Thorium Reactors for the Moon?
I was bored today, so I went browsing through the links on some of the sites linked to on this blog, when I stumbled across an interesting blog about nuclear reactors using Thorium, appropriately titled Energy from Thorium. The blog has lots of interesting technical posts about Thorium breeders and liquid floride reactors. Normally, anything related to flourine at high temperatures gives me the willies, but apparently Kirk (the proprietor of the blog) is talking about flourine salts (LiF, NaF, or BeF and various combinations of them) mixed with either tetraflourides of either Uranium or Thorium. I like the fact that the process has some very strong "passive safety" features, and the continuous thermal breeding/reprocessing that Kirk was discussing while a bit complicated is probably quite doable (while gaseous Flourine is involved, sparging and distillation aren't that complicated of processes--for a well-trained chemical engineer).
It's particulary of potential interest for lunar reactors, since there has been at least some evidence that there may be Thorium deposits on the moon (in the form of KREEP deposits), and also because Thorium breeders have the capability of being very, very efficient compared to a Uranium based fuel cycle.
The other good news is that if Thorium deposits are found on the moon, they'll be much easier to actually convert into useable reactor fuel than will Uranium. Uranium needs to be enriched to a much higher concentration of U-235 before it can be used in a reactor. This usually involves a lot of massive and expensive enrichment systems like gas-centrifuge cascades. Naturally occuring Thorium on the other hand is almost entirely of the Th-232 form, which is the form used in the "fertile" loop of a Thorium breeder. Now, to get the reaction started, you do need some U-233 to provide the initial reactivity to ge the breeder going. There may be creative ways of doing that in-situ, and without a lot of extra industrial base required, but in the near-term it's likely that it'd be far cheaper to have some made on earth and shipped to the Moon. I guess a lot depends on how difficult it is to ship (politically and regulatorally speaking) some U-233 compared to the cost of developing something clever in-situ. Even with the current environmental opposition to launching radioactive materials into space, my money would still be on getting it from earth being the easier route to start with.
Anyhow, check out Kirk's blog, and "read the whole thing".
It's particulary of potential interest for lunar reactors, since there has been at least some evidence that there may be Thorium deposits on the moon (in the form of KREEP deposits), and also because Thorium breeders have the capability of being very, very efficient compared to a Uranium based fuel cycle.
The other good news is that if Thorium deposits are found on the moon, they'll be much easier to actually convert into useable reactor fuel than will Uranium. Uranium needs to be enriched to a much higher concentration of U-235 before it can be used in a reactor. This usually involves a lot of massive and expensive enrichment systems like gas-centrifuge cascades. Naturally occuring Thorium on the other hand is almost entirely of the Th-232 form, which is the form used in the "fertile" loop of a Thorium breeder. Now, to get the reaction started, you do need some U-233 to provide the initial reactivity to ge the breeder going. There may be creative ways of doing that in-situ, and without a lot of extra industrial base required, but in the near-term it's likely that it'd be far cheaper to have some made on earth and shipped to the Moon. I guess a lot depends on how difficult it is to ship (politically and regulatorally speaking) some U-233 compared to the cost of developing something clever in-situ. Even with the current environmental opposition to launching radioactive materials into space, my money would still be on getting it from earth being the easier route to start with.
Anyhow, check out Kirk's blog, and "read the whole thing".
posted by Jon Goff at 10:46 AM
11 Comments:
Thorium reactors can burn at least half of their thorium, and can burn all of it if they use online reprocessing as Kirk suggests. The amount of energy per kg of thorium is huge: 11 million kW-hr per kg (read the whole thing), with 50% efficient conversion to electricity. A 10 megawatt plant could run for 5 years on 40 kg of Thorium. There is no need to find thorium locally, on the moon or on Mars or anywhere else either... refuelling would be irrelevant, as the initial reactor configuration would carry enough fuel to go for decades.
Because the Thorium would be shipped as a solid, hard, unreactive, non-radioactive fluoride salt, my guess is that even launch would not be a significant political problem, so long as a low-hazard way can be found to start up the reactor.
But like you say, starting it is the problem. Maybe a deuterium neutron source could do it.
On a seperate note, what do you think of the steam launch idea now?
Any idea how I could find out about the reliability and cost of really big explosive bolts (holding back 1-10 meganewtons each)?
Iain,
There is no need to find thorium locally, on the moon or on Mars or anywhere else either... refuelling would be irrelevant, as the initial reactor configuration would carry enough fuel to go for decades.
