The world as I see it
RSS icon Home icon
  • Accelerator-Driven Thorium Reactors, or An Easy Way to Eliminate Surplus Population

    Posted on September 10th, 2013 Helian 7 comments

    The Daily Telegraph has just taken thorium wowserism to a whole new level.  According to the title of an article penned by International Business Editor Ambrose Evans-Pritchard, Obama could kill fossil fuels overnight with a nuclear dash for thorium.  Continuing in the same vein, the byline assures us that,

    If Barack Obama were to marshal America’s vast scientific and strategic resources behind a new Manhattan Project, he might reasonably hope to reinvent the global energy landscape and sketch an end to our dependence on fossil fuels within three to five years.

    And how is this prodigious feat to be accomplished?  Via none other than Nobel laureate Dr. Carlo Rubbia’s  really bad idea for building accelerator-driven thorium reactors.  It would seem that Dr. Rubbia has assured the credulous Telegraph editor that, “a tonne of the silvery metal produces as much energy as 200 tonnes of uranium.”  This egregious whopper is based on nothing more complicated than a comparison of apples and oranges.  Thorium by itself cannot power a nuclear reactor.  It must first be converted into the isotope uranium 233 via absorption of a neutron.  Natural uranium, on the other hand, can be used directly in reactors, because 0.7 percent of it consists of the fissile isotope uranium 235.  In other words, Rubbia is comparing the energy potential of thorium after it has been converted to U233 with the energy potential of only the U235 in natural uranium.  The obvious objection to this absurd comparison is that the rest of natural uranium is made up mostly of the isotope U238, which can also absorb a neutron to produce plutonium 239, which, like U233, can power nuclear reactors.  In other words, if we compare apples to apples, that is, thorium after it has been converted to U233 with U238 after it has been converted to Pu239, the potential energy content of thorium and uranium is about equal.

    As it happens, the really bad news in the Telegraph article is that,

    The Norwegian group Aker Solutions has bought Dr. Rubbia’s patent for an accelerator-driven sub-critical reactor, and is working on his design for a thorium version at its UK operation.

    In fact, Aker has already completed a conceptual design for a power plant.  According to Aker project manager Victoria Ashley, the group needs a paltry $3 million, give or take, to build the first one, and another $150 million for the test phase to follow.  Why is that disturbing news?  Because the U233 produced in these wonderful new reactors will be ideal for producing nuclear weapons.

    In fact, it will be even better than the “traditional” bomb materials; highly enriched uranium (HEU) and weapons grade plutonium.  The explosion of a nuclear device is produced by assembling a highly supercritical mass of fissile material, and then introducing a source of neutrons at just the right moment, setting off a runaway chain reaction.  The problem with plutonium is that it has the bad habit of occasionally fissioning spontaneously.  This releases neutrons.  If such a stray neutron were to happen along just as the bomb material became critical, it would set off a premature chain reaction, causing the device to “fizzle.”  As a result, plutonium weapons must rely on a complicated implosion process to achieve supercriticality before the stray neutrons can do their dirty work.  Implosion weapons are much more technologically challenging to build than the gun-assembled types that can be used with HEU.  In these, one subcritical mass is simply shot into another.  However, the required mass of HEU is much larger than the amount of plutonium needed in an implosion-assembled weapon.  As it happens, the amount of U233 sufficient to build a nuclear device is about the same as the amount of plutonium, but spontaneous fission is not a problem in U233.  In other words, it combines the plutonium advantage of requiring a much smaller amount of material, and the HEU advantage of being usable in gun-assembled weapons.

    Why, then, you might ask, are we even giving Rubbia’s idea a second thought?  Because of people like Professor Egil Lillestol, who, Evans-Pritchard helpfully informs us, is “a world authority on the thorium fuel cycle at CERN.”  According to Lillestol,

    It is almost impossible to make nuclear weapons out of thorium because it is too difficult to handle.  It wouldn’t be worth trying.

