Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Liquid-fuel molten salt reactor?

  1. Jan 18, 2009 #1
    I've run into a some advocates (assorted nuclear engineers and grad students it seems) for an unusual nuclear reactor design. It is a thermal-spectrum thorium breeder reactor. The fuel is liquid - it is a molten salt, containing the fluorides of both U-233 fuel and thorium, as well as lithium and beryllium as moderators. (As well as graphite? I'm not sure)


    The proponents claim it is a viable alternative to fast reactors, in that the fuel cycle efficiently burns all thorium fuel, and relatively little transuranic actinides are produced. They suggest fluoride salts are convenient for reprocessing based on fluoride volatility, that the closed fuel cycle would be cheaper than that of solid-fuel fast reactors (because there are fewer chemical conversions, so less reprocessing waste would be created). They claim the hot, molten fluoride salts will not corrode away the whole reactor. And they say the idea is an offshoot of a 60-year old experiment in Oak Ridge, which tested molten U-235 fluoride fuel for a reactor designed for powering nuclear bombers (airplanes).

    The chief exponent is Kirk Sorensen, who is an NE grad student at UT-Knoxville.

    I am not a nuclear engineer, and I have no ability to evaluate any of this. But I am intrigued by this novel-sounding idea. So, experts: is this idea interesting, or even feasible?
  2. jcsd
  3. Jan 18, 2009 #2


    User Avatar
    Science Advisor


    I don't know about powering airplanes - but the reactor described above sounds like an experimental
    reactor studied / built at Oak Ridge National Laboratory called the MSRE - Molten Salt Reactor Experiment:


    [ The MSRE was housed in a building that was once used for aircraft reactor research ]

    I'm dubious of the claim that the cycle would be superior to the solid-fuel fast reactor reprocessing because
    of fewer chemical reactions. The Integral Fast Reactor [ IFR ] developed by Argonne National Laboratory
    in the '80s-90s employed a metal fuel. Metal fuel can be reprocessed by a metallurgical technique of
    electrorefining without a lot of chemical conversions. See:


    Dr. Gregory Greenman
    Last edited: Jan 18, 2009
  4. Jan 18, 2009 #3
    Yes, as I understand the MSRE was based on work done at the Aircraft Reactor Experiment - the wikipedia page mentions this.

    I see - so IFR is fuelled by metallic U/Pu, and so there is no oxide reduction step, as there would be for MOX fuel, or TRISO pebbles? (pardon my ignorance)

    Is the fluoride reactor design sound, or is this an obvious flaw, a reason why nothing ever came out of the MSRE?
  5. Jan 18, 2009 #4


    User Avatar
    Science Advisor


    The molten salt reactor is a nice design. As I understand it; one of the major drawbacks is that the
    reactor operator [ the electric utility company that would employ such a reactor ] would have to
    operate the molten salt reactor's chemical processing facility. That is; a facility to continually
    reprocess the reactor's molten fuel is an integral part of the operation of such a reactor.

    Thus the operation of a molten salt reactor is more complex than even operating a commercial
    light water reactor. This was more complexity than most electric utility companies at the time
    wanted to undertake - they were in the business of generating power, not operating a chemical
    processing plant.

    The IFR concept also includes an on-site reprocessing facility - but a facility based on electrorefining
    would be simpler to operate than a chemical processing plant.

    For both concepts; it is a matter of how much complexity the reactor owner wants to manage.

    Dr. Gregory Greenman
  6. Jan 19, 2009 #5
    Here at Ohio State University, my class in nuclear design (a team based approach) investigated the LFR. It is a research based class, and each year, the class builds upon the previous year's work. It was started (I believe) here at OSU in 2004 or 2005).

    The reactor does indeed use a FLiBe salt using thorium, with a blanket surrounding it as a breeder material. There were some promising results, but yes, the online processing turned out to be one of the major challenges, as well as initial startup. It required a substantial amount of U-233 to start, but eventually enough U-233 was bred so that it could supply startup uranium for additional reactors. I went through the class Fall 2007, so I don't know what advances (if any) were made this past quarter.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook