Seaborg Molten Salt Reactor Will Fit on A Truck and Cost Less Than Coal Power

Seaborg is the largest reactor design start-up in Europe and they are making an ultra-compact molten salt reactor (CMSR). Seaborg cannot meltdown and can use spent fuel. Conventional nuclear reactors have solid fuel rods that need constant cooling, typically using water under high pressure.

They are talking to the East Asian supply chain and are talking to customers who could buy 10-15 units for the late 2020s.

In the CMSR, fuel is mixed in a liquid salt that acts as coolant. This ensures it can always be cooled and it cannot melt down or explode. It will simply shut down by itself in case of an emergency.

The importance of this is not only the safety but also a significant reduction in complexity and cost.

They have a design for a molten salt reactor that is ten times smaller than the Terrestrial Energy IMSR. It would 20 to 30 times smaller than an existing pressure water nuclear reactor for submarines.

Seaborg CUBE reactor can use spent nuclear fuel (SNF) by adding thorium as a catalyst. The CUBE as a waste burner. Current conventional reactors use about 4% of the uranium fuel rods. This is because they use Uranium 235 and cannot use the Uranium 238. The thorium aspect is only useful in terms of very long term issues. Thorium with Uranium can extend nuclear fission to hundreds of thousands of years instead of tens of thousands of years.

Timeline aligned with standard IAEA reactor development method
• 2014-2016: Pre-conceptual Design Phase 1
• 2017-2018: Pre-conceptual Design Phase 2; 1.5 Million Euros
• 2019-2020: Conceptual Design Phase; 10 Million Euros
• 2021-2024 Technical Design Phase; 50 Million Euros
• Ready to build reactor blueprints

Delivered cost for 250 MW thermal MSR in 2025 in the $50 Million to $70 Million depending upon manufacturing scale. They are working towards a 50 MW thermal pilot plant and then would scale to 250 MW thermal for a commercial system.

SOURCES- Seaborg, Youtube – Gordon McDowell
Written By Brian Wang,

30 thoughts on “Seaborg Molten Salt Reactor Will Fit on A Truck and Cost Less Than Coal Power”

  1. How different do you think it would be if each task had performance bonuses for early delivery? I’ve noticed people tend to work exactly as they are incentivized to.

    Early in my career at a near management free webdev house a novice was left alone to work on a project and told to ask for help if he got stuck on anything. One day he was excited that he had solved a unicode translation problem after only 6 weeks of effort working remotely from home. The rest of us looked at each other and checked google, finding many answers to his exact problem in about 60 seconds. He got fired immediately.

    When I give fixed price quotes to build gadgets I perform literally about 5x faster than I do when on hourly contract. My hourly performance is great but I can muster extraordinary energy levels when I have the opportunity to maximize my income by burning brighter for shorter periods.

  2. Mixtures of salts have lower melting points than pure salts. BeF2/NaF solidifies at 385C. Running at ~550C would give supercritical steam, and still leave some headroom before the steel got stressed.

  3. And the graphite doesn’t do a particularly good job at slowing the neutrons… Note that the graphite core of the msre was perhaps 6-8 foot tall right circular cylinder and the moderation was so poor that the salt needed to be +30% enriched. Compare that to the possibility of criticality with low enriched uranyl nitrate in a 5 gal bucket of water.

  4. Have to remind people that UF4 melts (melts!) at over 1000C, and boils at 1417C. Honestly, there aren’t reasons to dilute the UF4 in MSR other than to lower this melting point (ZrF4 = 910C, BeF2 = 554C LiF = 845C).

    Wiki Superalloy:

    “State-of-the-art turbine blade surface temperatures are near 2,100 °F (1,150 °C); the most severe combinations of stress and temperature corresponds to an average bulk metal temperature approaching 1,830 °F (1,000 °C). Although superalloys retain significant strength to temperatures near 1,800 °F (980 °C), they tend to be susceptible to environmental attack because of the presence of reactive alloying elements (which provide their high-temperature strength). Surface attack includes oxidation, hot corrosion, and thermal fatigue. In the most demanding applications, such as turbine blade and vanes, superalloys are often coated to improve environmental resistance.”

    Those are nickel alloys from which you could fabricate the dump tanks. So, you want to dump brazing flux (MSR fuel) into a nickel alloy tank at a temperature where it can be said that “superalloys retain significant strength”? You have about a 700C window of temperature before boiling at 1420C. The lower end of the window is at the bleeding edge of what superalloys can handle in surface temperature, not to say bulk temperature.