Well, I guess I was suggesting the idea in case the political difficulty of getting U-233 and Thorium delivered to the moon/mars/wherever are too high. A lot of times when I mention that I'd prefer nuclear power for initial lunar power (since it gives you the flexibility to go wherever on the moon is interesting, not just chillin' at the poles) people bring up the environmental worries about launching reactors or anything nuclear, regardless of what the real dangers.
On a seperate note, what do you think of the steam launch idea now?
I read your article a few days ago, and have been thinking about it. Short answer is that I'm not sure what I think. For a system like ours that is supposed to be highly reusable, it sounds like a bad idea, even if it could increase payload a bunch. It just adds way too much extra infrastructure, extra ground systems, extra hassle, and extra failure modes. But if you're talking about ELVs? It might make sense in some cases, but the whole idea of a steam rocket just seems iffy to me. Keeping the whole thing hot when it's right next to a tank full of LOX....Not to mention that steam explosions can be a really, really bad day in their own right...
I guess I'm just not convinced it's a win.
Any idea how I could find out about the reliability and cost of really big explosive bolts (holding back 1-10 meganewtons each)?
Not really sure. Though at that point, you're probably talking about just putting shaped charges around a bit bar of metal or something. I'm not really sure. Have you looked at using a SeaCatch instead? They have standard systems up to about 150klb of force, and claim to be able to make bigger models. They're relatively easily actuated including via pyrotechnic squib, pneumatics, or even a pull-chain if you prefer ;-) (and they are reusable and testable).
~Jon
Nuclear fuel is one of the last things we'll mine in space, since the shipping cost from Earth will be quite affordable (unless launch costs are so high that space exploitation never starts at all.) The Earth is one of the best places in the solar system to obtain fissionable materials, since they are highly enriched over chondritic abundances (by several orders of magnitude) in the Earth's continental crust, and ore-forming processes involving water have produced even more concentrated deposits.
I expect space reactors will be fast reactors with highly enriched fuel to minimize mass.
There's only one scenario I could see in which something like a molten salt reactor could make sense in space: if large quantities of spent reactor fuel are being shipped up from Earth for disposal on the moon.
>>>>On a seperate note, what do you think of the steam launch idea now?<<<<<
>>>I read your article a few days ago, and have been thinking about it. Short answer is that I'm not sure what I think. For a system like ours that is supposed to be highly reusable, it sounds like a bad idea, even if it could increase payload a bunch<<
FWIW high energy density thrust seems to be safer than low energy density thrust when I try to run numbers on it. A ton of very hot water under high pressure can be more dangerous than 0.1 tons of LOX/propane at 30 psi cryo temps.
The only reason I see to go with steam is if combustion rockets are viewed as even more difficult and expensive. The boiler 'explosions' during the steam age were more dangerous and common than diesel engine explosions.
John Hare
The only reason I see to go with steam is if combustion rockets are viewed as even more difficult and expensive.
A steam '0-th' stage is simple and rugged and presumably easily recovered after unretarded ocean splashdown, but the same would be true, I imagine, of a low performance blowdown pressure-fed chemical stage (especially if operated with a large amount of water diluent or far from stoichiometric ratios so the temperature is down in the steam rocket range.)
So what if the space reactor is neutron-inefficient. You just ship up more fuel. There's no reason to try to make your space reactors into breeders. We don't design shipboard reactors to breed fuel; why space reactors? Indeed, for shipboard reactors we use neutron-inefficient burnable absorbers to increase the lifespan of a fuel load.
As for the safety issues you raise: low probability catastrophic accidents that would be unacceptable on Earth would not sigificantly affect the chance of the crew being lost in a space mission. In a situation where loss of the reactor for any reason threatens the crew, one may want to accept a higher chance of a catastrophic accident if it reduces the chance of lesser (but still reactor-disabling) accidents. Similarly, if the lower mass of a fast reactor enables one to spend mass on reliability elsewhere, the putative advantages of a thermal reactor may not be worth it.
On a more technical point: the MSRE (the only molten salt reactor to be operated for extended periods) the peak temperature was 650 C. This is rather low for a space reactor, where either waste heat needs to be radiated at high temperature (remember, radiator areas goes as T^-4), or you are heating propellant and want a very high temperature (> 2000 C in NERVA designs.)
Paul Dietz said...
