Thorium-fueled Molten Salt Reactors

The separate "seed" is present in the designs I've seen, and has the advantage of being removable, replaceable and reusable. You could surely mix in some U233 (or U235) to the fuel/salt mix to kick-start the reaction, although you'd need a significant amount.

I agree that they have some significant advantages. But light water reactors were supposed to result in electricity "too cheap to meter" as well. Often even 99.9% safe technologies (as LWR's were touted) turn out to need more protection/containment/leak mitigation than originally envisioned.

Most of the water in a heat exchanger is in the form of steam - and that, unfortunately, does indeed go uphill.

Also agreed - but again, a safety system that requires you 'blast a big hole in the reactor cooling system with explosive bolts like they use on the Space Shuttle, and then have a fire in a big unoccupied room' has some obvious drawbacks.

With a light water reactor, as soon as you drop the rods, the chain reaction stops immediately. You then have to deal with the decay heat as the short-lived isotopes decay, which is why you have to keep cooling them.

The thorium reactor is no different; the short-lived isotopes generate heat for a long time (minutes to years.) As time goes on, of course, the generation of heat declines. Another problems is that some of those isotopes generate hard gamma radiation (.2 to 2.6 MeV) and hard gamma radiation is both extremely dangerous and very hard to shield against.

The thorium reactor solves the problem by "spreading out" the core over a large area, acres in the case of a big reactor. This stops the reaction (too low a cross section to continue fission) and allows a larger area for cooling the fuel with its heat-generating isotopes. So now you have a gradually-cooling acre-wide pond of molten radioactive fuel emitting gamma radiation for decades. That will probably be a bit problematic to clean up.

Well, you have to reprocess it to remove the waste, and reprocessing has been a problem for decades. (We'd have almost no nuclear waste problem if we reprocessed spent uranium, but we've outlawed it.) You can do it with remote manipulation to avoid the gamma radiation problem, but again, that will take some development.

 

Log in or Sign up to hide all adverts.

We have a lot of U235 from old bombs. I think most of it is being mixed into fuel as "Topping Off" lower enriched levels to bring them up to reactor levels. Certainly could do that one time to start up a thorium reactor, rather than many times to make refueling rods for a uranium reactor. Also the first "seeded mix" reactors canbe designed to make more U233 than they burn so when adding more thorium fuel you just take off some of the over rich in U233, then mix in some thorium with it and presto, without any new separated U233 needed, you start up a new thorium reactor. -Sort of like a chain smoker only needs one match per day.

BTW, I agree spent fuel rods which still have 95% or the U235 un"burned" should be reprocessed as France does to recover that U235.

I remember that slogan too and it would nearly have been true if not for the excessive delays, in the approval processes (capital cost of 10+ YEARS before expenditures are allowed in the utilities "rate base.") The fear of nuclear power is way out of proportions to the risk. Coal kills many more (mine and transport accidents) and release more radioactivity to the environment (due to huge volumes with small part of natural radio activity in the coal - K40, Uranium and Thorium mainly I think). I am nearly sure that "coal bed methane gas" could not be used if held to nuclear power plant release standard as there is radon in it.

The rules on release are so strict that drinking a can of beer/ per day, which the hops have picked up K40 form the soil, is more exposure to ionizing radiation than standing permanently at the fence of a nuclear power plant. Some good cites on rivers for a new plant had to be rejected because the river water could not be used for the cooling tower without pre-process it to remove natural radioactivity in it (dissolved radon mainly I think) I.e. just taking a bucket full of the river water inside the fence and throwing back out would be a violation of the release standards! That is why nuclear power never came close to the "Too cheap to meter" goal.

assuming you have electric power for the controls and that the problem is not warped fuel rods blocking the space they were to drop into, which I believe is why the operators could not get control of Chenoble again when they knew they had lost it.

Yes but if the reactor core is now spread out in about 100 sq meters catch pan (10 meter on edge square) it loses the isotope decay heat as fast as it is produced, naturally - no active cooling required.