    MSR is nonsense just based on material science.

  5. There is no “market” for electrical generation. Power generation is municipal infrastructure – it is different.

  6. Fair enough, there are other ways to accomplish the same end. But I do think this is a reasonable way to accomplish it.

  7. Most of the molten salt designs floating around now sacrifice a few percent of thermal output to passive cooling of the fuel vessel, so in an accident, it’s there already. There are ways to increase the heat flow if the reactor shuts down, or overheats. Radiated heat increases as the fourth power of the difference between source and surroundings, so an overheated reactor will lose more heat anyway. Then you can do things like put a ‘fluid diode’ in the main circuit – if pumped circulation stops, convection will set up in the other direction. Terrestrial Energy want to surround the fuel vat with a tank of salt with a slightly higher melting point. In normal operation, it acts as an insulator, but if the reactor overheats, it melts, and convects heat away to the passively cooled outer surface.

  8. The moderator is there to accelerate the reaction, not create it. Get rid of it and you may no longer achieve criticality, but you still have nuclear chain reactions producing a slow radioactive burn that you don’t want to get anywhere close to for a very long time. The more you spread out the fuel, the less total radiation you get from the same amount of fuel.

    One way to address that is a large-diameter flat tank so that the evacuated fuel is spread thin and most decay neutrons rapidly escape vertically without triggering chain reactions. Alternately you can collect it into several more spherical containers, where each one is small enough to keep the spontaneous chain reactions below acceptable thresholds.

    Personally the big flat tank seems like the simpler solution, but if they’re designing this to fit in the footprint of the reactor itself (e.g. to fit the entire thing in a cylindrical underground “silo”, that’s probably not an option). Multiple smaller tanks are probably also easier to deal with safe for removal and recovery of the evacuated fuel.

  9. Assertion, not assumption.

    Doesn’t it mean something that none of following entities have anything to do with the MSR(s) you claim are “in the design phase”:

    Westinghouse, RosAtom, LockMart, Framatome, GE, KEPCO, Mitsubishi, Kawasaki, Hitachi, Dawoo, CNNC, CGNPG, RollsRoyce, Siemens, ABB, Electric Boat, etc., etc.?

    The fact that the Russians aren’t interested in MSR technology says sooooo much. The Russians do risky engineering historically – they have vast tracts of land that were contaminated by deliberate and accidental spills and weapons manufacturing/testing… One could argue they are the current leaders in nuclear technology, and they want nothing to do with MSR. LOL.

    I believe that it is inaccurate to say that some inordinate sum of money was spent designing and building the first PWRs. PWR are not simple machines, but they are not as complex as the ships, which they power. Considering that the S1W was functioning in 1953, I’d say the development process of the PWR was straightforward and reasonable. Some of the LWR fleet operating today were built for hundreds of millions of dollars during the 1970s, and not $13B in today’s money.

  10. That is a simplistic assumption. There are several MSR models currently in the design phase. The reason a LWR was designed so fast was uncle sugar spent BILLIONs on designing and building the first models. They need to cancel one of those 13 billion dollar white elephants that make great targets and spent the money on MSR and new fusion designs.

  11. You’ve painted the perfect picture that shows why you would make a terrible engineer. Stay with your strengths, don’t think, just type.

  12. Yeah, but the point of this sort of design is that the only thing keeping the fuel out of the dump tank is a fusible plug. That’s the basis of considering it walk away safe: If it overheats, the plug melts, and the fuel drains into tanks with enough surface area to be passively cooled. Without any active intervention at all. No human decisions, no control system, no power, nothing but basic physics.

    I suppose you could have a similar mechanism for dumping fission poisons into the reactor, but the main reactor chamber can’t be passively cooled enough for these emergencies because that wouldn’t leave enough heat to actually generate power.

  13. Yeah I tend to agree. Pipes = bad. Things that stretch when heated. Big potential hole in the bottom of the reactor.

    You can use boron rods to control the reaction. In the absolute worst case you can dump salt with boron in to it which would poison the fuel salt and probably brick the reactor.

  14. No moderator would mean no reaction. Putting the fuel together in a tub isn’t enough, you need graphite to slow the neutrons.

    I am not a nuclear engineer but I suspect that the issue isn’t surface area of the drain tank so much as how well you can passively remove heat from the drain tank. 250MWth with 5% decay heat is 13MW initial decay heat which seems quite manageable.

