The Integral Molten Salt Reactor And The Benefits Of Having A Liquid Fission Reactor

Although to most the term ‘fission reactor’ brings to mind something close to the commonly operated light-water reactors (LWRs) which operate using plain water (H2O) as coolant and with sluggish, thermal neutrons, there are a dizzying number of other designs possible. Some of these have been in use for decades, like Canada’s heavy water (D2O) reactors (CANDU), while others are only now beginning to take their first step towards commercialization.

These include helium-cooled, high-temperature reactors like China’s HTR-PM, but also a relatively uncommon type developed by Terrestrial Energy, called the Integral Molten Salt Reactor (IMSR). This Canadian company recently passed phase 2 of the Canadian Nuclear Safety Commission’s (CNSC) pre-licensing vendor review. What makes the IMSR so interesting is that as the name suggests, it uses molten salts: both for coolant and the low-enriched uranium fuel, while also breeding fuel from fertile isotopes that would leave an LWR as part of its spent fuel.

So why would you want your fuel to be fluid rather than a solid pellet like in most reactors today?

Historical Lessons

Aircraft Reactor Experiment. General diagram. (Source: ORNL)
Aircraft Reactor Experiment. General diagram. (Source: ORNL)

Even though many newly licensed or about-to-be-licensed reactor designs in the 2020s sound futuristic, virtually all of them have been conceptualized in some form or shape before the 1960s, and many have had prototypes built. So too for molten salt reactors (MSRs), which saw Oak Ridge National Laboratory (ORNL) create a number of prototypes, starting in 1954 when the Aircraft Reactor Experiment (ARE) reached first criticality. ARE was an off-shoot of the Aircraft Nuclear Propulsion (ANP) program that had its roots in the US Air Force, before it was transferred to the Atomic Energy Commission (AEC).

MSRE plant diagram: (1) Reactor vessel, (2) Heat exchanger, (3) Fuel pump, (4) Freeze flange, (5) Thermal shield, (6) Coolant pump, (7) Radiator, (8) Coolant drain tank, (9) Fans, (10) Fuel drain tanks, (11) Flush tank, (12) Containment vessel, (13) Freeze valve. (Source: ORNL)
MSRE plant diagram: (1) Reactor vessel, (2) Heat exchanger, (3) Fuel pump, (4) Freeze flange, (5) Thermal shield, (6) Coolant pump, (7) Radiator, (8) Coolant drain tank, (9) Fans, (10) Fuel drain tanks, (11) Flush tank, (12) Containment vessel, (13) Freeze valve. (Source: ORNL)

From there the project ended up at ORNL, where the original solid fuel design was changed into a molten salt/fuel mixture due to concerns over reaction stability at high temperatures, which the MSR design might solve. After the cancellation of the ANP program the MSR technologies from ARE and subsequent designs were used for a purely civilian project: the Molten-Salt Reactor Experiment (MSRE).

Like the ARE, MSRE used molten fuel, albeit with a different composition. ARE used 53.09 mole % NaF, 40.73 mole % ZrF4, and 6.18 mole % UF4 for its salt/fuel mixture, with uranium-235 as the fissile material. The neutron moderator also changed from beryllium oxide (BeO) in ARE to pyrolitic graphite in MSRE.

MSRE used 7LiF – BeF2 – ZrF4 – UF4 (65 – 29.1 – 5-0.9 mole %) following lessons learned from the ARE salt mixture. Initially 33% (enriched) uranium-235 was used in its primary coolant/fuel mixture, before switching to using uranium-233 bred from thorium in breeder reactors. Although it would have been possible to configure MSRE to use thorium salts to breed its own fuel from, this was omitted for the experiments, with instead neutron measurements being performed. This does however touch on one of the benefits of an MSR, in that they can be a fast neutron reactor unlike a water-moderated LWR, making them capable of breeding their own fuel from fertile isotopes, including the transuranics and actinides resulting from the original uranium fuel. The other advantage of MSRs is that they can operate at very high temperatures (820 °C for ARE, 650 °C for MSRE) due to the high thermal stability and heat capacity of the coolant, while not requiring the pressures seen with pressurized light water reactors (PWRs), which typically feature a outlet temperature of around 300 °C.