>>>The only reason I see to go with steam is if combustion rockets are viewed as even more difficult and expensive.<<<
>>A steam '0-th' stage is simple and rugged and presumably easily recovered after unretarded ocean splashdown, but the same would be true, I imagine, of a low performance blowdown pressure-fed chemical stage (especially if operated with a large amount of water diluent or far from stoichiometric ratios so the temperature is down in the steam rocket range.) <<
I think we agree on the diluted proppellant booster possibilities.
A major water film cooling would protect the chamber walls with the main burn at near stochiometric for maximum temps. A 3/1 ratio water/propellants would seem to cut Isp in half for a potentially very robust system. I would go for a bit more performance than the orriginal though.
John Hare
Jon, if your bored, I have a suggestion/question
Last week, under your Space Access Notes Day 3, I posted a short bit on launcher independence, and was wondering your thoughts on pushing for that for CEV and the Lander, given what Jim Muncy said
Jon,
SeaCatch -- cool! So I could theoretically prototype this thing with no controlled technologies at all. Wow. I may have to do a budget. It would be pretty cool to be able to fire a wrecked school bus a thousand feet into the air with a pull on a rope.
John,
Not like I'm an expert at this stuff, but:
- Steam explosions happened because people did not quantify their margins and/or pushed them too low. Nowadays, steam explosions do not happen, even in the pressure vessels of boiling water reactors which have thousands of tons of very hot water, because folks use higher margins.
The flaw here is that when the steam stage crashes into the ocean, it may take some invisible damage. And then the question is: is it cheaper to melt it down and refabricate, or to test it? For instance, if simple hydrostatic testing is sufficient, then reuse should be pretty cheap.
- A super-simple combustion 0th stage may end up lifting more mass per buck to orbit. But even the simplest combustion stage needs mixing (an injector) and ignition (or toxic cleanup). The really neat thing about a steam stage is that it has *no* injector, *no* mixing, *no* L*, *no* reaction of any sort, in fact, *no* moving parts. That's a big deal if you want to scale up.
- If the upper stage (last of stage 0, 1, and 2) is reusable, then perhaps it makes sense to combine the bottom two stages into a Beal-type pressure-fed rocket. But if the upper stage is expendable, then my sense is that you want that stage as small and cheap as possible, and make the recoverable first stage do much of the delta-V. And, a two-stage steam then Kero/LOX rocket seems cheaper than a single higher thrust Kero/LOX rocket.
Ambivalent,
>>>> Steam explosions happened
because people did not quantify their margins and/or pushed them too low. Nowadays, steam explosions do not happen, even in the pressure vessels of boiling water reactors which have thousands of tons of very hot water, because folks use higher margins.<<<<
I am not expert on these particular systems either. My opinion should be read as 'relative safety & difficulty.' I am not in the serious space business at this time. I do build tools and light equipment for my company that I design. At my current level of operation, the designs are scale sketched and built the same day with mostly intuitive design. My intuitive feel for your concept is that I would rather take my chances with combustable propellants in smaller cooler tanks and design an injection system. This is not a suggestion that you should stop work based on my opinion.
>>>>>- A super-simple combustion 0th stage may end up lifting more mass per buck to orbit. But even the simplest combustion stage needs mixing (an injector) and ignition (or toxic cleanup). The really neat thing about a steam stage is that it has *no* injector, *no* mixing, *no* L*, *no* reaction of any sort, in fact, *no* moving parts. That's a big deal if you want to scale up.<<<<
Jon did a good deal of work on MCD boosters a few years back. I think it shows in the relatively rapid progress of Mastens' pintle injector rockets. If you were to specify say 1,000 m/s for the zero stage, including losses, they could probably deliver it with a mass ratio of 1.5. Your steam stage for the same performance could easily have a mass ratio of 6 or more. While simple is cheap, a quarter of the GLOW mass could possibly be even cheaper.
Even worse. If you standardize on 1 ton for the upper stage(s), then Mastens' zero stage could be a half ton to your 5 ton steamer. If you care to offer more accurate numbers for your system, I can put a lot more accuracy in my replies.
A number of years ago I was convinced of the usefulness of launch assist platforms by Len Cormiers' papers on the subject. I think you are on the right track. If I ever get in the right financial position again, I intend to pursue an LAP concept of my own.
John Hare
John,
I design circuits, not mechanical things, so my intuition probably counts for less than yours.
Elon Musk's rocket has valves that cost $25,000. Presumably not many of these, perhaps two. As long as you have liquid combustion, you can't get away from this kind of thing. That kind of money will buy an awfully big pressure vessel, which is why solids are so popular.
In any case, I'll not be building this stuff, I've got other things to do. :)
Post a Comment
<< Home