Liquid fuel molten salt thorium reactors

thorium fluoride melting in a laboratory burner flame

Last week I wrote a little bit about how fertile thorium can be turned into reactor fuel. A discussion of how a thorium reactor might differ from a uranium reactor seems a reasonable follow-up topic.

The reactors I worked with in the Navy were light-water pressurized water reactors. Like virtually every reactor we have on Earth our reactor core was filled with fuel elements – uranium fuel was sealed into a zirconium cladding and these metal fuel elements were arrayed in the reactor vessel in a precise pattern. As the uranium fissions it produces highly radioactive fission products – these are trapped within the metal fuel elements and if the metal melts (as happened at Fukushima) all of this radioactivity is released.

The reactor vessel was filled with light water – normal water (heavy water consists of water in which some of the hydrogen atoms are replaced with a heavier isotope, deuterium).  But it’s not enough to simply sustain a fission reaction, to generate power it’s also necessary to find a way to use the energy produced by that fission. In the case of a reactor plant that means heating water to boiling and using the steam to spin a turbine – since this happens most efficiently at high temperatures it helps to have the coolant as hot as possible. And if we are going to try to heat water to high temperatures then that water will have to be at high pressures to keep it liquid. In turn, this requires building components that are not only able to withstand temperatures of several hundred degrees, but also pressures of up to a few thousand pounds per square inch. And at these extreme pressures even a small flaw can cause problems. Not only that, but water at such high pressures will be forced from the reactor plant at tremendous speeds, eroding the pipe and making the hole larger, as well as emptying the plant all too quickly. Not all water-cooled reactors are pressurized water plants – boiling-water reactors are another “flavor” of light-water reactor – but all water-cooled reactors operate at high pressures and temperatures and all suffer from the stresses and worries that go along with those conditions.

Back in the 1950s an alternate reactor design was tested – one in which the high temperatures didn’t require concomitant high pressures. Not only that, but this reactor also eliminated the need for high-precision fuel elements and many of the problems associated with them. And to top it all off, it ran on thorium. Part of the novelty comes from using molten salt and the rest comes from using homogeneous liquid fuel. With the government’s love of initials it was called an LFTR – liquid fluoride thorium reactor – and here’s how it worked – and how it could work today.

One of the interesting things is that the reactor uses liquid fuel – instead of putting an array of metal fuel elements into a tub of water a liquid fuel reactor has fuel particles suspended in the liquid coolant. When the coolant is circulating through relatively narrow pipes the geometry prevents a fission chain reaction from taking place; the fission chain reaction occurs in the more bulbous reactor vessel, where enough of the liquid can collect to sustain criticality (remember that, in a nuclear reactor, “criticality” simply means that the rate of fissions – the reactor power – is remaining constant over time; ALL reactors are critical when they are operating). Anyhow – returning to the main point, a liquid fuel reactor dispenses with fuel elements, control rods and their associated drive mechanisms, and so forth because reactor criticality is controlled solely by the shape of the pipes and other components through which the fuel passes.

Another advantage of liquid fuels is that they can be continually processed to remove fission products as well as to adjust the amount of uranium or thorium it contains. So rather than fission products building up in fuel rods they can be scrubbed from the fuel as it circulates through the reactor system – doing this can help to reduce radiation levels in the reactor plant and buildings as well. So – in theory – a liquid fuel plant should be radiologically “cleaner” than an ordinary reactor, reducing radiation dose to the workers. Not only that, but any reactor accident would likely release far less radioactivity than did the Chernobyl and Fukushima accidents.

The other novel part of the LFTR is the liquid fluoride – the thorium is in the form of a thorium fluoride salt (“salt” is a general chemical term that refers to a specific category of chemical compounds of which table salt is only one); the salt is heated to its melting temperature of several hundred degrees by the fission reaction and the hot salt circulates through a heat exchanger to boil water, forming the steam to drive turbines. What’s nice about molten salt is that it doesn’t have to be at high pressure to be at a high temperature, so the system is under less stress and leaks are less quickly catastrophic. All of these features – lower pressure, less in-core machinery, simpler emergency cooling systems, and so forth – mean that LFTRs are also likely to be much less expensive to build than our current light-water designs.