    Rubbia has made similar statements, based on the same “logic.”  The rationalization for the claim that U233 is “too difficult to handle” is the supposed presence of U232, an isotope of uranium with a half-life of about 69 years, one of whose daughters (elements in its decay chain) emits a highly energetic and penetrating, and hence deadly, gamma ray.  In fact, avoiding the production of U232 in accelerator-driven reactors would be a piece of cake.  Rubbia and Lillestol must know this, making it all the more incomprehensible that they dare to foist such whoppers on unsuspecting newspaper editors.

    Only one neutron absorption is needed for the production of U233 from naturally occurring Th232.  Two are needed to produce U232.  Thus, one way to keep the level of U232 within manageable levels is to simply extract the U233 before much U232 has a chance to form.  However, there’s an even easier way.  Very energetic neutrons, with energies above a threshold of around 6 million electron volts, are necessary to produce U232.  Not many fission neutrons have that much energy, and slowing down the ones that do is simple.  Simply pass them through a “moderator” rich in hydrogen or some other light element.  Think of billiard balls.  If one of them going at a good clip hits another dead on, it stops, imparting its energy to the second ball.  Neutrons and the proton nuclei of hydrogen atoms have nearly the same mass, so the same thing can happen when they collide.  A fast neutron will typically lose a large fraction of its energy in such a collision.  In other words, the “secret” of avoiding the production of dangerous levels of U232 is as simple as passing the neutrons through a layer of hydrogen-rich material such as paraffin before allowing them to interact with the thorium.  All this should hardly come as a surprise to people like Rubbia and Lillestol.  It’s been old hat in the literature for a long time.  For a more detailed treatment, see, for example, U-232 and the Proliferation-Resistance of U-233 in Spent Fuel, a paper that appeared in the journal Science and Global Security back in 2001.

    In other words, the idea that “it is almost impossible to make nuclear weapons out of thorium” is a pipe dream.  That does not necessarily mean that thorium technology should be rejected root and branch.  It will always be necessary to exercise extreme care to insure that U233 isn’t diverted for illicit purposes.  However, managing the risk will be considerably easier in “conventional” thorium breeders, which rely on assembling a critical mass to supply the necessary source of neutrons.  Such reactors have already been built and successfully operated for years.  The U233 they produce will always be mixed with highly radioactive fission products, and can also be “denatured” by mixing it with U238, from which it cannot be separated using simple chemistry.  Such reactors would produce few of the transuranic actinides that are the main culprits in nuclear waste, potentially requiring it to be stored securely for millennia.  They could also consume the actinides produced in the current generation of reactors, so that the remaining waste could potentially become less radioactive than the original uranium ore in a few hundred years, instead of many thousands.

    If, on the other hand, the accelerators necessary to provide the neutron source for Dr. Rubbia’s subcritical facilities were to become readily available, they would be much easier to hide than conventional reactors, could be configured to produce U233 with almost no U232 contamination, and with much less radioactive fission product contamination.  In other words, they would constitute an unacceptable risk for the proliferation of nuclear weapons.  One must hope that the world will wake up in time to recognize the threat.


    7 responses to “Accelerator-Driven Thorium Reactors, or An Easy Way to Eliminate Surplus Population” RSS icon

    • Thomas Jam Pedersen

      Thanks for a well written article and good links. The world need more of that… Please

    • It is not really comparing apples with oranges.

      As tons of fuel in, and quantity of electricity out are the same units.

      Sure you need different machines to do the conversion.

      Or another way look at the (amortised) cost (both in money and environmental) of Th vs U vs Coal to power X number of homes.

      In this area Th would be a win.
      (As a side note coal power plants are responsible for dumping far more radioactivity into the biosphere than Nuc power (even when including accidents). When you need to burn 3,500,000 tons of coal (compared to 200 tons of U fuel) even the tiny amount of radioactive elements in coal adds up)

      Also the risk of ‘gah you could use it to build a bomb’ is over played here.

      Here I will add in the piece that may have been missed from the article:

      There is a difference between an accelerator driven sub-critical thorium reactor and using an accelerator to turn thorium into U233 for bomb.

