How Seawater Can Power the World
nytimes.com
nytimes.com
I think people believe that fusion is somehow "way more powerful", because they know that fusion bombs are way more powerful than mere atom bombs. Actually, in many of the most powerful fusion bombs, the vast majority of the energy is produced by fission, with fusion serving as only an intermediate step to release neutrons and create more fission. A step you don't need in a reactor, which can control the fission rate itself.
My guess is that a lot of people who haven't looked at the numbers think that fusion is hundreds of times more powerful than fission. Well, let's look at those numbers: the fission of a uranium-235 atom produces 202.5 MeV. The fusion of D+T produces only 17.6 MeV. That's more energetic per unit mass of fuel, since D and T are just hydrogen isotopes, but if you do the math, it's only about 3.5 times more energetic even using this measure. And fuel mass is a tiny portion of the mass of the whole powerplant. To make it worse, hydrogen is far less dense than uranium (by a lot more than a factor than 3.5), so you actually end up with a lot more fuel volume and containment volume. There are other reactions than D+T, but none of them look as good as plain old fission. That's before we even get into the fact that for most fusion reactions, a lot of the energy goes into extremely dangerous neutrons.
The talk about "getting the fuel from seawater" is also misleading. It's technically true, but the process is very expensive, and, if we are going to talk about seawater, uranium can also be extracted from it.
Finally, the idea that the danger from proliferation would go away is also false - anything producing a neutron flux as strong as a fusion reactor can be used to produce isotopes suitable for a bomb. And even aneutronic fusion (which is technically even more difficult) can be made to produce neutrons with suitable additions to the fusion fuels.
EDIT: P.S. I am not an expert. I do think I can understand numbers and technical arguments, and I would love to hear expert analysis. But one thing: for every technology which you assume will help fusion (such as wonder-containments for the neutrons), please analyze how fission reactors would benefit from it, as well. Then let's see if we can make fusion come out "better" than fission in some meaningful sense.
EDIT: I can't resist adding that the talk of "harnessing the reaction that powers the Sun" is aimed at people who don't know the numbers, and is very annoying. The Sun is not a good model; it has a huge mass (duh), and a tiny surface-to-mass ratio, which is why the energy it produces makes it so bright. But pound-for-pound, the Sun produces a lot less energy than slightly warm chicken soup. Do the math yourself, and be surprised. We can not build a reactor that works like the Sun.
The Sun produces 3.84e26 Watts and has a mass of 2e30 kilograms. So suppose you can harness all the energy produced by a kilogram of the sun for 100 years, or 3.5 billion seconds (a human lifetime and then some). You will have harnessed 672000 joules, or 160 dietary calories. Less than a kilo of chicken soup.
The whole point is that the time scale and the mass scale of the Sun are incompatible with human energy production.
And also, I think it's a surprising result - again, ask a layperson who hasn't done the math to tell you how much energy a good kilogram-sized chunk of the Sun produces in 100 years, and he'll probably tell you it's enough to destroy whole cities or boil seas. Actually, it's barely enough for a small breakfast.
http://www.sal.wisc.edu/~khn/ast100/lectures/lecture16.pdf
the energy produced by a kilo of the sun is orders of magnitude higher.
(something like 2.3×10e13 Joule/Kg).
Of course it doesn't produce it in a measly 100 years (more like 100.000).
At one time, only a very tiny fraction of the sun is being processed in it's fusion reactor at the core ( The core, inside 0.24 solar radius, generates 99% of the fusion power of the Sun. From http://en.wikipedia.org/wiki/Solar_core ), and that's what people think about when talking about the sun's energy (the 2.3×10e13 Joule/Kg thing, not the 672000 one).
And yes, that does make things more difficult :) .
Look at this another way: ask a layperson, who consumes 3000 kcal a day approximately, how much of the Sun's mass would be required to supply him with this energy input. The answer, unless I mislaid a zero, is 755 tons or so.
Now, I will absolutely be a monkey's uncle if laymen give you an approximately correct answer. My guess is that people will say that it takes a tiny drop of solar material, or a microscopic amount. Can someone try this?
The density of the Sun is approximately 2 * 10^-4 kg/m^3 in the photosphere ( http://en.wikipedia.org/wiki/Sun ). We're ignoring the core for now, because that's at more like 15000000 K, so. A kilo thus takes up 5 * 10^3 m^3, which is a sphere with radius ~12 m, having a surface area of ~1.4 * 10^3 m^2. The power radiated by a kilo of plasma from the Sun's photosphere is thus is j' = area * sigma * temperature^4, where sigma = 5.67 * 10^-8 is the Stefan-Boltzmann constant ( http://en.wikipedia.org/wiki/Blackbody_radiation#Stefan.E2.8... ), so j' = 1.4 * 10^3 * 5.67 * 10^-8 * (5 * 10^3)^4 = 4.96 * 10^10 watts, or well over a person's daily caloric intake every second (In fact, this is approximately the energy consumption of a small country).