I think that wildly excessive, 100 m^2 should do as now there is much greater surface with neutrons escaping, but if I am wrong about that also being adequate for passive isotope decay heat cooling, then there are some internal fluid convectors vertical in the pan like used to cool viscus oil pipe lines, in Alaska to keep permafrost frozen. I.e. the part of the convector in the pan contacting the hot thorium mix has a liquid inside the convector that boils and its vapor condenses in top section of the closed convector heating fins for transfer to the air.

 

Log in or Sign up to hide all adverts.

That's definitely doable, but keep in mind that to do that we have to reprocess the fuel+salt mixture to maintain the ratios of U233 vs thorium vs waste products, and currently that is difficult (i.e. humans can't be anywhere near the process.) You'd need even more reprocessing to get pure enough U233 to start a new reactor, and it would be nearly impossible to transport - gamma radiation is hard to shield against, so it might be best to use easily transported U235 seeds.

From what I've read about that accident, the neutron poisons built up quickly enough that by the time they managed to restart a reaction in the core, the rods were so far out of place that the reactor went prompt-critical i.e. it blew up like a nuclear bomb, and no control rods in the world can move fast enough to stop a prompt-critical reaction. (Fortunately light water designs can't go prompt-critical by design, due to their void coefficients.)

Depends on the size of the pan. A 100 sq m pan is going to have to dissipate a LOT of heat in a small area; a 1GW reactor is going to put out something like 3% of its power in decay heat right after shutdown, so you'd need to dissipate 30 megawatts in a small pan. You'd need active cooling for that. To get around that you would have to spread it out over a much wider area, so that natural convection/conduction could get the heat away from the fuel.

I agree. But now you're getting into something that would work just as well for a conventional reactor, and something that damaged the passive convectors (like an earthquake) could cause serious problems.

Again, I want to emphasize that these problems are all solvable via solutions like the passive heatpipe systems and blowout ports you mention above. However, now you're approaching the complexity of light water reactors with all their emergency cooling systems and containment vessels. Would thorium let you remain less complex overall for a given level of safety? Perhaps, but much remains to be done.

 

Log in or Sign up to hide all adverts.

Billy T said (About how to enhance cooling of a 100m^2 pan of hot (by residual isotope decay only / no chain reaction in progress) thorium reactor fuel that drained into a catch pan after something went wrong and the reactor core started to overheat. I.e. after automatic reaction kill by gravity alone.):

“There are some internal fluid convectors vertical in the pan like used to cool viscous oil pipe lines, in Alaska to keep permafrost frozen. I.e. the part of the convector in the pan contacting the hot thorium mix has a liquid inside the convector that boils and its vapor condenses in top section of the closed convector heating fins for transfer to the air. "

billvon replied:
"That's definitely doable, but keep in mind that to do that we have to reprocess the fuel+salt mixture to maintain the ratios of U233 vs thorium vs waste products, and currently that is difficult (i.e. humans can't be anywhere near the process.) You'd need even more reprocessing to get pure enough U233 to start a new reactor ..."

BT now comments:
"pure enough U233" Do you know what percent of U233 is in the thorium of a homogeneous core reactor as "match" at start up of a new thorium reactor? I note for others, as you understand, that the "match" can not start the chain reaction fire until the fuel geometry is right (i.e. until only a few neutrons are escaping prior to transmuting thorium).

Well can't be too difficult as the French have been doing fuel reprocess for years. They still mainly store spent fuel rod in "swimming pool" as US does for all. This because if you could recover all the U235 from one spent rod that would be nearly enough to make a new one. I.e. 95% of the U235 in a new fuel rod is still un"burnt" when the rod must be removed from the reactor (to get the neutron absorbing fission fragments out, I think)

It may be true that spent fuel from a thorium reactor is a stronger source of harsh gamma rays, but the French need to keep people save too - I.e. you can't kill the same person 100 times more dead so even if the flux is 100 times greater from thorium spent fuel - that makes no greater the reprocessing without nearby humans requirement.