      A sub-critical reactor uses an accelerator to keep a chain reaction going. Most of the reaction is U233 splitting, releasing (most of the time) 2 neutrons. Some of those neutrons will convert Thorium to U233 (ready for the next round in the chain reaction). Some will hit some previously converted U233 (e.e., the start of the next round). But some will be lost out the walls and some will end up creating U232. The accelerator is there to make up for those losses. But you still get U232 produced.

      The U232 can’t be used in a bomb (wreaks the firing and flight electronics, and humans when building and handling it)

      And it is not practical to remove the U232 from the U233, as it is be easer and cheaper to remove U235 from regular Uranium ore.

      Now conceivably you could use an accelerator to convert thorium to U233 with little U232 contamination. However a machine in that configuration, would require an accelerator that was thousands of times bigger(my back of the envelope calculations point to 10K times the size). With the power requirements to match. So you would need a small fleet of power plants to run it.

      I would bet that it would be more expensive to run that machine, than using one of the more conventional ways of building a bomb.

      Now if someone had that much money to throw around. They would have the money, to work out how to make an implosion (rather than the simple gun type).

      So the technology could be used for evil it would not be simple, cheap or practical.

      I guess with that out in the open it makes for al less alarmist post. It is a pain when the facts get in the way of a good story or bias…

      now forsure there will be risks with any reactor. But you need to make realistic estimates of what they are and weigh them up (along with the other costs) against the benefits (along with the alternatives) and go form there.

      Facts and the full story help with that.

      (Another fun fact: More people die (falling when installing mostly) from solar power than nuc power(radiation from accidents included) (when compared per GWh) Similarly for hydro & coal (which is the worst; mining is dangerous). )

      Hmmm… Interesting quote (of the day) by the way:

      We seldom find any person of good sense, except those who share our opinions.
      –Francois de La Rochefoucauld

    • You should actually learn something about nuclear engineering before you start lecturing others about it, Darcy.

      A. Claiming that a ton of thorium is the equivalent of 200 tons of uranium is patent nonsense. It is indeed comparing apples and oranges. You can make a pile of thorium as big as the planet, and it will never go critical. It is useless as a reactor fuel unless it is first converted to a fissile form of uranium, U233. However, the U238 that makes up the lion’s share of natural uranium can just as easily be transmuted into the fissile isotope plutonium 239. It is utterly absurd to claim that a ton of U233 is the equivalent of 200 tons of Pu239.

      B. The notion that it will be perfectly easy to breed enough U233 to run a reactor, but out of the question to breed the much smaller quantity needed to make a bomb is patent nonsense.

      C. You have simply ignored everything I’ve said, and everything set forth in the papers I’ve linked about the ease of avoiding U232 contamination of the U233 produced from thorium, and simply repeated the false claim that U232 will make U233 useless for weapons production. This argument has never been anything but an easily refutable red herring. Try actually reading the papers and doing the math.

    • I wouldn’t say that it’s not possible to come up with a way to get bomb-making U233 from a thorium-“fueled” reactor, but I think it’s important to note the difficulty of doing so, relative to other, well-known (and used by states) methods to come up with enough of the right type of material to make some sort of bomb.

      If we are going to say that any possibility, however remote, that any quantity of bomb-grade material can be produced, is going to “nuke” a particular technology or approach that is otherwise promising, then I think we need to go back to our risk-assessment drawing board. Every year that we avoid working in earnest on these types of next-gen reactors is a year that results in our existing energy production killing 100’s of thousands of people (from coal pollution, etc.), and brings us a year closer to our CO2 emmissions tipping point.

      I happen to believe that MSR-based designs hold more promise (and we have more experience), and it sounds like you do also. However I worry that the popular press will pick up on this type of post and over-generalize the issue so the public ends up being told that thorium-based energy is a bad idea…

      My (non-expert) $.02.

    • I know of no reason why it should be more difficult to get bomb-making U233 out of a thorium breeder than it is to get plutonium out of a U238-based breeder. As I pointed out earlier, the U232 issue is a red herring. Excessive U232 production can be easily prevented. I think the MSR is an extremely attractive concept. The U233 can be denatured with other isotopes of uranium, and MSR’s would be far more difficult to hide than accelerator-based sub-crits intended to circumvent international controls. They would consume transuranic actinides, and could only produce them through a long neutron absorption chain, and could even destroy some of the longer-lived fission product waste.

