Hard SF discussion

General Information > Fusion May Not Be Practical

Comments Showing 1-17 of 17 (17 new)    post a comment »
dateDown arrow    newest »

message 1: by Rikhard (new)

Rikhard Von Katzen (rikhardvkatzen) | 18 comments My blog post:

As Discovery Magazine noted, for some reason fusion power is always 30 years away. Every fusion reactor has been at less than 1:1 I/O for power, and none of them have worked for more than a few minutes (and the most efficient ones for far less time). Fusion power is a staple of most hard science fiction, but what if it's just not possible? Much like the 'fusion torch' of Heinlein it seems at least possible that this is an engineering red herring, something that we'd like to have but can't actually build. Writers tend to treat fusion power as the natural next step beyond fission but it may be more like antimatter.

In fact most realistic antimatter power plants are fusion plants that use antimatter boosted fuel to enhance reaction temperatures, not pure M/AM reactors which would basically be bombs - but fusion reactors are already basically bombs. The fact is that we have only two examples of really effective fusion: stars, which use their mass for slow and weak fusion reactions, and thermonuclear bombs.

Volumetrically the sun has about the same energy output as a goat, only its superdense insulation, sheer size and ancient age that makes it so hot. The sun is pretty fuel efficient but not very powerful. It's also very difficult to build a star, much less move one you already have. It's vastly outdone in the time-energy equations by the only other example we have of fusion, the thermonuclear bomb. Thermonuclear bombs reach incredible temperatures for a tiny fraction of a second. Since heat difference is the quintessence of reactor efficiency this makes thermonuclear bombs a good candidate for super-reactors. The unfortunate drawback being that they instantly vaporize all known forms of matter and throw off huge amounts of hard radiation that can punch through feet of solid lead to kill anything on the other side. There is also the interesting fact that the majority of the energy in a thermonuclear bomb is not from fusion: it's from fission. Fission-fusion-fission layers are built in the bomb, with the fusion stage being used to rapidly generate neutrons which then boost the next fission stage's reaction efficiency. The actual 'boom' in a thermonuclear weapon is almost entirely the product of that final stage of fission, with all the intermediate fission/fusion stages being used to generate heat for fusion and neutrons for further fission.

In many ways a fusion power plant suffers from the same sort of quixotic design features as a 'plasma gun'. The idea of a plasma gun is to set off a nuclear weapon and then shoot it. The problems with this should be obvious. A fusion reaction is a superheated explosion and by definition is very difficult to contain. As I said earlier the quintessence of reactor efficiency is heat differential, a high concentration of heat in a small space in a short time compared to the external world is the thermodynamic principle that allows power generation, no matter whether you're talking about an internal combustion engine or an antimatter reactor. The problem is that for extremely high powered reactors the heat levels have to be so high that it's almost impossible to design anything made of real-world materials that won't simply blow up or, at best, prevent the bomb you've set off from simply dissipating as free plasma.

One of the supposed super-fuels for fusion is Helium-3. Unfortunately Helium-3 is not efficient. It skips a step in the fusion process, making it inefficient compared to Hydrogen-Deuterium reactions. And although it does not release any neutrons in its fusion process directly it will still create neutron secondary radiation because it does release hard radiation that is sufficient to free neutrons in the surrounding matter. So while it probably would produce less ultra-deadly radiation than a normal H/D reaction it still produces enough to require a radiation shadow shielding. And since it's less efficient than H/D, requiring you to carry more of it for fuel, it becomes sort of pointless: you could just put more shielding on a H/D reactor, since radiation shielding is vastly cheaper than the virtually non-existent He-3 isotope (there may actually be entire oceans of He-3 on some of the gas giants, but getting anything off a gas giant is vastly more difficult than simply building enormous platinum walls) it becomes questionable that there would actually be any reason to use He-3 fusion reactors - if, indeed, fusion reactors are ever practical in general.

Earlier I mentioned the fusion torch engine. This is a theoretical (more likely mythical) form of fusion reaction engine that has high delta-V and high acceleration. Aside from the fact that it would destroy anything near it because it is literally a thermonuclear bomb that goes off continuously for days it also doesn't seem to be practical from any known real-world fusion or rocket engineering processes. It's good Unobtanium because you can calculate how it would work, but that doesn't mean it ever will work. Fusion power might be the same way. The reason sci-fi authors want these things is because they give them a lot of power to play around with, and as Charles Stross has observed doing anything in a speedy manner in space requires outrageous amounts of energy. Of course energy is energy, and if you can build fusion torch engines and fusion reactors you ought to be able to reduce entire planets to radioactive rubble pretty easily. The fact that sci-fi weapons often fall far short of their power plants (the Lensmen series being a good counter-example, where at least Nonsensium Power translates into planet-smashing genocide on a regular basis) is a perfect example of the author not doing his research.