The operating temperature ultimately determines what processes and turbines it is compatible with, as industrial processes often require temperatures far above what LWRs can deliver. An MSR capable of providing a constant source of >600 °C heat would be exceedingly practical for these applications, as well as increase the thermal efficiency of electricity generation via steam turbines.

Over the course of the MSRE’s five-year lifespan, it provided significant information on the behavior of both 235U and 233U fuel, as well as the production and handling of xenon gas (a neutron poison), the stability of the graphite moderator and the immunity of the used salt to any kind of radiation it was exposed to. It also validated the new metal alloy developed at ORNL to resist the corrosive effects of the hot salt, which is a nickel-chromium-molybdenum called Hastelloy N.

When the MSRE was shutdown in 1969 for decommissioning, one unexpected finding that was that of embrittlement of the metal exposed to the salt, which was traced back to tellurium, one of the fission products. This finding led to an adjustment of the Hastelloy N alloy, to counteract these effects. Yet despite these successes, the US would virtually abandon further development in MSRs, despite ORNL creating a number of follow-up designs.

The use of salts as coolant would however continue worldwide, mostly for fast neutron reactors, with examples like the Russian BN-series of sodium-cooled, fast neutron reactors being prime example of this. Although they use solid fuel, they demonstrate the viability of long-term use of corrosive, hot salt as coolant, with the BN-600 operating since 1980. The similar US EBR-II operated from 1964 to 1994, including simulated emergencies like a sudden shutdown at full power, demonstrating the passive safety of these pool-type sodium-cooled reactors, much of which also applies to MSRs.

The Integral MSR

Schematic describing a few possible applications for the IMSR. (Source: Terrestrial Energy)
Schematic describing a few possible applications for the IMSR. (Source: Terrestrial Energy)

The Terrestrial Energy’s IMSR is a fully self-contained MSR, with the molten salt, the pumps, primary loop and graphite moderators are contained within what Terrestrial Energy calls the IMSR Core-unit. A single Core-unit produces 440 MW thermal (MWth) with a claimed 44% thermal efficiency when generating electrical power, due to the 700 °C outlet temperature. A plant with a typical dual Core-unit  configuration thus generates around 390 MWe, with the remaining heat conceivably being used for cogenerating purposes (e.g. heating), although Terrestrial Energy currently envisions dedicating a single Core-unit to thermal energy, making for 195 MWe and 440 MWth from a dual IMSR plant that is also employed for process heat, thermal storage, etc.

What is interesting about the IMSR is that it is a purely thermal spectrum Generation IV reactor, despite using molten salt fuel. This fuel itself is standard low-enriched uranium (LEU) at <5% 235U, the same as used in virtually every commercial reactor in use today. There are no fast neutrons being used to breed fuel from fertile fission products or thorium salts – with the graphite moderator moderating all fast neutrons to thermal neutrons – making it very much akin to an LWR.  The advantages of using molten salt here come mostly from the much higher heat capacity at ambient pressure, as well as online refueling, with each Core-unit expected to remain in 24/7 operation for seven years. During this time fresh fuel is gradually added to the primary loop to maintain reactivity.

After the operational time is over, the entire (sealed) unit is left to cool off for a while before it is returned to the factory for recycling. This means far less stringent requirements for the operator, as the unit is essentially maintenance-free, which is part of Terrestrial Energy’s pitch towards commercialization. Such a focus on simplicity of operation is popular with small modular reactor manufacturers – including GE-Hitachi’s BWRX-300, which is a more conventional boiling water reactor (BWR) type LWR – but also with a range of other upcoming MSRs.