  15. Another program I checked used recursion to verify connectivity during BWR offload/shuffle (i.e. that all fuel assemblies had some path of connection through adjacent assemblies to a functioning detector). Boss told me it took him 2 weeks to write that and get it right. My routine simply read through the 2D array and flipped 0s to 1s if face adjacent to a connected fuel assembly. By the 5th sweep through the array, it stopped flipping digits and the lack of 0s was evidence of satisfactory connectivity. That’s why you need big companies working on these things you can’t have some Viking working in isolation with a 10-person team. You need a Westinghouse or a General Electric or LockMart. You need 6 eyes on the most trivial detail.

  16. And if it were true, it would be done, right? The technology has been on the shelf since 1969, so what is the logical conclusion? The logical conclusion is that the Viking’s CATIA models are just CATIA models and that he is just another grant money grifter.

  17. If it really is smaller and safer than military reactors, then that is the business model.

    The Navy is willing to pay a lot more for energy than the power system is. Sell it as a way to make more of the fleet nuclear.

  18. Actually Brett, you don’t need to dump it at all; dumping the fuel introduces more places to leak or creep at 1000C. Chrissake; we’re only talking about 1% of the nominal thermal power after a period measured in hours post shutdown.

  19. You need to divide it up enough to passively cool it while minimizing the continuing reaction. Dumping it in a single tank leads to too low of surface area to volume.

  20. When you tell an engineer working in a silo or vacuum, to bring you a rock…. they will bring you a rock – their rock – whatever arbitrary solution occurs to the engineer in his/her state of head-down sensory deprivation. This “solution” is an early iteration; it doesn’t address the six other parameters of which they are ignorant. I’ve seen it before (vacuum distillation). I once checked an elegantly written piece of software that employed newtons method to search out a minimum of an empirical function with pseudo-random inputs (manufacturing variance); it found the minimum in about 6 iterations. The developer was allowed 4 months to develop this code. I checked the results by running a stack of evenly divided cases over the solution space and plotted his 50,000 results against my 50,000 results. I got a line y=x, with a few outliers where his method didn’t converge or erred. My method took about a week to write and execute (not 4 months); the code (FrapCon) was I/O limited. Once the input was loaded; code could run a case in about 6 femto seconds; it took less time to run a stack of 200 cases over the space than it took 6 iterations using his elegant but I/O intensive method. Sometimes things that look odd actually are one dude’s oddball/ignorant/arbitrary choices. Examples: beryllium oxide reflector in GA’s EM2 reactor concept became zirconium oxide. Moltex rectangular geometry is now cylindrical per recently released pubs. I called BS on both these things.

  21. “Beware any MSR article that doesn’t spend 90% of its words on off-gas and fuel handling.”

    Yep. I go back to wondering about negative pressure reactors where the Xe and Kr are pulled out of fuel salt due to low pressure and their nature as Noble gasses. Periodically you could vent the gas in the reactor to a different vacuum chamber, let it sit for a few weeks and then process it.

  22. I don’t understand why their drain tank is so “involved”. Lots of piping, multiple tanks, etc.

    Really just needs to do two things: move fuel away from moderator, allow for cooling.

  23. “No dangerous radioactive gasses.”

    Well, only since it will remain a pulped cellulose reactor proposed by dilettantes in the most greentarded windmill state of the EU.

    In other words, it ain’t getting built:

    probability << snowball’s chance
    …on Venus
    …after the sun goes red giant.

    “Seaborg CUBE reactor can use spent nuclear fuel (SNF) by adding thorium as a catalyst. “

    That makes no sense, because thorium would only further dilute the fissile content of SNF. Never heard anybody call the 238U in normal fuel a “catalyst”. Stretching things there… but hey, this is pop science.

    Ahhh Denmark. Here we have the red-bearded long-haired Viking PhD in Denmark that wants to build nuclear reactors… nice hoodie; looks hygge. They wear collared shirts and ties at Westinghouse BTW (respectable – old school).

    Curiously, this auditorium on the molten salt tour appears to be fully half full – how optimistic. That is [half jam] packed compared to Lars Jorgensen’s (ThorCon) crowd at the other MSR ComicCon from last week’s NBF article.

    Beware any MSR article that doesn’t spend 90% of its words on off-gas and fuel handling. It’s not that I know where the bodies are buried; the problem is that the coffins are empty.


Leave a Comment