Storing fuel for a few years in swimming pools near reactor is good idea for removing short half life isotopes cheaply, BUT IS NOT A PERMANENT SOLUTION TO THE SPENT FUEL ROD PROBLEM. The US (and all users) need to reprocess fuel for the reusable part and for the rest, adopt the suggestion I have made in several posts (No human near "glassification" of post-pool, long-lived, isotopes in to disk about 0.3 to 0.5M diameter and ~ 10cm thick* transferred to ship with automatic handling and "disk hurlers" on the stern that throw them into the ocean over a deep ocean trench to start their billion year journey down into the molten core of the Earth.)

-------------
*Thickness is set by requirement that the core of the disk should not melt – i.e. heat transfer from center to water cooled surface rate not too slow. Note also the outer mm or so of disk is only glass to stop all alphas inside the disk as conceivable they could active some of the automatic handling equipment, but the escaping gamma cannot.

 

I just wanted to make a quick point about the "catch-pans" The addition of carbon rods soaks up excess neurons and speeds the cooling process by slowing the reaction even more. As long as you built well designed containment vessels it would be safer than having to pump water into a reactor to cool the rods or... Anyway... What happens if a catastrophic even causes a reactor to loose power. It switches too diesel generators to keep the pumps going, once those fail or their fuel runs out you then face the meltdown problem...Also if the rods get warped, which we know has happened in the past, you cant stop the process. With the thorium MSR you lose power, the frozen salt plug melts, the thorium salt mix runs into the designed catch bottles then it starts to cool.

 

Nope. I assume that it's significant, since it has to sustain a chain reaction in the midst of a lot of non-fissile (but fertile) material, but I have no idea how you'd figure that out.

I don't understand this. I mean, you realize that nuclear workers are exposed to some (low) level of radiation, right? 20 mSv/year is the limit for nuclear workers in the US. A reprocessing plant that requires human intervention, and results in an exposure of 10mSv/year, might be acceptable, but one that exposes workers to 1000 mSv/year would not be.

That's during normal operation. During emergencies where public health is at risk (a fuel spill for example) most countries allow higher exposures for a short time so workers can "save the day" so to speak. An exposure to 250mSv during an hour-long emergency might increase your lifetime cancer risk, but results in no detectable radiation sickness. An exposure to 25000mSv (25 sieverts) would kill you within minutes. That seems like a big difference!

 

No. For absorbing neutrons there are few things LESS effective than carbon.

“… Chicago Pile-1 (CP-1) was the world's first artificial nuclear reactor.[4] CP-1 was built on a rackets court, under the abandoned west stands of the original Alonzo Stagg Field stadium, at the University of Chicago. The first artificial, self-sustaining, nuclear chain reaction was initiated within CP-1, on December 2, 1942. … The pile consisted of uranium pellets as a neutron-producing "core", separated from one another by graphite blocks to slow the neutrons. …” From: http://en.wikipedia.org/wiki/Chicago_Pile-1

The fast neutrons released by one uranium atom will very rarely cause another to split. Only very slow ones do that frequently. Fermi chose carbon blocks because they would absorb very few, if any neutrons while “thermalizing” neutrons from their release energy (17 Mev as I recall) I.e. they had to bounce off carbon atoms to drop their energy down more than 20,000,000 fold without being absorbed!

It is the new flat geometry when in the catch pan that immediately stops the reaction - nothing else is needed. The heat come not from any continuing chain reaction , but just the decay of some isotopes.

 

Originally Posted by Billy T
"pure enough U233" Do you know what percent of U233 is in the thorium of a homogeneous core reactor as "match" at start up of a new thorium reactor?

Here are my uniformed thoughts/ guesses: The main atoms in the mix are thorium and Fluorine19, the only natural isotope. (Fluorine18 can be made, but has a half live of less than two hours.) Thus fluorine, being much lighter than Thorium will scatter the neutron without absorbing it until it is thermalizied enough to be captured by thorium 232 (assuming that few simply escape the core). I.e. as you told me:
neutron + thorium-232 -> thorium-233 -> protactinium-233 -> uranium-233.

