How Might LFTRs Fail?

All devices have failures. We have to consider the possible failure methods of any nuclear reactor design.

We don’t know the ways LFTRs would fail, just like we didn’t know decades ago that a 9.0 earthquake and tsunami could knock out all diesel backup generators leading to loss of coolant accidents, or how engineers’ recommendations would be ignored in construction of sea walls, at Fukushima-Daiichi.

But we do know many ways that current reactors fail that LFTRs can’t — we know LFTRs can’t fail from core meltdowns, loss of coolant, or high pressure explosions.

Components that fail, or that sensors detect are out of specifications, would be replaced. Molten fuel that can be easily drained from the reactor, and atmospheric pressure, make replacing components easier than in LWR. Multiple small reactors supplying a region, all normally operating below peak, would eliminate power disruption during maintenance.

Since LFTRs could be mass-produced and installed everywhere a 100MW – 1000MW power source is needed, small problems such as leaking pipes would be much more frequent (though less likely per gigawatt).

Current reactors have leaking pipes, often. Easy to detect, easy to fix, easier to clean up in a LFTR since radioactive material is trapped in a salt that cools to a solid, instead of leaking radioactive water. At atmospheric pressure and with chemicals that don’t react to air or water, repairing the pipe would be the same or easier than for a solid-fueled reactor.

Fire a cannon at it, and the fuel would spill out onto the building floor and collect in the drain tanks. Damages the reactor, you’d have to repair it, but not a radiation disaster. Loss of electric power generation would be annoying to the city; but most cities would have multiple smaller reactors instead of one larger one.

LWR needs fuel for 18 months, so need to have “excess reactivity”. Since all MSRs can be fueled daily (or even continuously), the amount of uranium and all other radioactive materials in a MSR at any time would be less than in any solid fueled reactor. For a LFTR there would be no radioactive fuel stockpile (thorium is pretty harmless unless ingested). No MSR has “spent fuel storage”, just short-term storage of fission products (83% of them are below background radiation in under 10 years).

(MSRs designed to consume LWR spent nuclear fuel would be located at the waste storage facilities, which are heavily guarded.)

“MSRs will have an increased potential [compared to solid fuel reactors] for small-scale radioactive materials leaks because the highly radioactive fuel material is liquid and comparatively more accessible than solid fuels. The leak probability will be increased for on-line reprocessed reactor design variants as a result of more intensive fuel salt manipulation. The fuel salt reprocessing manipulation will need to take place within a hot-cell type environment, providing an additional containment structure within the primary reactor containment.” Fast Spectrum Molten Salt Reactor Options, Oak Ridge National Laboratory, July 2011

(MSR leaks would quickly cool to a stable solid, rather than the radioactive water leaks in a water-cooled reactor.)

Terrorists with plastic explosives, to send radioactive fuel droplets into the air? They would quickly cool to glass-like solid, wouldn’t stay in the air (too heavy, about 2.4kg/liter), wouldn’t dissolve in water, wouldn’t flow well in water, but of course would have to be collected. Since there is much less radioactive material in a LFTR than a PWR, just enough fuel to sustain fission, and radioactive fission byproducts would be regularly removed, there would likely be injuries only for people close enough to be hit by explosives spraying shrapnel and hot liquid salts. Installing LFTRs underground would make placing those bombs much harder to do. Put the electric turbine above ground, and in tours of the facility call it “the reactor” so most terrorists would strike the wrong device.

Insiders who have access to the reactor, and can install some high-tech “fire hose” that would handle 700° C highly radioactive liquids and gamma rays, for at least a few minutes?

There would be heat burns from very hot liquid hitting people. I don’t know what happens if radioactive FLiBe gets inside you, from drops landing on you; seems that would be the worst. So we would want to protect the facility against people installing explosives or other things.

Looks like the point terrorists would target is the attached equipment to use heat, air and CO2 to make methanol and gasoline; “It’d be like blowing up a big tank of gasoline, we’d get a big ‘splosion and be in the news!”. There would need to be an above-ground truck loading station or gas pump, easy to attack. Of course, the reactor would be shut down, and methanol/gasoline production would cease.

