Terrestrial Energy Molten Salt nuclear plant regulator review progress

Terrestrial Energy’s Integral Molten Salt Reactor (IMSR) has entered the second phase of a vendor design review by the Canadian Nuclear Safety Commission (CNSC). The design was the first advanced reactor to complete the first phase of the CNSC’s regulatory pre-licensing review.

Walk away safe – no meltdown reactor

Terrestrial Energy aims to commercialize its walkaway safe molten salt modular reactor design in the late 2020s.

The use of a molten salt is at the heart of many virtues of the IMSR® and directly leads to IMSR®’s key commercial advantages – a cost-competitive and “walk-away” safe nuclear power plant.

The IMSR® uses a fundamentally different reactor technology — a liquid fuel, a molten salt, rather than the solid fuel used exclusively in conventional reactors. This provides a fluid medium to carry a nuclear fuel, a uranium fluoride salt. Molten salts are thermally stable and are excellent heat-transfer fluids, ideal for capturing and dissipating heat from the fission process.

An IMSR® power plant generates 400 megawatts of thermal energy (190 MW electric) with a thermal-spectrum, graphite-moderated, molten-fluoride-salt reactor system. It uses standard-assay low-enriched uranium (less than 5 percent 235U) fuel. It incorporates many aspects of Molten Salt Reactor operation researched, demonstrated and proven by test reactors at the Oak Ridge National Laboratory.

Swappable Core

The IMSR® improves upon earlier Molten Salt Reactor designs by incorporating key innovations that create an industrial reactor ready for commercial deployment.

The key challenge to MSR commercialization was graphite’s limited lifetime in a reactor core. Commercial power reactors require high energy densities in the reactor core to be economical, but such high-power densities significantly reduce the graphite moderator’s lifespan. Replacing the graphite moderator is difficult to do safely and economically.

The distinct IMSR® innovation is an elegant solution to this problem — integrating the primary reactor components, including the graphite moderator, into a sealed and replaceable reactor core. The IMSR® Core-unit, which has an operating lifetime of seven years, is simple and safe to replace. It supports high capacity factors of IMSR® power plants and hence high capital efficiency. It also ensures that the materials’ lifetime requirements of other reactor core components are met, a challenge often cited as an impediment to immediate commercialization of MSRs.

The result is a small modular reactor that delivers a combination of high energy output, simplicity of operation and cost-competitiveness necessary to drive broad commercial deployment.

Pre-licensing of Terrestrial Energy

The three-phase pre-licensing process will verify the acceptability of a design with respect to Canadian nuclear regulatory requirements and expectations.

Phase 1. a pre-licensing assessment of compliance with regulatory requirements

This was completed November 2017. CNSC said the company had demonstrated an understanding of the regulator’s requirements applicable to the design and safety analysis of the 400 MWt IMSR, known as IMSR400.

Phase 2. an assessment of any potential fundamental barriers to licensing

This is happening now. This involves a detailed follow-up of phase 1 activities, and an assessment of the IMSR design’s ability to meet all 19 focus areas of power plant licensing. It is expected to complete late in 2019.

Phase 3. a follow-up phase allowing the vendor to respond to findings from the second phase.

In June 2017, Terrestrial Energy began a feasibility study for the siting of the first commercial IMSR at Canadian Nuclear Laboratories’ Chalk River site.

In March 2018, Terrestrial and US utility Energy Northwest agreed a memorandum of understanding on the terms of the possible siting, construction and operation of an IMSR at a site at the Idaho National Laboratory in southeastern Idaho.

In September 2018, Terrestrial Energy USA partnered with the big utility Southern Company ($46 billion market value) and several US DOE national laboratories to investigate the production of hydrogen using its IMSR. The two-year research and development project will examine the efficiency, design and economics of using the IMSR to produce carbon-free, industrial-scale hydrogen using the hybrid sulfur process.

97 thoughts on “Terrestrial Energy Molten Salt nuclear plant regulator review progress”

  1. Will these reactors use thorium like the ones they were expirementing in the 70’s? If they are then we should start making them and soon.

  2. Terrestrial Energy states shows on their website that they aim for the cost of electricity to be cost competitive or slightly cheaper than CC natural gas https://www.terrestrialenergy.com/technology/competitive/ Besides, out of the two electricity generation is much easier than selling process heat, so if the IMSR’s not competitive on electrical generation then it probably wouldn’t be able to pull of process heat in the first place.

  3. If they can do that on first of a kind, nth of a kind should be competitive. Gas turbines have had decades, and billions of dollars worth, of development, much of it paid for to start with by the military. ( The US Air Force paid to develop the KC135 flying tanker, Boeing developed that into their 7n7 s, and GE adapted the engines for power plants.) Uranium is cheaper than gas per unit of heat released, even in North America.

  4. 1. Beating natural gas prices in the US is a high bar. 1a. Even if you can’t beat natural gas prices in the US you can be quite competitive elsewhere. 2. FOAK costs are much higher than nth of a kind. Nth may well be competitive with US gas. 3. OTOH there are still costs associated with transporting/handling the spent “fuel pods” but that is 13 years off from commercial use (7 years of use, 6 years of cooling down before shipping).