Compact, Safe, And Hot

Perhaps unsurprisingly, the IMSR isn’t the only MSR in town today, with a few more contenders looking to commercialize their own designs over the coming years. These include the Danish Seaborg Compact Molten Salt Reactor (CMSR), the Moltex Stable Salt Reactor (SSR) and a number of designs that also use salt as coolant, but with solid fuel such as the TerraPower Natrium and Kairos Power KP-FHR, What’s also interesting is how unique each design is.

The IMSR, for example, is designed to fit in with a more LWR-like once-through fuel cycle (following ORNL’s denatured MSR design), just much more convenient, more compact and with much higher temperatures. The CMSR meanwhile does not use a graphite moderator, but sodium hydroxide in parallel tubing, which makes the design even more compact and adjusting the neutron moderation easier. The SSR does not target the use of thermal neutrons alone, but also fast neutrons in its SSR-W configuration, for ‘wasteburner’. This uses liquid fuel rods, suspended in a salt coolant pool, that would use spent fuel reprocessing remnants (like actinides) for its fuel, along with plutonium, MOX fuel and so on, befitting the name.

Although MSRs have the potential to be refueled continuously without shutting them down, this is not a common design feature yet. Even so MSRs and similar high-temperature reactors are likely to become a common sight over the next years, with larger reactor types on the scale of traditional LWRs (1+ GWe) an interesting prospect, especially if they are operated in a fast neutron configuration, able to use both spent LWR fuel and fertile isotopes like thorium.

Regardless of the exact configuration, MSRs have the intrinsic safety feature of a negative fuel temperature coefficient, meaning that the reactivity of the fission process is inversely linked to the temperature of the salt coolant. This makes an MSR not only intrinsically resistant to high reactivity events, but also of load-following, as the extraction of heat from the secondary cooling loop immediately impacts the core’s reactivity.

It’s hard to believe that it’s been seventy-odd years since ARE’s first test run, but as this decade dozens of MSRs go into commercial service, it should feel as validation for the researchers at ORNL and colleagues for all the work they put into making molten salt work as reactor coolant as well as fuel.

(Heading image: IMSR plant as imagined by Terrestrial Energy. (Credit: Terrestrial Energy) )

44 thoughts on “The Integral Molten Salt Reactor And The Benefits Of Having A Liquid Fission Reactor

  1. Well done!
    I’ve been following the Gen IV nuclear technology for some time and this article is a good review. However, for anything to happen in California (where I live) the state needs to change the state laws to allow for power companies to even apply for a license through the NRC.
    Many in California believe we can solve our energy needs with solar and wind alone but without an energy dense solution like MSR, it’s hard to see the light at the end of the tunnel. Doug Fletcher

  2. There’s a lot to unpack here but until we get stable fusion this sounds like a good stop gap solution, especially in remote areas where transmission is an issue. If refueling can be streamlined, or at least made less of a logistical hurdle than traditional solid fuel reactors then I look forward to seeing how this tech is developed.

    1. Frankly, for all the difficulties with fission, it’s still so much more energy dense than chemical fuels that I really don’t mind a nonzero amount of nuclear waste. The big problem is that everything has to be kept secure against anyone who would misuse it, and that includes the waste. I don’t want small-scale fission, and I think the minimal-waste types are important. Apart from that, we can get enough material not to run out of fuel for ages, and it doesn’t wreck our environment in normal operation the way fossil fuels do. Fusion can potentially be better, but if we did fission properly that would meet our needs just fine – and it seems much easier to expand it quickly into making up a large part of the grid, along with using it to make up for the burning of carbon in places not served by sufficient other sources.

      1. Nuclear waste would be generated even if we had zero fission reactors running. It’s one of the reasons why crucial minerals production were outsourced from western countries to Africa and China.

        See for example, Monzanite.

        1. That’s also the reason why we can’t ignore fission power anyways, if we are to return supply and production back to the countries that at least care about doing something about nuclear waste instead of tipping it down somewhere nobody’s looking.

          The best way to get rid of it – use it.