There are dozens of terrorist targets that would be easier and less expensive to attack. It’s easier and more dramatic to blow up a school bus or bridge or restaurant. But we still have to look at ways to prevent attacks.

Since all MSRs could be factory assembled, sensors for security and reliability can be installed throughout the reactor. These could communicate security status via satellite, electric power lines, local alarms. Tracking devices can be in each component, each batch of fuel, each batch of fission products. Molten Salt Reactors can be extremely difficult targets for theft or terrorists.

8 thoughts on “How Might LFTRs Fail?”

  1. Two things: @ TEA, it is noted in one video that the easiest way to get Th is to go to a Florida beach and scoop it up ___ and ___ Above it is stated that Th isn’t toxic unless ingested. Seems to imply that Florida beach sand is toxic. Comments? Thanks

    1. @John Shilling: Pure Th dust is different than Th chemically bound to the other components of sand. Sand will very likely pass right through you. But don’t eat sand.

      One use of Thorium was for early medical imaging, where tiny particles of thorium was suspended in a liquid and injected into organs. We’ve stopped doing that, knowing that radioactivity can damage cells. But the thorium in a little sand isn’t going to cause much harm (from radiation, I mean; who knows about the cutting edges). Thorium is among the Least radioactive elements.

      Beach sand is easy source of Th, but not in industrial quantities. The concentrated sources of thorium are the same places as rare earth elements, mined for use in TVs, cell phones, headphones, windmills, etc.

  2. Many parties developing and proposing LFTR seem naive to the risks of LFTR and their numerous potential failure modes. These LFTR are potentially just as dangerous as contemporary solid-fuel reactors. Even solid-fuel accelerator driven Thorium reactors (as proposed by Rossi) will become critical if their structural integrity of their coriums are lost. [Accelerator driven reactors can’t sustain fission without the accelerator supplying neutrons; it would be stopped before the fuel could melt, no corium could form. — George] At Fukushima Dai’ichi, the coriums are still melting and lots of spatially local criticalities are occurring as the coriums form lava tubes and melt down, eventually to reach the water table at Fukushima Dai’ichi. All the present time, radiation is leaking into the Pacific Ocean from Fukushima Dai’ichi and all life in this Ocean will soon have been destroyed. Hundreds of million of years of biological evolution will be ruined by just a few years’ of radioactive leaks from Fukushima Dai’ichi, as delicately evolved DNA material is butchered by the released radiation. Sorry, not enough fuel in the reactors to cause that much damage. He’s repeating nonsense found on the Internet

    If we have LFTR’s everywhere, there will be lots of accidents and breakdowns, as presently proposed designs are not adequate.

    1. I approved the rant “Many parties developing and proposing LFTR seem naive…” because it is so useful in showing how to evaluate safety.

      Molten Salt Reactors such as LFTR can’t have corium. Wikipedia “consists of nuclear fuel, fission products, control rods, structural materials from the affected parts of the reactor, products of their chemical reaction with air, water and steam, and, in case the reactor vessel is breached, molten concrete from the floor of the reactor room.” MSR has no fuel rods (has molten fuel dissolved in the salt). MSR has no need for control rods (thermal expansion of the low-viscosity fuel/salt strongly regulates the fission rate, and faster than any control rods); and if you need to stop the reactor (e.g. maintenance) operators would simply drain the fuel salt into storage tanks where fission is impossible.

      In LWR, a nuclear meltdown is from loss of water coolant leading to control rods heating above their melting point. In MSR, loss of coolant is impossible. The fuel is molten, chemically bound to the molten salt coolant (fluoride salts chemically bind very strongly to uranium and most fission products). The salt is minimum 400C below its boiling point (the hottest operating temperature of LWR is about 350C, for reference).

      In LWR, overheated fuel and fuel rods can melt through the reactor vessel. In MSR, the reactor materials can withstand the hottest the reactor can get in any circumstance.