      However, they will produce U233, a very attractive bomb material, and that danger cannot simply be ignored. There will always be alarmists who like to strike poses as noble saviors of the world, and I’m sure it will always be possible to find many of them exaggerating the dangers of nuclear while ignoring those of the alternatives. Nuclear will always be an easy target for cheap shots. However, I doubt that my little blog will be of much use to them as a source of agitprop.

      I agree that we should start building thorium breeders now, not necessarily in large numbers, but enough to insure that we can master the technological challenges and not be left in the dust by countries like India and China. However, we need to do it right, with eyes wide open to the potential dangers, and that means passing on Dr. Rubbia’s sub-crits.

    • Hi.

      I am puzzled by some of this though I am the first to acknowledge that I do not have a nuclear background, just a curious mind and a minute amount of RaSO4 to amuse myself with.

      My main query is how you propose to limit U232 production by simply converting fast neutrons to thermal neutrons.

      When using a RaBe system to initially start breeding Th233 to decay to PA233 and then to U233, you have to slow the neutrons down or next to nothing happens at all anyway. So the hydrogen moderation is already there surely.

      Yet using this very simple method of moderation, either U232 is being generated OR the U233 seems to be considerably more active than it should be.

      As for the 200 – 250 times more energy dense than uranium, I agree this is misleading to some extent – however my basic comprehension of this, was that is was largely a result of the more efficient extraction of energy from a molten salt – compared to the solid pellets that self-poisoned quickly – and thus need to be replaced more frequently.

      So perhaps although the statement that it is “200 times more energy dense” is incorrect, is it fair to say that around 200 times more energy can be effectively extracted before the fuel needs to be replaced?

      Either way, even if just from an abundance aspect, I think Thorium holds more promise than uranium (as in mined rather than bred.

      I do wonder though, that if the thorium cycle uses molten salt, could a modified fuel “pipe” not have a centre feed of Th232 and then, as the fuel is liquid, would it not permit top and bottom drains of different isotopes?

      ** for your eyes only, please delete next two paragraphs before plonking on site – if you wish to plonk on site at all.

      As for nuclear weapons, I think my concern would be more along the lines that armed with little more than a decent RaBe seed and a LOT of time, a dirty bomb consisting of very little more than the seed, moderator (bog standard parrafin wax at a push) – and a sack of thorium nitrate mixed with finely powdered Beryllium could be produced.

      Add some powdered lithium too and the fast neutrons and thermal neutrons can play with the Li6 and Li7 to generate some a little tritium to further drive the Be to make neutrons and hasten the breeding of Th233,Pa233,U233 & U232.

      After a while, you are going to have an exceptionally cheap and very unpleasant volume of nastiness, which, used with conventional explosive would make a real sod of a clean-up job.


      If you have time to respond to limiting production of U232 when just trying to produce U233 I would be most interested.

      Kind Regards

    • Mark,
      To produce large amounts of U233 with minimal U232 contamination you need lots of slow neutrons, with little or no “hard” or “fast” component. In my article I considered two potential sources; the proposed high flux accelerators and “conventional” nuclear reactors, meaning ones that require the assembly of a critical mass. If you were just trying to produce enough U233 for a bomb, the reactor wouldn’t have to be a breeder. However, it would necessarily produce highly radioactive and easily detectable fission products, and would be relatively difficult to hide because of the size, detectable fission signature, etc. Using the accelerator-driven approach, you wouldn’t need to assemble a critical mass, and could potentially use thorium alone. The U233 could be extracted chemically as it was produced, minimizing parasitic fission and the resultant fast neutrons, fission products, and U232. In other words, the accelerator-driven approach isn’t the only one. It is just the simplest one and the easiest to hide.

      There are, of course, other potential sources of slow neutrons, but none that I am aware of are capable of producing enough U233 for a bomb anywhere near as quickly. or within a reasonable time at all, for that matter.

    Leave a reply