I don't the half lives so don't know the rate at which U233 which decayed to give the original neutron is replaced by a new one. I think that new one decays and gives at least 2 new neutrons. Thus the neutron number doubling time is no more than the average time to step thru the above nuclear chain. less if some neutrons are also released in the above chain or is U233 decay gives three neutrons.

I.e. if willing to wait long enough, it would seem you only need one atom of U233 (or U235) to start the fire. If that is correct, the optimum percentage of U233 is a capital cost question. I.e. you keep buying U233 atoms to mix with each Kg of Thorium until the cost of another one does not reduce the cost of the duration that your capital is tied up in a reactor (more than the atom cost). I.e. buy more U233 to mix in so long as the reactor is not producing enough heat to make electric power it is generating worth more than the rate the interest charges are accumulating.

It is late and I am going to bed in Brazil now, so this may not be correct, or even clearly stated, but hope you get the idea that it may be an economics problem, more than a physics problem. Basic idea is you don't want to wait many years, paying interest on the reactor capital, until the (exponential?) growth in density of U233 has made a fire hot enough to make it profitable.

 

That would be true if you could achieve 100% neutron capture, but alas, I don't think that is the case. (Which is a good thing. If the above were true, any thorium, even natural thorium, would end up in an uncontrolled chain reaction if even one neutron was captured by it!)

 

I would think that any nuclear system which can sustain a chain reaction has this "run-away" problem so all must have some form of control that adjust the neutron production rate to be equal to the absorbion rate. (I.e. like "control rods") I would not expect the thorium reactor to be an exception, but because replacing the U233 (or 235) is a chain of steps, which surely has some inherent delay (more than seconds, perhaps more than minutes?) it should be easier to control than a U235 reactor.

Most do not know that when U235 splits giving 2 or sometimes 3 neutrons, not all are immediately made. I forget the fraction (a few percent, I think, come after many seconds). That makes time for the control system to act. If all produced neutrons came immediately with the split, uranium reactors would be more difficult to control. Basically, without that fortunate natural delay in neutron release, the time constant for exponential growth would just be the time for thermalization to slow neutrons, which have large capture cross section - less than a second I would guess. The inherent time constant for the thorium reactor may be long enough (because of that transmutation chain) to allow a man controlling it time to light his pipe before lowering the control rod(s) a few cm deeper into the core!

In nature, I am sure the thorium does not explode as bombs probably because it is "cover controlled" - i.e. the are many neutron absorbers around. Note in the thorium reactor (other than the man made control "rods"?) the only other atom is F19 and F20 does not exist so F19 can not eat up the neutrons, but is much better than the heavy thorium in slowing them down, until either some thorium atom start the path to becoming U23* or is eaten by the control "rod" or escape from the core. As you know, pure thorium can not sustain a chain reaction, thus needs no control, but in natural deposits may have some neutron source mixed in.

One or both of us need to find out how the thorium reactor really works. I am just guessing with the physics I already know.

---------
* I don't know but suspect there are some detours on that path so not all that start down that chain make it to U233.

 

A "pro-thorium reactor" video here, worth a look: http://pro.oxfordclub.com/METAL49BR...8899&s=19462&u=291435&l=79657&r=MC&g=0&h=true despite being a promotion.

The video fails to mention that India, which has very little uranium, does have at lest 25% of the world's thorium, an lots of cheap labor to mine it. I.e. don't get interested in the stock they are promoting. See instead my post 2 and 15, at: http://www.sciforums.com/showthread...alt-Reactors&p=2769261&viewfull=1#post2769261

 

I can't comment on thorium fueled reactors.

But my understanding is that both the US and Soviets experimented with naval reactors that used molten sodium salts as their heat-exchanger fluid and that both countries ended up abandoning the idea. The engineering proved to be a nightmare as I recall, the things kept developing very nasty leaks, and if the systems were ever allowed to cool down without careful preparation, the molten salt coolant would solidify in the pumps and piping, permanently ruining them.