        2. Fair. IIRC coal also can contain enough that you might see more radiation dose near a coal plant than a nuclear one. And I know from formerly living near a refinery that petrochemicals are no good for you even without any radiation. But even apart from that, if I could push a button and revert the atmosphere to a preindustrial state at the cost of the equivalent amount of nuclear waste to be kept contained and guarded… it all depends on how competently we would do that, or whether some would ‘grow legs and walk away’.

          1. Well, clearly the nuclear waste isn’t a problem.

            Even the people who oppose nuclear power find that the present solution of doing absolutely nothing about it is good enough. Otherwise they wouldn’t oppose every attempt at processing or burying it somewhere.

    2. Don’t hold your breath for a stable or really practical fusion reactor – with Tritium at $30,000/gram and the huge plant costs currently envisioned, it’ll never be panacea that scientists and politicians make it out to be.

      Remember: their is no such thing as renewable energy – solar panels consume quartz and metals as well as coal for the energy of their manufacturing. At least molten salt nuclear plants can be built cheaper than current plants and also have a very long life span, potentially 4x solar panels or wind turbines. Thorium is basically dirt, and there’s literally mountains of the stuff piling up for free as it’s a left over from rare earth metals production – thorium reactors would run on garbage!

    1. What a vindication for advanced reactors that consume a fair more minor actinides than people realize. In your face Clinton and Von Hippel. Go ahead and bury actinides. Me, I’d rather destroy them in reactors never to be on the earth ever more, and claim their energy content. Go ahead and bury it. People will find a way to dig it back up. For my money, it’s worth more than gold. Who buries gold? Idiots. And before you chime in its proliferant. It isn’t. Even-even nuclides aren’t fissionable (some were used as batteries for pacemakers)and you can’t separate them from odd-even very easily. Read ANSI/ANS 8.15, criticality of minor actinides. Then get back to me when the light goes on in your head.

      1. Right-on!

        I never understood the non-proliferent argument in the case of a country like the US – what, are there no guards and security at the reprocessing plant?

        Seems like Carter just wanted to lead the way, but most other nuclear countries saw it for the stupid idea it is and continued reprocessing.

        Apparently it’s still cheaper to bury the stuff and get fresh uranium – but since the US doesn’t actually bury the stuff how is that true? I believe Finland is the only country to actually have an official, operational storage facility and it’s brand new, so no cost history yet.

        The amount of used fuel is so small, that even for a natural uranium reactor like the CANDU, that has more volume of fuel because it’s not enriched, can store all the fuel rods from decades of operation onsite in the cooling pools of the nuclear plant.

    1. Ask the people surrounding the Pickering Nuclear Reactor if they mind a nuclear rector in their back yards, they’re mostly cool with it.

      43.814538254269166, -79.05925603148403

      Check out the satellite view.

  3. The biggest problem with liquid fuel reactors is proliferation monitoring. Once you break the pin you have no way of tracking it.

    Another problem with this type of reactor is the use of fluorine based salts. Nasty with a capital N.

    My preference is the use of liquid metal fast breeder reactors (LMFBR) where transuranics are burned up. They have been demonstrated as walk away safe. They also don’t need a containment dome.

  4. There is no reason why the MSR’s can’t be immediately licensed ! There is no need for the NRC to spend 10 more years “studying” & writing more regulations & forcing companies to spend more money for a license that it costs to manufacture a MSR.
    ThorCon’s founder, Dr. Robert Hardgraves is from New Hampshire (USA). It would be fitting if someone established another Electric Co-op in NH so we could pressure Congress/President to tell the NRC to give them a license or get FIRED ! Any size unit needs to provide heat & power to an industrial park or large shopping mall with heat & power being sold “over the fence” to abutters. It’s estimated that without all the interference by the politicians & having Local Control and no need for Interstate Transmission Lines, electricity could be sold for about 6 cents/kwh, instead of the present 23 cents !

    1. You don’t read do you. The iMSR is in a can and no workers get near the hot parts. The can stays in the ground a few years to cool down and can then be sent back for recycling. Yes they wear out as planned in 7 years. And they are unpressurized and no need for containment of steam explosion. So 1% of the price of a LWR. And 1:10 th the lifespan. So 10x cheaper in CAPX. But everything is built on an automated assembly line so it’s just like rebuilding a heavy diesel.

