar collecting area, which is only collecting a kw per sq m. Thus total surface area of collectors/radiators is ~160 sq m per kw. This is before other losses, e.g. transmission to earth.

That's why the collectors will be in solar orbit. After all, almost all solar energy is now wasted from our point of view. Moreover, the collectors would be useful in regulating the amount of light that gets to Earth, if global warning is a problem, or can be matched with big mirrors to reflect more light Earthward, if we head into global cooling.



PJ is right, of course, that the transmission problems are huge (actually, the only part of this scheme that isn't possible with current technology), but let's just posit some material that is an incredibly strong filament with a diameter of about 100 picometers that is also a superconducter in vacuum. QED.

The 2nd Law problem isn't global. It's

going to be local. Already is, in some

places.



The orbital solar program doesn't evade

that problem.



And you are right about the rather low

density of solar input. That means another

strategy will have to be to continue to

reduce the amount of energy needed for

particular processes.



Great progress has been made on this

front over the past generation, without

getting any attention.

David - you can't run a wire to orbit - even if the satellite were in geosynchronous orbit (22000+ miles) so the wire isn't wrapping around the earth, the tensile force would snap any wire.

PJ: Well, I know when I've been outclassed. I follow you, but I'm not gonna leap ahead of you.



David: I'm not sure exactly what material would have those properties, but I understand that we're positing here, not taking a look through the Merck Index
. So let me make a small observation: Given the Earth's rotational speed, orbital speed, etc., and given that we're basically tethering this mirrored monstrosity to the Earth's surface, and given that we can set an imperfect (but close) geosynchronous orbit, and (finally) given that we're worried about orbital decay, orbital drift, etc.: (Deep breath) Wouldn't there have to be a massive directional apparatus in place to keep the station from whipsawing around? And wouldn't that consume, just in the moment-to-moment corrections, onboard computer, etc., a good bit of the energy it's "generating"?

Ok, PJ beat me to the punch (he got his in while I was typing), and I don't want to seem like I'm piling on, so please don't take my post as just more of the same.

Of course, since 100 picometers is the diameter of a single atom, David must have had his tongue in his cheek when making that post.

Is an atom really that large (1 * 10^-10 m)? I never really thought about it, I suppose.

What you're saying, if I understand you, is that a space elevator
is impossible. On the contrary, I think there's at least one chance in ten that a space elevator will be operating this century.



However, I'm talking about something slightly different: putting huge solar collectors in orbit around the sun for one or more of the following three purposes: collecting sunlight that would not otherwise come to Earth and sending it to Earth as usable energy (in other words, positioning the collectors so that they are not between the sun and Earth); putting collectors between the sun and Earth to shade the Earth and reduce global warming while generating energy; or putting a mirrored fabric in solar orbit and directing additional sunlight to Earth to counteract global cooling. The second two are possible with current or near current technologies. The first requires some way of getting the energy to Earth orbit before it is transmitted to Earth, either on wires down the space elevator ribbon or by microwave.

Yeah, I was actually thinking of the old "assume a can-opener" economics joke when I wrote that. But I've always thought of 10 picometers as the diameter of an atom. Was I wrong?

And don't get me started on Dyson spheres.

A hydrogen atom is approximately 106 picometers
in diameter.

A space elevator? I hadn't even thought of that since the last time I played Civ II.



Ok, I just reviewed the plans on that site, and I'm thinking something along the lines of, Did NASA know that their grant money was going for crack cocaine purchases?
My degree was hardly in high-energy physics, but phrases like "Launched in pieces, it will assembled in LEO and will then begin deploying the ribbon down to Earth" make me wonder who was smoking what when they came up with this idea.



I can't imagine that a space elevator would be practicable, even if it's possible.



As to the rest, with all due respect, I don't see any of that as remotely possible. Take your first option: Assuming arguendo we could deploy the collection apparatus around the sun, transmitting the energy across space in the form of a microwave seems, at first blush, no more effective than just letting the sun send it in the form of heat, light, and the rest; and sending it by wire is science fiction.



As for the second two, the logistics are mind-boggling (and that's without getting into questions about dark matter, comets, asteroids, and space debris ripping holes in these collecting materials).



As an additional reservation, we know way too little about the climate to go (trying to) actively mess with it.



Final point: I mean no disrespect; I simply think you're wrong. And in case someone else posted while I was (slowly) typing this out, I don't mean to pile on.

Christopher-



You're being way more solicitous than you need to be. Pile on, this is fun. Just think of me as a jackass ([INSERT JOKE HERE]): you need to hit me over the head with a two-by-four just to get my attention.



As for space elevators: Arthur C. Clarke hasn't been wrong yet.