Real space travel, colonization and warfare seems unlikely to rely on such theoretical unbotanium sources, and will probably be limited to some improvements on what we already have: chemical rockets and nuclear fission power plants. HEDM rockets and generators present some engineering difficulties, but these are trivial compared to containing a 900 million degree plasma in the space of a shoebox. Fusion reaction engines seem more practical than fusion reactors, because they do not have to be sustained and contained to the same degree. But whether of the Orion method (tossing hydrogen bombs out the boot of the spacecraft) or a hot plasma engine these are not going to be the torchships of Heinlein and Karl Gallagher - they're going to be high-efficiency, slow-burning, and probably under 1G of acceleration. The upside to this is that it saves us the trouble of having to worry about other science fiction hand-wavium such as inertial dampners - we will not be accelerating fast enough to even maintain normal biological homeostasis, much less plaster ourselves to the walls.

message 2: by Esha (new)

Esha Nas | 5 comments I've never heard of a plasma gun using a nuclear bomb...at all.

As for fusion being '30 years away', isn't the problem a mixture of funding and applied technology? Where's that graph that shows the estimates for Fusion technology with aggressive, moderate, and low funding with a 'Fusion NEVER' line and our own allocated funds far below that line? Not to mention the forty years of advances since the 70s in materials technology to contain the apparatus and the plasma that basically forces engineers and scientists to reinvent the wheel, familiarize themselves with the new tech, and apply it in the real world. Apparently MIT or somesuch developed a new magnetic thing to contain the plasma and thus even ITER may very well be outdated...before it was even completed. That's a problem.

Since we've never given the fusion...'community' the funds they need since the 70s, we can't really say that its outright impractical. It's the same thing with antimatter: to my knowledge, we've never built a dedicated antimatter production facility, just got it as a sort of waste product from other experiments and then a few direct ones, and thus we can't really say it'll always be prohibitively expensive to produce when we're not, say, putting nuclear and solar power'd antimatter production facilities right in orbit around the sun.

It takes more than just a few papers saying we can do something to put that thing out there. You need engineers, you need materiel, you need to build prototypes, test them, then tweak them, build on them, build better ones. Hell in Nuclear power technology we still have liquid and gas core tech to even build, let alone proliferate.

message 3: by Rikhard (last edited May 10, 2018 07:53PM) (new)

Rikhard Von Katzen (rikhardvkatzen) | 18 comments Daniel wrote: "I've never heard of a plasma gun using a nuclear bomb...at all.

As for fusion being '30 years away', isn't the problem a mixture of funding and applied technology? Where's that graph that shows th..."

There has been a lot of money blown on fusion over the past 60 years. And if it were actually likely to be useable there would be plenty of private money, instead of having to rely on handouts. Any power company would love to produce tremendous amounts of power from relatively cheap hydrogen and helium. However, much like space travel, it doesn't seem like there's much good reason to actually bother with it. I really don't believe the whole muhFunding complaint from academic welfare cases. I think that stuff like String Theory, NASA, etc. are just money pits and welfare for an excess number of layabouts of the middle class. The dole for PhDs.

For that matter, fission is cheaper, it produces less problematic radiation (x-rays instead of neutrons), it's easier to turn into electricity (neutrons have no charge!) and the fuel is super-abundant. Fusion is not only impractical and possibly impossible for any sustained use but it appears to be entirely unnecessary and at least as dangerous as fission reactors on a number of grounds. Shielding against most fission products requires a few inches of water, shielding against neutrons requires feet of lead.

message 4: by Esha (last edited May 10, 2018 08:26PM) (new)

Esha Nas | 5 comments How much money? Because it seems we barely give it a billion a year. That's NOT a lot. But that is more than what the private sector would cough up - it's something only a state can do. Even energy corporations don't have that much money just lying around to spend on projects when they have stockholders to appease and more infrastructure to build for more profit.

We're talking about dozens of billions over decades. So no, I don't buy that it's off the table just because it hasn't been done because it hasn't even been tried. ITER has a cost tag of 20 bil over as many years. Then a successor to take advantage of all that was learned in DEMO. Another few bil. That's easy for a first-world power, or group of powers, to do if they actually gave a damn, with their trillion dollar yearly budgets; but far beyond any single person or company.