      At Fukushima:

      “the coriums are still melting” is nonsense, fission stopped years ago, the most radioactive fission products have already decayed to a fraction of their original radioactivity. The concrete in the reactor floor makes the molten fuel less dense and also stops neutrons; fission doesn’t occur in spread-out fuel, not enough neutrons. It may still be molten; it isn’t melting more.

      “lots of spatially local criticalities are occurring as the coriums form lava tubes and melt down, eventually to reach the water table” — also nonsense. “Of the 10.2 metres of solid concrete that makes up the floor of the reactor building, the corium is thought to have melted and mixed with the first 70 centimetres only. The natural spreading and expansion of the corium, plus the addition of compounds of concrete, would have reduced the intensity of the heat produced until it reached an equilibrium and solidified in place.”

      “all life in this Ocean will soon have been destroyed” — There is far more uranium in the oceans than in all the world’s spent fuel rods combined. Far more radioactivity from radon (in natural gas) and many other sources, than all the world’s nuclear reactors. Far more radioactive material in a single coal plant’s ash ponds, than has been released by all nuclear power plants combined — the one coal ash pond collapsing into a river in North Carolina recently, released more radioactive material (found in all coal) into the water than Fukushima did.

      We may kill off all of most species of plankton, not from nuclear power, but from increased carbonic acid dissolving their microscopic cell walls. That carbonic acid is formed when CO2 enters the oceans. The increase of CO2 over what had been stable for centuries, is from burning fossil fuels, making cement, and deforestation. Look at all the possible actual causes of the problem, not imaginary ones.

      “solid-fuel accelerator driven… become critical if their structural integrity of their coriums are lost” — Maybe he meant the reactor core would lose structural integrity if the coolant is lost? But an accelerator-driven reactor is deliberately low density fuel, so there can not be fission if the accelerator is not providing neutrons; cut electric power to the accelerator, and fission stops long before the core would approach melting point of any materials.

      Why did the Fukushima LWR reactors lose coolant? Because TEPCO didn’t put the diesel backup electric generators where they couldn’t be flooded (unlike other nuclear operators), and because TEPCO decreased the natural seawall (unlike other nuclear operators). The Onagawa reactors, closer to the center of the earthquake, with proper diesel backup and seawall, is undamaged. LWR design, as presently implemented by almost operators, is adequate (to follow Tim’s phrasing).

      LFTRs have no risks from water (no water used at all in any MSR). MSRs have no risks from water pressure. MSRs have no risk of chemical explosion. MSRs have no risk of melting their containment vessel, there is no way for the fuel to get dense enough to get nearly hot enough. With no high pressures, a pipe leak in a MSR would be measured in square meters, and quickly cool to solid, with the fuel and fission products trapped, and be easily cleaned up.

      If we have LFTRs everywhere, we will have Fewer accidents and breakdowns, as the nuclear industry has had fewer accidents over 50 years than the coal and oil industries have each year, per gigawatt-year electricity.

  3. George. It is respectfully inappropriate to refer to my comment as a “rant”. I have spent thousands of hours studying ORNL, CNRS, patent and other scientific literature regarding MSR and LFTR.

    Firstly, there is general mainstream media coverup about the severity of the Fukushima Dai’ichi situation. Detection of radioactive isotopes presently leaking from the Fukushima plant indicate that the nuclear reaction is on-going. Radiation inside what is left of reactor buildings 1, 2 and 3 is so severe that humans would be killed within minutes, even shorter, if they attempt to work in close spatial proximity. TEPCO admits that it does not know the status of the coriums, so your assertion that the concrete floor has halted the circa 5000 oC coriums as they bore down is without actual support. Steam has been observed from ground fissures around the reactors 1, 2 and 3 rather indicating that something is now happening deep underground.

    The nuclear industry has, over the years, come with many false statements:
    (a)”… electricity too cheap to meter”;
    (b)”… inner containment cannot be breached” (i.e. NRC assumption when licensing new nuclear plant);
    (c) Fukushima disaster has not resulted in any loss of life;
    and so on.