The power radiated from 1 square meter of the sun's surface (63 megawatts) comes from a cross section of the core, far below, that is 250,000 kilometers deep!
If we're going to analyze the performance of our "reactor", then we need to include the fusing fuel, the "confinement", which prevents us from being very quickly destroyed by rather nasty fusion products, and the photosphere, which is very beneficial to us.
I continue to claim that if you ask laymen about the mass of their "personal portion" of this reactor, you will get answers that are about 8 orders of magnitude smaller than in reality. I actually tried this just now, with expected results ("some kind of a tiny pinhead thing"). There is no point to this exercise, other than to show that people have no intuitive vision of these scales; it doesn't make them stupid, the scales are just unintuitive and outside of their experience. I don't claim to have discovered anything about astrophysics.
Well, you also clearly weren't aware of the point I was making: if you ask a misleading question, you get a misleading answer.
>I continue to claim that if you ask laymen about the mass of their "personal portion" of this reactor, you will get answers that are about 8 orders of magnitude smaller than in reality. I actually tried this just now, with expected results ("some kind of a tiny pinhead thing"). There is no point to this exercise, other than to show that people have no intuitive vision of these scales; it doesn't make them stupid, the scales are just unintuitive and outside of their experience. I don't claim to have discovered anything about astrophysics.
The 'laypeople' are basically right, though: if you actually extract the energy from a bit of thermal plasma it only takes a "pinhead" to supply energy for an entire human life. Obviously the energy being actually emitted is much smaller, just as there is no light emitted from a box with a bright light in it, but that doesn't mean that the light in the box couldn't easily light up a room. Like, the amount of 15000000 degree plasma at the core of the Sun that could destroy a small city if it were released is probably no larger than my fist.
>rather nasty fusion products
There really aren't any seriously nasty fusion products -- tritium and positrons, the latter being instantly destroyed and the former already present in the atmosphere.
The power produced by nuclear fusion inside the Sun turns out to be very small compared to the mass of the Sun. I don't claim that this is some kind of a discovery, and I am sure that this is obvious and intuitive to astrophysicists, but it's not intuitive to laymen, who think that the Sun emits a lot of power relative to its mass. That's not misleading, it's only unintuitive.
In their immediate experience, anything this bright is indeed producing tens of thousands of watts per kilogram at least.
And I never doubted that you would get a whole lot of power if the energy contained in the Sun was released on a human time scale; this is something where intuition serves us well.
The Sun is a self-balancing equilibrium. Gravity causes the matter to collapse, which increases heat and pressure at the core and causes matter to fuse, which releases energy that pushes matter away from the core. As such fusion only occurs at the rate needed to maintain this equilibrium, and that rate is less than you'd think intuitively.
Not only that but our Sun is relatively small and weak. Type-O stars, for example, can have a luminosity a million times greater than our Sun's.
My secondary point is that people do not have an intuitive understanding of these scales, which is why I used my "chicken soup" examples. I stand by that - I think that people hear "fusion" and do not understand how much unlike a star the reactor would be.
But let's not write off fusion over fission because the most common reactor in the solar system has a comparatively low energy output. It does kind of underemphasise the immense amount of energy that nuclear power generates per unit of fuel.
It would take a few kilos of seawater to power a family for a year (a single person drinks tonnes of water each year). It should be very safe compared even to non-fission options. The output is relatively clean, especially compared to non-fission options. Excluding initial R&D costs, it should be more efficient than fission in therms of $/W.
The problem with fission is that the fuel is really nasty stuff that requires all kinds of special precautions when you mine it, when you process it, and when you dispose of or recycle it. Even when you shut down a fission reactor, you still have a bunch of highly radioactive material sitting there that has to be dealt with. Also, the world will eventually run out of uranium, just like it will eventually run out of fossil fuels. I'm not going to try and predict exactly when that will happen (I'm guessing probably not in my lifetime), but it will happen eventually.
Fusion, on the other hand, uses hydrogen as fuel and leaves you with helium as waste. I don't know what kind of radioactive isotopes come out of fusion, but I'm willing to bet that, unless you're deliberately trying to produce them, you probably get way less than you do with a fission reactor. In short, waste disposal for a fusion reactor is trivial compared to a fission reactor, and when you shut down a fusion reactor the radiation hazard pretty much goes away immediately. Hydrogen is also a lot more common than uranium, so long after all of the uranium is gone we'll still have plenty of hydrogen around.
Also, any system of safety and containment that will make this look good, will make fission look even better.