Another nice feature comes when you combine both of these factors – the inherent safety of the LFTR design. Consider – energy is only produced when the liquid fuel is in a critical configuration in the reactor vessel, and that heat is given up when the liquid salt is pumped through the heat exchanger to create steam. So what happens if the pumps fail? In a typical water-cooled reactor the coolant stops flowing through the core, the fuel heats up to the melting point, and havoc ensues. In a liquid-fuel reactor the fission reaction continues, the liquid heats up, and it melts a drain plug in the bottom of the reactor vessel – the liquid fuel then drains into a shallow tank where it loses its critical geometry and the reactor shuts down. This should – in theory – prevent the sort of meltdown that destroyed the reactor cores in Fukushima and Three Mile Island and that caused the massive releases of radioactivity last year in Japan. I know that a number of more recent reactors are designed to be “inherently safe” but the LFTR comes closer to this ideal than does any water-cooled reactor if only because the fuel is already molten and, when it loses its critical configuration, the laws of physics dictate that it will shut down.

So with all of these advantages it’s natural to wonder why our reactors are still the “conventional” designs loaded with solid fuel rods and cooled with high-pressure and high-temperature water. Martin (the author of Superfuel) contends that the liquid fuel designs were torpedoed in part by the established water-cooled reactor designs that were being pushed by Hyman Rickover, the architect of the US Navy’s nuclear power program. That, plus the nuclear establishment’s level of comfort with a proven technology (the light-water reactors) might have sealed the fate of the liquid-fuel thorium reactor design, in spite of its advantages. I suspect that the reality is a combination of these factors. Regardless of the reasons why these reactors have not been built in the nuclear nations, it seems likely that thorium will become part of the Indian and Chinese nuclear programs – more on that in a future posting!

Tags: , ,

9 Responses to “Liquid fuel molten salt thorium reactors”

  1. Seth Nawk August 18, 2012 at 3:50 AM #

    You might think sometimes that the so called ‘waste’ is a tradable product
    and could become well needed in a warfare at the time it was discovered/invented..

  2. Mark L. August 18, 2012 at 11:05 AM #

    I have a hard time buying the argument that LWRs put paid to the LFTR. The LFTR – as far as I can tell – was too complicated and advanced to be developed in time for the first generation of commercial power reactors, so LWRs (or some other Gen-II design) were always going to dominate the initial field. I would think the real competition for the LFTR would be the Liquid-Metal Fast Breeder Reactor and other breeder reactor designs, which beat the LFTR in the race for R&D dollars.

    Another aspect I wonder about is whether it might make sense to begin with uranium-fueled Molten Salt Reactors rather than LFTRs. As far as I know we’ve never built an online reprocessing system of the type a true LFTR would use – I don’t think the two US prototypes incorporated them. That seems like an extremely complicated (and expensive!) piece of hardware. I wonder if it might be better to build non-breeding MSRs before moving on to true LFTR designs, especially since we don’t seem to be in imminent danger of running out of cheap uranium.

    Oh, and you left out my favorite part of the story, the LFTR’s beginning as the Circulating Fuel Reactor, part of the ANP Program to build nuclear-powered airplanes. Is that awesome or what?

    • Shmoe August 19, 2012 at 3:19 PM #

      @Mark L.:

      I think you have a point, about timing and Gen II reactors. I think there’s a tendency for advocates to revert to conspiracy theory when talking about their preferred tech.

      However, I would also say, since Gen II’s promulgation, regulators have become hidebound, and the industry is basically a rent-seeker that makes money by making expensive fuel (uranium-based, or not). This doesn’t benefit anyone, except, maybe, the industry itself.

    • Alberto R. August 20, 2012 at 7:19 PM #

      You’re right that an uranium MSR without reprocessing (non breeding) is a good idea to start. David LeBlanc (a canadian nuclear physicist) calls it DMSR and is working on it.

  3. K August 20, 2012 at 10:19 AM #


    I encourage you to watch some of the many thorium videos available online (thorium remix 2011) and read the book “Superfuel”. The material provides an in depth look into how the current nuclear industry ended up with light water reactors versus any other.

    The short version is that uranium solid fuel reactors were created to produce the raw material necessary for the atomic bombs. After that, the scientists who developed the LWR started looking for a better design for purely civilian use. What the group at Oakridge created was the MSR (Molten Salt Reactor) experiment and ran it for many many hours to prove it works. Unfortunately, waiting for the MSR experiment to complete before pushing the concept for funding proved to be its downfall. By that time the gov’t had already made up its mind to go with LWR’s and pursue another reactor technology based on the fast spectrum. Fast spectrum reactors designs had not yet been proven but had a large funding base for research thanks to its many supporters. In the end a fast breeder reactor was never built and consumed a copious amount of $’s (see chalk river). This left us with only LWR’s as even though the MSR reactor experiment was a HUGE success, it materialized no funding.