    If there are modular LFTR around (in 50 MW to 100 MW class), these will need supervision. There will be accidents. In the continuous chemical processing of the nuclear fuel, there are multiple areas where serious accidents could occur, and your comment does not address this. A leak from a Thorium LFTR, in view of the hard Gamma radiation flux, would be specially hazardous. Moreover, neutron embrittlement will occur and associated cracking, even with advanced materials. In present nuclear plant in the USA, especially older plant from the 1960’s and 1970’s, embrittlement and cracking of metal components is a big problem.

    Unless new materials are developed, for example Silicon Carbide ceramic composites, it seems that many of the components for a LFTR would be Hastelloy-N and/or graphite. These materials are known to fail from the conventional nuclear industry.

    I kindly suggest that you regard reader comments with the respect they deserve rather than merely referring to them as “rants”.

    1. @DrTim:

      TEPCO blew it, I’m not going to refute each little detail you give about them. Of course Fukushima disaster resulted in loss of life, but there are no documented cases of human life, and very low chances of any human deaths. I recommend we move away from LWR to Molten Salt Reactors, largely because there are less ways stupid management companies can get it wrong. Obviously “inner containment can be breached” if management is so stupid to keep diesel backup generators in the basement while removing natural sea wall. With all forms of nuclear (and chemical) plants, look to the actual source of the problems.

      Yes, there is modeling and evidence there the reactor cores are likely to be, even though TEPCO doesn’t know precisely (e.g. photographic evidence). The cement floor is thick, and the cement prevents fission, and the more the core spreads the faster it cools. Of course radiation inside the buildings is high, and the radiation is almost all from short-term fission products (half life of milli-seconds to 1 year), already a fraction of the original radiation. The biggest problem now is the water, from TEPCO not diverting ground water from the hillside they removed, away from the reactor.

      Today’s “nuclear industry” that you mention never talked about “too cheap to meter”, the Light Water Reactor was never capable of that, it was a modified nuclear submarine propulsion reactor modified to demonstrate electricity from controlled nuclear fission was possible; water and steel were common materials for testing with, never meant to be a utility scale reactor! “To cheap to meter” was from before Congress got their hands on the reactor designs; if we had gone with Molten Salt Reactors, or other designs the Atomic Energy Commission recommended in 1960s to President Kennedy and Congress, we would likely have cheap non-polluting electricity now. (Who made sure Congress decided on the most complex reactor, that physicists knew would have loss of coolant accidents? Follow the money…) Today’s Nuclear Regulatory Commission is more accurately the Light Water Reactor Regulatory Commission. We can have nuclear reactors, with all costs included from mining through decommissioning, cheaper than coal (including health and pollution costs).

      The gamma radiation from a MSR major accident (the reactor vessel is broken by bomb or cannon ball, since there is no internal pressure that could do so) would be stopped simply by having: a) drain the fuel in minutes to tanks with radiation shielding, and b) reactor underground where a few meters of dirt stop gamma rays.

      Hastelloy-N was certified in 1960s for several years in nuclear reactors, long enough for reactors to be very profitable before being rebuilt. Materials testing hasn’t yet been done on those new materials, which are expected to last longer. LWR is taken out of service about every 18 months for refueling; MSR could work taken out of service every few years for a week to replace the reactor parts (and another reactor could be connected in the same site before taking the that one off-line, minimal power disruption).

      Graphite can be used in designs for rapid replacement (e.g. LeBlanc), less frequent than LWR refueling. Graphite is only needed as a moderator, in modern designs, not for keeping chemicals separate; and there is atmospheric pressure; doesn’t matter if small cracks appear before replacement. Fast-Spectrum Molten Salt Reactors would not use any graphite (no moderator); we don’t know the chemistry of the salts needed for FSMSR as well as for FLiBe, more materials testing needed first. Remember, except for neutron bombardment, few of the problems in LWR apply, since the physics and the chemistry and the pressure is so different.