1) Meltdown. Chernobyl, Fukushima, TMI. This type of problem can't happen with fusion reactors, as the reaction stops as soon as reactant stops being fed in. This is not the case for fission, because of the need to attain near-critical mass before the chain reaction can occur.
2) Waste disposal. As fissile material is consumed by a fission reactor, the resulting waste products are still highly radioactive. This waste has to be stored safely (the famous cooling pools from Fukushima were the first step in this process), for very long periods of time, generally measured in decades and centuries.
3) Fuel availability - there are not huge quantities of fuel-quality fissile material available on Earth. This limits it's usefulness going into the future
4) Theft of fuel - although reactor fuel is not generally very useful for bombs, sometimes the waste is. Some reactors, eg breeder reactors were in fact designed to generate bomb material, more so than to generate electricity.
Have a look at that list, and think about how it applies to nuclear fusion. Meltdown is impossible, waste disposal is far easier (it's just the reactor's shield that gets radiated, which is a much more contained problem than that of radioactive reaction by-products. The amount of fuel for a fusion reactor on this planet is just huge, and you can't make bombs out of the material coming in or out of a fusion reactor.
These are the reasons that fusion reactors are highly interesting, not because of the possibility to generate more power.
Articles like this are a strawman, and Stewart C. Prager, director of the Princeton Plasma Physics Laboratory, knows better and should be ashamed of himself. To take up just the matter of catastrophic accidents: it is said correctly that a fission powerplant can never explode like a fission bomb - the atoms of fuel are in the wrong arrangement for that. But any continuous system for producing fusion power is theoretically capable of "running away" - it is a self-reinforcing process, after all - and producing an actual nuclear explosion. They will design it to make this spectacularly unlikely? OK, sure they will, but let's not talk about this being danger-free.
Or the "fuel from seawater" thing. Seawater contains 3.3 ppb of uranium, and it would be commercially feasible to extract it if the prices were higher. It's actually easier to do this than to design fusion reactors. And you can use more advanced fission cycles, especially the thorium ones, to extend the energy you get out of fission fuels.
EDIT: let me clarify the actual quote from Dr. Prager that I am complaining about:
Harnessing nuclear fusion, the energy that powers the sun and the stars, has been a goal of physicists worldwide since the 1950s. It is essentially inexhaustible and it can be created using hydrogen isotopes — chemical cousins of hydrogen, like deuterium — that can readily be extracted from seawater.
This is so incredibly misleading in many ways, without containing an actual false statement. This kind of manipulation of popular opinion makes me angry. Does anyone want to defend this, even as a popularization?
Indeed. There's about a factor of 100,000x right there -- 500x for increased uranium supply (5e9 tons in oceans, vs. 1e7 tons in "conventional" <$130 kg/ton reserves (IAEA est.)), and 200x for increased fuel efficiency.
But any continuous system for producing fusion power is theoretically capable of "running away" - it is a self-reinforcing process, after all - and producing an actual nuclear explosion.
Don't think this is true. Fusion plasmas need active confinement -- a hotter plasma is no longer confined, and will no longer fuse.
This is so incredibly misleading in many ways, without containing an actual false statement.
For example, the limited resource is not deuterium but tritium (a misdirection). The plan to create this is by fissioning lithium, which is actually a rather scarce element (comparable with uranium, ironically). To get more interesting sustainability you need to switch to something like D+D fusion, which is much harder to achieve.
I still think he is being disingenuous when he says that there is no danger of accident with fusion reactors. Proposed designs have huge irradiated masses and very large neutron fluxes. There is going to be a ton of active safety systems needed, any way you imagine it.
"Less than 1g of fusion fuel is present in the Vacuum Vessel at any one moment."
Neutrons are not really more penetrating than other sorts of radiation, just differently penetrating, as they are blocked not electrostatically (by metals, as in the case of alpha/beta/gamma radiation) but by nuclear interactions (i.e. absorbed by atomic nuclei) requiring the use of different sorts of materials to block them (aka neutron modulators). It is largely a solved problem.
>It's technically true, but the process is very expensive
Well... yes, heavy water costs a good $600 / kg for the super-pure sort used in nuclear reactors. Still, this isn't really more expensive than plutonium or enriched uranium -- I can't get cost figures mostly because civilians like me can't buy plutonium, surprisingly. Tritium is a lot more expensive mostly due to a current lack of production and use, but it can be produced easily from lithium-6 or potentially from heavy water itself.
Fusion, however, has a few significant advantages:
The only radioactive product is tritium itself: a beta emitter with a short half-life which, unlike radioactive strontium and iodine, has zero potential to accumulate in biological systems and the environment is already full of the stuff thanks to cosmic rays.