 

Nope, they used molten sodium METAL, you know, that stuff that if it touches water it explodes? Sodium metal is highly UNstable. The liquid fluoride salts used in MSR are both chemically AND radiologically stable.

As far as I know, the US Navy's LMR was a fast spectrum reactor which have other problems besides just the liquid metal. Molten salt reactors tend to be of a thermal spectrum design which are more inherently stable and safe. And even the MS fast reactor designs tend to be more stable than the LMFR units.

 

Just as liquid flouride salts do when they come into contact with water at their normal operating temperature. (It is a steam explosion rather than a chemical explosion, but there's not much difference when it comes to the damage the explosion does.)

 

Not so. Chemical explosions contain quite a bit more energy than thermal ones, especially when the chemicals are also very hot. And if some of the LM gets spread about it will continue to burn in contact with water while the MS just solidifies.

 

Oh come sodium literally explodes in contact with water chemically, be 700 k just adds to the explosions. Sodium will burst into flames on contact with air let alone water at those temperatures! And yet Sodium cooled reactors use heat exchangers that have to transfer heat from sodium to water, a disaster just waiting to happen!

Reactors should be designed more passively safe, like not use components that would burst into flames on contact with air or explode in contact with the secondary coolant! Molten salt reactors use a coolant that is physically safer then sodium or even water! The salt will not combusted or react with water exothermically, in fact many fluoride salts have low solubility in water. The molten salt reactor can also be made "passively" meltdown proof such that if it gets too hot it melts a plug that drains the fuel into passively cooled neutron absorbing storage containers, that a level of safety physically impossible with conventional reactors.

 

ANY molten metal/salt paired with water in a heat exchanger is a disaster waiting to happen.

Agreed. Which is why something like the AP600 design is, currently, a better choice than a molten salt reactor.

What happens when the molten salt at 600C comes into contact with water in the secondary coolant loop?

 

No its not, in the case of fluoride salts or molten lead as a leak would not result in catastrophic explosions for 2 reasons

1. The primary coolant is not under high pressure
2. The secondary coolant water is already boiled/in steam form.

More so the molten salt can be use with brayton cycle CO2 or Helium turbogenerators such that the molten salt can be pump into the high pressure zone and pumped out again removing the need to high pressure tubing outside the brayton cycle pressure vessel.

No because it to uneconomical to mass produce: the molten salt reactors present the possibly for far small reactors without the need for huge containment domes. Also 3rd gene plus reactors can manufacture their own fuel in a break even manner as well and self process as the molten salts can.

Contact with STEAM and nearly the same temp, not much at all, the steam would likely leak into the salt pipes creating massive voids until reaching a release valve, the reactor would shut down as the steam pressure pushes the fuel out of the reactor, the "released" steam could in fact be piped to an emergency cooling loop made to such a contingency. Without reactor heat the and all the fuel pushed into containment vessel the steam would cool to water and depressurize. Actually steam or water are not likely coolants for molten salt reactors because of likely need to operate at temps beyond 500 C. CO2 or Helium would be the coolant.

Please Register or Log in to view the hidden image!

http://www.softinway.com/news/articles/article.asp?ArticleID=95&Page=1

 

With modern HEXs, mixing of the fluids is almost impossible. The issue is whether the fluid leaks out of the HEX to the outside. If that happens, liquid sodium is a major problem. Liquid salts will tend to just solidify and plug the leak.

The AP series is much better than older designs and is walk away safe... for 72 hours. MSRs like Liquid Fluoride Thorium Reactors (LFTRs) are walk away safe... period. Ok, you do have to do something with the dumped fuel within a decade or so, but other than that...

Same answer as before. Modern HEXs don't let that happen. Leakage OUT of the HEX, yes, between sides, no.
But even if there was one, there would be a clean, thermal energy steam explosion, and the plant would shut down. No combustible liquids thrown around the place. Some solidified salt and that's pretty much it. Replace the HEX and start up again.
I would point out that despite HUGE amounts of money spent on LM reactors, there has yet to be a successful one based on an alkali metal. Lead? Yes. Sodium? No.