Again: cash is the key. Same with space travel. It needs cash. Constant, yearly cash in the tens of billions, in face of shifting governments and economic boons and bursts. Interplanetary ships - at the least, their projects as a whole - cost billions at the low end. Bases: the same. Their supporting industries have to be built up and expertise garnered. Money and time politicians nor corporations want to give up. It's not due to lack of capability but lack of drive.

message 5: by Outis (new)

Outis | 64 comments Dear Feline,
So it turns out your problem is political in nature.
I don't disagree with you about basic physics (it's no secret the main point of fusion research hasn't been to develop better reactors) but writing stuff like "String Theory, NASA, etc." is indicative of an engagement with value judgements at the exclusion of facts.

I agree one shouldn't rule out things which are possible in theory just because they have never been tried. By the same token, don't rule out corporations or individual capitalists playing a major role in funding these fields. Certainly that's looking a lot more plausible nowadays with the ethos of the IT billionaires.

message 6: by Rikhard (new)

Rikhard Von Katzen (rikhardvkatzen) | 18 comments Outis wrote: "Dear Feline,
So it turns out your problem is political in nature.
I don't disagree with you about basic physics (it's no secret the main point of fusion research hasn't been to develop better react..."

No, I think Fusion is just an unrealistic technology and muhSubsidies is a shitty excuse for its failure.

message 7: by Outis (new)

Outis | 64 comments This subsidies notion is your excuse to avoid engaging with facts. You've brought it up in two different threads now, only the political motivation was less obvious the first time.

message 8: by Steven (new)

Steven Jordan (stevenlylejordan) | 3 comments The fact is, the physics say it's possible. So it's a matter of developing the apparatus to make it happen. No, we haven't cracked that yet... but advances in materials, better understanding of elements like plasmas and their control, workable scale, etc, bring us closer every day.

So I wouldn't say it can't happen. I think it can... when our technology is advanced enough to generate, sustain and control the reaction. And that day may not be as far away as you think.

message 9: by M.D. (new)

M.D. Cooper (mdcooper) | 13 comments Regarding the assertion that fission is better:

There is a MASSIVE problem with fission, and that is the mining process. Because of the halflifes of the common isotopes of uranium found in uraninite (formerly called pitchblende) much of the uranium has already decayed. This means that mass-wise, most of the unused material in uraninite is not used, and left in mining tails. Radium, which is common in uraninite tills, for example, is about 400,000x more radioactive than the common uranium isotopes.

Because of the cumulative halflifes of most of the radioactive decay products of uranium is roughly 80,000 years, we now have water-covered, dam retained, mining tails all over that are ridiculously radioactive. We have to maintain these tails for the next 80 to 100 thousand years--or extract the material and get it off planet.

So...we should work harder on fusion.

message 10: by Outis (new)

Outis | 64 comments In this thread's context, fission fuel would be mined (or produced) off-planet to begin with.
And I think this is the wrong group to try to scare people with random large numbers.

message 11: by M.D. (new)

M.D. Cooper (mdcooper) | 13 comments I brought up the risks of mining uranium because of the OP and the thread's general references to near-term power generation. There was also reference to the risk of fission vs fusion. It's noteworthy that when you take U-238 and U-235 and use them in fission, you get the decay chain, much of which is far more radioactive than U-238 or U-235. So, mining or using in fission reactions, these decay chain isotopes are real concerns.

Regarding where you'll find uraninite (the primary uranium-containing ore): Because of the temperature variances in the protostellar disks that form stars and the planets, denser/heavier elements are closer to the star, so uraninite is most likely to be found on terrestrial planets. Given that Earth is the densest object in the Solar System, and is likely to contain the largest deposits of uranium in our star system, it may very well continue to be mined here.

Regarding my numbers: these aren't random large numbers. You can look up the decay chain of Uranium 238 (the most abundant form of uranium due to it's half life of 4.5 billion years) for yourself and do some simple math.

When U-238 decays, it goes through two quick intermediary steps and then hits U-234. That isotope has a half-life of 245k years, and from there it hits Thorium 230 (75k year half life) and then Radium 226 (1600 year half life). This is where my 80k year number comes from. The more precise number is 76,600 years given the half lifes.

When Uranium is mined, they leave the thorium and radium in the mining tails because its not useful. Both Thorium and Radium decay with alpha particles, so their radioactivity isn't not too hard to block--a sheet of aluminum foil is almost enough. However, if you get any of that inside of you, its going to ruin the rest of your life. But they also decay with gamma radiation, and you really don't want that around.