The reaction is not a runaway -- if the confinement somehow fails, fusion stops instantly. This is not great to harp on because nuclear fission is still one of the safest sources of energy, statistically speaking.
The fuel is ubiquitous -- there are no terrorists selling heavy water a la Back to the Future; both deuterium and lithium are recovered from seawater, and doing so is far easier than recovering uranium from seawater. This means far less trade issues and no Iraq War type stuff in the interest of securing the Earth's supply of heavy water. Uranium, by contrast, is distributed unfairly throughout the Earth.
The fuel is inexhaustible -- at last count, if producing tritium from deuterium can be achieved, the Earth's supply of heavy water will last for several million years. Even the most starry-eyed predictions from thorium only have around a 10000-year supply prediction. ( http://en.wikipedia.org/wiki/Energy_development#Nuclear is the source here -- Wiki claims 150 billion for deuterium alone, but I'm assuming our energy usage will keep increasing)
Let Dr. Prager say that in his article, then. He won't, because it will take away any sense of urgency from supporting his research; there are hundreds of years before it becomes necessary, and there is no reason not to leave this for people in the year 2371. I think his motivation is good, in the sense that he has real scientific curiosity about this, but it's still unethical to bamboozle people with talk about "harnessing the power of the stars" and "fuel from seawater", neither of which is really the point here: the stars are irrelevant, the tritium doesn't come from seawater in any meaningful sense. Current proposals for fusion reactors are also uneconomical, which is why people invent hybrid fusion-fission economies.
> This is not great to harp on because nuclear fission is still one of the safest sources of energy, statistically speaking.
Then don't harp on it. With immense shielding and cooling mass that needs to get irradiated in current fusion reactor designs, the safety is oversold; and some of that mass will get activated, since you can't economically make it only of materials that won't be.
Not to mention the writer forgets there are countries without seashores. How are they supposed to generate their energy in this fusion-driven utopia?
Fusion has been about the same number of decades (two) away since I was a young boy. I don't think it deserves the faith we are depositing in it. Much like space colonization and jetpacks, it is a broken promise.
I'd also not forget other alternatives like molten-salt solar, hydro, wind and even photovoltaics that can offset some of the needs for power generation.
Edit: You can see actual progress on ITER http://www.iter.org/doc/all/edit/img_galleries/chantier_25.0... But, wait that's just an amphitheater the actual fusion power plant is still a hole in the ground http://www.iter.org/doc/all/edit/img_galleries/dsc_0767.jpg (June 2011).
PS: The US program has been derailed by attempts at fusion using lasers which had nothing to due with power plants and everything to do with simulating bombs. As to Europe they are just really slow Japan did make some interesting progress but they don't really allocate the money to fund such things.
I would rather people replace that overused line with the statement: "Technology T has barriers B, C and D preventing its successful implementation and mass production". Leave feel-good speculation out of it and focus on overcoming the obstacles.
Might be more technical, but it should be able to be operated without such a high technological level. So you could sell a system to a third world country without a great deal of concern over their nuclear capabilities.
I would believe the "It's 20 years away" figure if, 20 years ago, the same folks would have said we were 40 years away, instead of 20. I heard the 20 number a lot and I have seen very little actual progress beyond successive discoveries that the problem was much harder than they thought.
Why not try to be on the leading edge of something with practical applications 5 years down the road (like the thorium cycle, for instance)?
To push forward the state of the art? Sure, you can wait for someone else to do it for you, but you give up quite a lot in doing so. Namely: you pay a premium for the technology having not taken the risk, you lack local expertise with the technology as it was developed externally, you will suffer knock-on effects in unrelated areas as research and development resources move to where there is demonstrated ease of funding. Note that you suffer the last one regardless of whether the investment is actually successful.
This is one of those cases where the expense seems disproportionate compared to the risky return. Betting on fusion without hedging the bets on, say, thorium fission, is a very risky investment to make. Remember that after achieving sustainable fusion on the required scales, you still have to deal with collecting the energy released in the form of fast neutrons and transforming it in something more manageable.
Millions of years from now after the human race fails and alien archeologists are picking apart the carcass of its long-dead civilization, they are going to find an enormous number of exactly these circumstances. From education to space travel and power generation this kind of folly is so ubiquitous that it is more newsworthy when it doesn't happen. There must be some aspect of human and group psychology that ensures it. As such it's hard to have much hope for the future.
Just kidding about the aliens though. It's unlikely any would ever discover our ruins.
The project would be doomed to failure without intermediary objectives. Long term planning is difficult as it is and politics adds to the implausibility.
If you put more realistic probabilities in (hint: substantially less than 100% in both cases), and then realistically account for all the other things going on and their realistic probabilities of success and the values when they succeed, human behavior looks a lot less irrational. Maybe not entirely rational, but certainly not blindingly stupid.