      Of course LFTRs or any other nuclear reactor should have supervision, even if mostly electronic surveillance and internal sensors. MSR could be made in factories with higher safety and monitoring standards than used for the oil/coal processing industries, and there would be far less chemicals (total under 1000kg/gw-yr) than any LWR, coal, natural gas plant.

  4. Hello George

    I hope that our public dialogue helps progress on LFTR and MSR. I am personally convinced that a suitable form of MSR is vital for transmuting, and therefore safely disposing of, enormous contemporary stockpiles of nuclear waste. The present stockpiles are quite insane, for example from LWR. It is the most pressing problem. In the USA, there is 77000 tonnes of such waste, and in Japan 17000 tonnes of such waste. This is environmentally no trivial problem that has been created during the past 50 years.

    I also kindly draw your attention, and that of readers to our blog, to the news aggregator ( which shows quite how dreadful contemporary nuclear technology is in practice, and why improvements are so very necessary, for example via MSR and LFTR. However, we need to be realistic, and accidents will MSR and LFTR will happen in future.

    A common error is “… if the reactors overheat…”, the contents are drained via the melt plug into a safety dump tank. No, no, no I say! For example, a terror attack or unusual fracture of the separating wall between primary and second LFTR circuits can cause fragments which could block the melt plug and prevent safe draining of the reactor. Such a mundane situation could theoretically occur, or an earthquake causes fracture of the pipe after the melt plug, so that the contents of the reactor create a terrible general mess beneath the reactor, rather than ending up in the dump tank. I fully agree with you that such a scenario would be far less disastrous in comparison to what has happened at Fukushima Dai’ichi. However, it is worth bearing in mind that the reactors at Fukushima Dai’ichi were very large capacity plant and the afterheat of radioactive decay is very considerable. A modular 50MW to 100MW LFTR would result in less afterheat, and could potential survive via air convection cooling in an event of a serious accident.

    Thus, MSR and LFTR need considerable safety protection, in addition to their intrinsic safer characteristics than LWR and BWR. For example, dumping Boron into the melt in an event that the melt plug fails due to blockage would slow down the reaction, so that something could be done to ameliorate the situation.

    It is also worth mentioning that some of the proposed accelerator-driven solid-fuel Thorium reactors have control rods, because the potential capability of the fuel self sustaining criticality changes as the fuel composition changes during the lifetime of the fuel. Where some of the fuel rods include a MOX/Thorium mixture, a structural failure of the reactor core could bring fuel rods sufficiently spatially close together in gross failure conditions (e.g. a bunch of rods mutually touching) that a local uncontrolled criticality could potentially occur. Thus, even with accelerator-driven Thorium reactor designs, there are worst-case dangerous failure modes. Harrisburg disaster clearly teaches that one should not assume that the structural formation is maintained in an event of a meltdown.

    Having stated above, reaction by-products in a Thorium LFTR tend to quench reaction if not continuously removed, because they are neutron-grabbing in general. Moreover, as the fuel is a melt, it cannot become more spatially compact than under its normal operating conditions, such that a very gross failure where the molten fuel gets ejected and thrown about within a building (secondary containment of building outer walls), there is far far less of a risk of a criticality (although it would be terrible mess to clean up). In contradistinction, a solid fuel rod system in a gross structural failure can result in fuel rods being crowded together with potentially disastrous consequences if local criticalities occur.

    Thus, it is clear, also when many other factors are taken into consideration and not discussed above, that LFTR and MSR should (!) be safer than conventional LWR and BWR, but the hazards of operating a LFTR and MSR should not be underestimated.

    Here ends my “rant” !

    Kind regards

  5. When I look at the two fluid Flibe reactor I have two reactions. First I love the internal processing of waste products, the chemical kidney. My second reaction is more critical, looking at a the Flibe design I say, that’s a lot of plumbing, what happens if any part of it fails or needs replacement? Every tank, every separation column, every pump, heat exchanger, two loops driven by pumps. All of it has to work well in severe conditions, a caustic, high temp, high radiation environment. I’m not saying it can’t be done but I think that the biggest challenge to constructing it will be material challenges. To Flibe I say good luck.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>