    As for online reprocessing, this is not some complex operation. The reactor design is completely different than today’s reactors that I could see why you might think it is hard. There are also 2 different approaches to consider. First, you can choose to do no online reprocessing and leave almost all the fission products in the reactor and they will eventually burn up or you could shut down and “refuel” just like we do today with LWR’s. Or, you build the plant to enable online reprocessing. Which is far simpler since you already have a liquid fuel which is circulating out of the reactor. This is trivial chemistry. There are far more complex chemical plants out there than what is needed to remove the fission products continuously from a LFTR design. (Also, the amount of fission products is extremely small!)

    • Mark L. August 20, 2012 at 11:09 AM #

      I have read Superfuel, and I’ve watched some of Kirk Sorenson’s videos.

      Don’t get me wrong, I’m a huge fan of the LFTR. As a complete non-expert, it seems like our best bet for a safe, cost-effective breeder reactor, which in turn seems like our best long-term bet for an energy source. I’m just A) skeptical that LFTRs could have been ready for commercialization by the mid to late 60s. That’s when we saw the first wave of commercial construction of nuclear reactors. LFTR seems like a more long-term project, whose real competition was other long-term projects like the LMFBR. And B) I can’t help but wonder if the difficulties of turning this into a commercializable technology might be being downplayed.

      Anyway, I hope I’m wrong about item B, and I hope Kirk Sorenson becomes a billionaire selling safe and cheap LFTRs to the world. Because that would be awesome.

      • K August 20, 2012 at 2:13 PM #

        I would love to see Kirk succeed here in North America! However, China and India are likely to be the sole IP owners of the technology at the current rate we are headed in…

  4. Ted Bernstein August 20, 2012 at 2:24 PM #

    Cheap, clean, by far the safest way of generating electricity currently know to humankind. When I first learned about LFTR, this was what I concluded. Like so many others, I found myself asking one question. “Why in the world didn’t we develop it after it was invented back in the 1950′s and proven to work in the 1960′s?”

    The anwers, are of course complex, but also very simple and mundane. We didn’t develop LFTR’s primarly because of the personalities and the egos of a few very important people. The full story may never be known, but the likeyhoood that this policy failure by the leadership of the AEC and the US Congress at that time, in deciding not to fund the furthering of our knowledge in liquid salt fuels, has changed our world for the worse can not be logically refuted.

    Support LFTR. What we do now matters. The world can be better than it is, and LFTR can help us get there. If you doubt it, you need only to imagine what the world world would look like today if every major city in the world were powered by affordable, clean, safe power. We can’t go back and change the past, but we CAN change our future.

    I too am rooting for Kirk Sorensen and Flibe Energy, but we should do more than sit back and cheer. We need to make certain our government does not miss a chance to explore this technology again. I’m sure that many in office today don’t have a clue about LFTR or what it represents. Let’s have a converstation about that and what can be done about it.

  5. Chris Brudy September 21, 2012 at 6:30 AM #

    A LTSR reactor was run successfully at Oak Ridge for about nine years, and was shut down, supposedly, because it would not produce weapons grade materials. Actually, it was shut down because it was a viable alternative to LWRs. The industry had gone with LWRs because there was much more money to be made building and servicing the LWRs, and LTSRs threatened the monopoly the industry had developed.
    We know this because there is a large stock of U233, which is not weapons grade but can be used to “seed” the LFTRs, and the industry furiously lobbied to have Congress spend tens of millions destroying the U233 stockpile. So, the industry is afraid of LFTRs, which means the technology must be viable.
    I suggest a spherical containment vessel, so that if any of the scalable LFTRs ar ripped from their foundations by storms, earthquakes or tsunamis, would automatically scram (shutdown) wherever they came to rest.
    The true beauty of the system is that waste from LWRs, now sitting in vulnerable pools of water at nuclear plant sites, has a half life of some 50,000 years. This waste can be processed and used to seed the LFTRs, which gives us electricity and reduces the half life of the fuel to 400 years, once it has been used up. We could take the spent salts, bottle them securely, and dump them to the bottom of the oceans. Four hundred years? That’s nothing with a mile and a half of ocean above….

Leave a Reply