Putting things in solar orbit isn't particularly tough. In a sense, everthing we've ever put in space has been in solar orbit. Energy gathering systems could be put into position at 1 AU, just slightly ahead of or behind Earth (although this is obviously not the most efficient placement.) In fact, think of the moon as a satellite that reflects sunlight towards Earth and even, at predicable intervals, blocks sunlight from reaching us.



I agree that taking energy in as light is the best return on investment, because there's no investment. But my point is that almost all of the energy put out by the sun misses Earth entirely, so if we can figure out cost efficient transportation of energy across space and, a seperate question, through the atmosphere, we will be able to access completely new energy that would not otherwise reach Earth.



As for not fooling around with the climate because we don't understand it, amen, brother. But what I have in mind is, for example, if we unmistakably enter a new ice age (one is due any moment).

Boy, go back to work for a while and miss all the fun.



I'll deal with Mr. Badeaux first. Yes, it will probably be several decades before space based solar power is feasible. I thought we were discussing this issue on that kind of time scale i.e., 100 or 200 hundred years out. There are some losses for transmission, but again the key point of orbital structures is that it if you can build them at all, it's very cheap to make them very big, so it's not hard to make up for such losses just by building bigger. As for orientation problems, the power stations don't rotate in the solar reference frame, they always point at the sun.You may need some tricky bits to connect them to the wire to the ground, but you can use short range microwaves for that if you want, or beam straight to the ground. Currently we can build microwave power connections with about 85% effeciency, which is plenty good enough for this application. The combination of that and the ability to build larger collection surfaces makes space based solar power far more attractive than trying to do it on the ground (without even considering weather related issues). As for space elevators, there's lots of real research on the topic and it certainly doesn't look fundamentally impossible.



PJ; Space based solar power may well be better from a 2nd Law point of view because it can be set up to intercept energy that would have arrived on Earth anyway, and not create additional
waste heat. You also complain about the low value of the solar constant, but as mentioned that's why you build it in space. Also, the cold reservior is about 4 degrees Kelvin - that's what I meant by "the dark sky".



Mr Cohen; One can collect non-Earth incident solar energy without putting the power stations in solar orbit by putting them in polar orbits around the Earth, or even the Moon. In the former case, there will be some attentuation of Earth incident solar energy, but larger orbits can greatly reduce that.



Finally, on the time scale of decades, we may well get to the point where there's little need to ship the energy back to Earth as most of it may be consumed in orbit and the manufactured goods shipped down instead of raw energy. This is where my environmentalism kicks in - why worry about clean industrial processes when we should really just move them off the planet entirely?

With the caveat that we're rapidly passing beyond my minimal expertise in the subject:



AOG: Hi. The wire-to-ground idea is impractical -- at least within the foreseeable future -- for the reasons PJ and I enunciated. My big objection is, aside from the enormous stresses on the wire from the earth's gravity and internal atmospheric events, which is to say, put those to the side, you still have to have these apparatuses (apparati?) use a significant fraction of their energy absorbed on managing their orbits. But assume we don't bother with wire: How much of that microwaved energy is going to dissipate across the atmosphere, ionosphere, etc.? And the amount of energy you're talking about sending in discrete packets would, it seems to me, raise the danger of altering local
climate (or at least, long-term weather) at some level or another (superheat some wind, etc.) through simple energy transfer. (See "liability" below.)



The space elevator may very well be doable. Like I said, this isn't my field of (ha ha) expertise. But on the face of it, as they're currently designing it... well, it looks like an incredibly expensive project, at best, and there are huge liability issues wrapped up here (now we're in my specialty). I'll spare you the gory details, but let's just say you don't want to be the company that "loses" a manufacturing shipment in what used to be a teeming suburb.



I will concede that, assuming arguendo we can get everything spaceside, your points about power collection and building seem valid.



DC: Thanks for the kind words.



Putting something in solar orbit (I presume on the same plane as Earth) does not necessarily mean putting it at the same speed and orbital time ("year") as Earth. Even assuming arguendo that you put these bad boys on the same orbital path as Earth (to the "sides" of old Gaia, one supposes), I still don't easily see how to get the energy where it needs to go; and given how large those collectors would have to be, it seems likely they'd take a fair-sized pounding over time from space debris of various sorts.

AOG - (1) The solar constant is already above the Earth's atmosphere -- that's the most you can get without moving closer to the Sun. (2) You have to dissipate waste heat by radiation from a blackbody to the 4 degree sky - the power you radiate rises as the 4th power of the temperature - if your cold reservoir (the black body) is at 4 degrees, you are dissipating NO waste heat at all, absorption equals radiation. If you make the blackbody warmer, e.g. 300 degrees, you radiate more away so you need less blackbody surface area. (3) By the way, I wrote too quickly, for a 300 degree black body and a 6000 degree hot reservoir you need a dissipation surface area of 160 sq m per watt of energy received from the sun. With conversion and transmission efficiency of 10%, you need 1600 sq m per watt delivered to earth, so a single light bulb on earth requires 160,000 sq m of dissipation surface. If the useful life of this satellite system is 30 years (= 260,000 hr), then 1600 sq m produces 260 kw-hr over its useful life. If nuclear power costs 3 cents/kw-hr, then you have to put the 1600 sq m in orbit for less than about $3 to break even.