  5. I remember i saw a documentary on TV (on Arte, a french/german TV channel) about molten salt reactor something like 15/20 years ago. Beside the advantages of this technology, the documentary explained that the currently dominating technology was favoured only because it was the only one to be able to produce some isotopes needed by military, not because of its advantages or cost efficiency.
    Maybe the documentary did not mention the drawbacks, i don’t remember.
    But since then, i always wondered why almost nobody talks about molten salt reactors (i think i heard once that China was considering building some).

    1. Worth considering the land area required for wind and solar. The energy density is so low that we will have to dedicate huge areas of land in order to generate enough power, which surely has detrimental effects on local habitats. Saving nature by destroying nature… Doesn’t make any sense. Makes more sense to have compact energy generation that takes up a fraction of the land area, ie modern/coal/oil/gas/nuclear

    2. Nuclear is already cheaper than wind and solar.

      Take away the government subsidies, properly compare the long-term costs of entirely replacing your energy grid every 15 years (And the batteries more often than that), the costs of recycling all those piles of trash (It’s far more expensive to recycle cobalt than to mine it), the enviromental and social costs of the absurd amounts of land that need to remain deforested for solar and wind to occupy, how much more continuous mining of rare resources they require, the rising costs of said resources as the cheaper reserves get depleted, etc…

      Even the one-off projects staffed by inexperienced engineers and absolutely covered in red tape that are used by anti-nukies as benchmarks of how much nuclear power costs will compute as being cheaper.

      Now do a proper comparison to nuclear reactors that can be built at scale and it’s just silly. Not to mention the fact that nuclear reactors are mostly just concrete and steel.

      Outside of the western bubble there are already nuclear installations that were cheaper to build (To build! That is, just comparing short-term costs) than local solar/wind power plants. And as they achieve economies of scale, they’ll only get cheaper.

      Meanwhile anti-nuclear Germany is burning more coal to keep the lights on.

      1. “Cobalt is cheaper to mine than recycle”
        The majority of cobalt has historically been consumed in Petrochemical Refineries, where the cobalt is contaminated by the processing.

        Recycling cobalt from lithium ion batteries is a distinctly different issue with only highly refined materials within the cells.

  6. China is building a MSR for a Coal to nuclear upgrade. If they like the taste of that, they will upgrade all their coal plants quickly and become net zero. India will do likewise. Note that building a new nuclear facility in India doesn’t reduce carbon output. But upgrading a coal to nuclear plant does! Because they’re not going to turn off any power plant for a long time.
    The US department of Energy says we have 300 coal fired power plants that could be converted to nuclear. It’s cheaper than free! Do you know how much a house is worth that’s next to the railroad tracks that are constantly carrying cool? Not much. But shut down those coal trains, and that whole part of town is suddenly worth more. The increase in property values and property taxes were more than pay for the cost of the power plant upgrade.

    Only the molten salt reactor can fit inside the footprint of an existing coal plant. The light water reactor is too pathetic in temperature, and too large due to containment and security needs.

    650C steam can also be used to repower all of those beautiful frack gas powered turbine generators all over America. Uranium and thorium are infinitely renewably available and way cheaper than methane. There’s more uranium than we could possibly use, cheaply recoverable by polymer nets from the ocean. You don’t need a mine. And when you extract it it comes back from geothermal hot springs. It’s renewable!

    1. Considering china is building on average 2 new coal plants per week, upgrading wont go quickly. Also, the sheer volume of msr reactors involved, should china upgrade, would end in serious accidents from a statistical viewpoint. The gods of progress are indifferent and demanding.

      1. Nick Touran Ph.D. works for TerraPower, LLC the nuclear power outfit started by Bill Gates. Maybe there is just a bit of conflict of interest……
        Besides, most of his talk deals with thorium based MSRs. Not the fuel cycle for most of TerraPower’s competitors.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.