The way these mining tails are managed is by covering them with water in things called tailing ponds. There are tens of thousands of these tailing ponds at old mine sites that government agencies have to maintain after the companies that mined them shut down. There are many uraninite ones.

Regarding my statement that Radium 226 is 400,000x more radioactive than uranium, you can quite simply verify my statements.

R-226 has decay rate of 37 billion becquerels, and U-238 is 25.4Bq. Becquerels measure decays per second in 1mg of the isotope.

However, there is decay, and there is decay. It depends on whether or not you get alpha particles, beta particles, or gamma waves from the atomic decay process. Alpha and beta are more damaging, but not as penetrating. This is why R-226 is only about 400,000x as damaging radioactively as U-238 and not 1.4 million times more so.

message 12: by Outis (new)

Outis | 64 comments The OP is about power generation in spaceships much more powerful than existing ones.
You don't choose mining locations by looking at where most of the stuff is. Like, we don't get iron from the Earth's core. There's more than enough fissible isotopes outside of our planet's bothersome gravity well.

The issue isn't whether your numbers are accurate but with what purpose they serve. Hopefully everyone here realizes that it should be no surprise that isotopes with very long half-lifes are very weakly radioactive. I'm somewhat less hopeful that everyone here realizes that we breathe radon all the time. And I know for a fact that there's at least one person here who doesn't know what a becquerel is.

message 13: by M.D. (new)

M.D. Cooper (mdcooper) | 13 comments Thank you for being so reasonable and thorough in your responses, Outis, it does you a great service. I did mention radium, not radon. Radon is the next down on the decay chain. You're right that the ppm we breathe is not harmful, though.

message 14: by M.D. (new)

M.D. Cooper (mdcooper) | 13 comments Regarding where the Uranium in the solar system is, we know for a fact that its in the continental crusts (and not too much in the mantle) of Earth. However based on the distribution of metals across the system, it is likely that sources outside the Earth hold 12x the uranium that Earth does (not including the sun). The thing we don't know yet is whether or not that uranium is close to the surface like it is on Earth. Comes to around 640 trillion tons.

We should work out how much thrust you can generate with that much uranium, and how much mass you can move around with it.

message 15: by Outis (new)

Outis | 64 comments But most of these 12x would be in even more bothersome gravity wells than the Earth's, yes? So I think it's not preposterous to assume that most of the accessible deposits are on Earth. The point is that only a small fraction of such deposits would need to be accessed so the ones on Earth can for the most part be left alone.

Around here, radon can be harmful. I brought it up not only because it gets in our bodies but specifically because it decays a lot more often than Ra-226 which goes to show none of your numbers mean much on their own.
Giving people a fair picture of how dangerous these tailings actually are would require putting a lot more work in digging up numbers which aren't found in textbooks, and they'd be contentious to boot. You'd have to put them in context of natural hazards as well as coal tailings and so forth. Thankfully the big picture of energy generation in the near future is pretty obvious: non-radioactive toxicity of waste products and climate change are both bigger issues. Radiation would on the other hand be quite relevant to the durability of certain parts of nuke-powered starships.

message 16: by M.D. (new)

M.D. Cooper (mdcooper) | 13 comments You get radon-222 from radium-226, so much of the radon you're worried about is coming from U-238 in the crust of the planet. There's just more of it where they principally mine uranium (Canada and the Congo).

Yes, radon-222 decays in days vs years for Radium-226. But what that means is that the radium in the tailings pump out radon for millennia. I didn't get in to that at first because I was picking an example out of the decay series, not walking through the whole thing.

I've personally spent a lot of time reading the research being done on this in Canada because the tailings there are particularly dangerous and the country is strapped to maintain them all after the mining companies shut down. But folks can certainly research that on their own.

You make a good point about the decay of the isotopes in your fuel supply. All of the U-235 and U-238 in your starship's fuel is decaying into thorium and a few other things at all times. The alpha and beta particles aren't a big deal, but you'd need decent gamma shielding around your fuel otherwise its going to bork semiconductors.

Granted, you get a lot more gamma from the proposed antimatter alternative, so its going to be an issue no matter how you slice it.

message 17: by Outis (new)

Outis | 64 comments How do tailings compare to climate change adaptation in terms of public maintenance costs?
Even harmless-sounding energy sources can be dangerous: en.wikipedia.org/wiki/File:Boston_pos...

back to top