And with that, we're officially beyond my minimal expertise.

Although I should point out that $3 per 1600 m^2 may be possible in a long-run economy of scale. Massive losses in the short run, steady, low profits in the long. But I'm speaking without hard data here.

I might add - the world's electricity production capacity was 2 million MW in 1993; replace gas by fuel cells, add in some growth in usage, call it 5 million MW in a few years. At 1600 m^2 per watt, that's 8x10^15 m^2, or roughly 10^10 km^2 of surface. That's about ten times the surface area of the Earth.



It could be done more efficiently with a warmer cold reservoir, but there are various engineering difficulties too. In short: I don't think solar power will ever be significant.

AUTHOR: Annoying Old Guy
EMAIL: aog@thought-mesh.net
IP:
URL: http://blog.thought-mesh.net
DATE: 01/09/2003 05:59:00 PM
AUTHOR: Annoying Old Guy
EMAIL: aog@thought-mesh.net
URL: http://blog.thought-mesh.net
DATE: 1/09/2003 05:59:00 PM

Ok, I found the equation and the right constants. P/A (power per unit area) = sigma * (T^4 - S^4) where T is the temperature of the radiator and S the temperature of the sink. We take S to be 4K and sigma is .567*10^-8W/m^2-s. From this, we get that at 300K a surface radiates 460W/m^2 and at 400K 1453W/m^2. So if the cold side is 400K and I have 10% conversion effeciency then I can still push ~150W per square meter of radiating surface to the ground. For your 5 terawatt number for the entire planet, that's only 30 billion square meters which is a square 182 km per side. Big, but not impossibly huge. And I think the efficiency will be much better than 10%. The numbers don't support your scepticism.

AOG - You're right on the main point - I conflated two numbers, solar constant is at 1 AU and 6000K is effectively solar constant at sun's surface, so I had an extra factor of (1 AU)/(solar radius) squared. Switched too quickly from the thermodynamic issues to the energy ones and confused myself, I guess.



Looked quickly at your numbers - sigma * T^4 is 46 W/m^2 at 300 K, not 460; at 300 K and 10% efficiency you need 10^6 km^2.



The radiating surface has to be a blackbody on the side away from the sun, aluminized mylar won't do.

PJ;



You're still off.

sigma=5.67E-8 W/m^2.

300^4 = 81E8 = 8.1E9

sigma * T^4 = 5.67E-8 * 8.1E9

= (5.67 * 8.1)E1

= 46E1 = 460



I'd expect a number in this range because of the Earth being able to cool itself at around 300K. The effective radiating area of the Earth is 4 times the collecting area, so you'd need about 1/4 of the solar constant in radiating energy at 300K which is close to 460 W/m^2.



Also, note that we don't have to cool the mirrors because as long as they don't melt they work fine at any temperature and have no effect on the conversion efficiency of the turbine system. Because of the T^4, even a much lower emissitivity is compensated for with a small rise in temperature. Only the cold reservoir requires cooling. How much we need there depends on what part of the inefficiency is due to the mirrors vs. the turbine.



So, it looks physically feasible. I am not claiming that we'll do this next week, but in 100 years? Easy. It'll be easier, cheaper, safer and more ecologically sound than any alternative.

I was using your sigma - .567x10^-8 in the post below mine.



Most of the waste heat not converted to microwaves is going to go to the cold reservoir.

PJ;



Ah. I'm sorry, I must have typo'd the post. Your way of stating the energy that needs to be dumped is a good one - expressing the cooling surface area required in terms of delivered watt / m^2. Even at 10% for the turbine (which a very inefficient turbine) we still get 50 W delivered / m^2 of cooling surface. I think it's reasonable to assume that all of the waste heat goes into the cold reservoir - the amount radiated from the turbine and infrastructure will be negligible.



The one thing I don't know is, can we do better than 4piR^2 for a radiating surface? I think so but I'm not willing to bet on it. This makes a big difference in how much it costs to get a square meter of radiator.



Finally, there was a comment on the local climate effects from beaming the energy down. I don't consider that an issue because of the high conversion efficiency of microwave rectennas, which is significantly higher than any existing large power plant technology. So a receiver for space power will dump less heat into to the local environment per watt delivered to the consumers than existing power plants.

The best radiating surface would be a sheet, not a sphere, that is mirrored on the side toward the sun and blackbody on the side away. Consider using two mirrors, telescope-like, to focus solar radiation onto the hot side of the heat engine and then the radiating surface can be on the back side of the solar-collecting mirror.