  5. Well the original idea was to create carbon free process heat to enable Canadian tar sand oil extraction.

  6. Thorium isn’t fissionable. It’s bred into U-233 which is what actually fissions. The breeder versions of MSRs are considerably more complex than what TEI is proposing.The best way to think of the IMSR is as if it’s just an alternative to an LEU fuel rod but with the reactor built into the fuel unit. When it reaches the point where the U-235 is burned up a crane pulls the whole unit out and replaces it with a fresh one similar to the way that a light-water reactor swaps out fuel rods when they’re burned up.

  7. 1. Beating natural gas prices in the US is a high bar.1a. Even if you can’t beat natural gas prices in the US you can be quite competitive elsewhere.2. FOAK costs are much higher than nth of a kind. Nth may well be competitive with US gas.3. OTOH there are still costs associated with transporting/handling the spent fuel pods”” but that is 13 years off from commercial use (7 years of use”””” 6 years of cooling down before shipping).”””

  8. Well the original idea was to create carbon free process heat to enable Canadian tar sand oil extraction.

  9. In a TEI IMSR? No. The only thing that a thorium breeder has in common with the IMSR is that they both use molten salt.

  10. Thorium isn’t fissionable. It’s bred into U-233, which is what actually fissions. The breeder versions of MSRs are considerably more complex than what TEI is proposing. The best way to think of the IMSR is as if it’s just an alternative to an LEU fuel rod, but with the reactor built into the fuel unit. When it reaches the point where the U-235 is burned up, a crane pulls the whole unit out and replaces it with a fresh one, similar to the way that a light-water reactor swaps out fuel rods when they’re burned up.

  11. In a TEI IMSR? No. The only thing that a thorium breeder has in common with the IMSR is that they both use molten salt.

  12. It’s a throw-away nuclear battery, where a Thorium cycle plant can keep feeding the fertile material into the outer jacket of fluid, converted to U233 in a month. A LFTR reactor can run a lot longer than these seven-year throwaways.

  13. But no Thorium cycle reactor. How lame. India, China, and the Netherlands are going to hand us our lunch with LFTR’s, a design that America pioneered.

  14. It’s a throw-away nuclear battery where a Thorium cycle plant can keep feeding the fertile material into the outer jacket of fluid converted to U233 in a month. A LFTR reactor can run a lot longer than these seven-year throwaways.

  15. But no Thorium cycle reactor. How lame. India China and the Netherlands are going to hand us our lunch with LFTR’s a design that America pioneered.

  16. ‘why graphite?’ Because if the heat ever equalises across your fuel/moderator wall, the water will explode, and the salt will freeze. You might be able to figure out how to ensure that never happens, and what effects phase changes might start to have on the neutronics and thermohydraulics, but why bother ? Graphite isn’t perfect as a moderator, but it works ok in eleven GW of Russian reactors and eight of British ones, with another starting up in China this year.

  17. ‘why graphite?’ Because if the heat ever equalises across your fuel/moderator wall the water will explode and the salt will freeze. You might be able to figure out how to ensure that never happens and what effects phase changes might start to have on the neutronics and thermohydraulics but why bother ? Graphite isn’t perfect as a moderator but it works ok in eleven GW of Russian reactors and eight of British ones with another starting up in China this year.

  18. Actually, the tube would just boil dry and thereby give negative reactivity feedback. Remember, it isn’t pressurized – none of it is. Try again. Graphite makes reactors BIG and of low power density. If you don’t want it to be BIG then you have to raise the enrichment BIG.

  19. Actually the tube would just boil dry and thereby give negative reactivity feedback. Remember it isn’t pressurized – none of it is. Try again.Graphite makes reactors BIG and of low power density. If you don’t want it to be BIG then you have to raise the enrichment BIG.

  20. Weren’t the Japanese going in that direction with their Reduced Moderation Reactor proposals – tighter lattice, with more uranium and less water ? On another tack, you answered a comment of mine about the Moltex paper reactor ( I can’t find the original thread.) I read somebody criticizing their design, saying they should get rid of the fluoride coolant as it eats too many neutrons, and use sodium metal instead. Reckoned that would get it up to breeding. Ian Scott’s reasoning was for ease of licensing. I’m not sure how much difference running on a fast spectrum would make to the hafnium’s high neutron cross-section, or to the effects of fuel shuffling. LWR fuel rods are over 1000 C at the centre, tapering off to 500 or so near the cladding. Moltex claim their plutonium chloride would be at average 760 C across all fuel tubes, with a centreline peak of 1075. That should give them a more even, and faster, negative temperature feedback compared to solid fuel, and maybe level the heat output. Ian Scott says the hafnium in the coolant will reduce neutron flux at the tank wall by four orders of magnitude, so if it’s about a metre from the fuel rods to the edge, and half a centimetre between rods, how many would you lose from the adjacent rod ? It wouldn’t be the first British reactor to fail to start up. When they were trying to make plutonium for their initial bomb, they got the neutron calculations sightly wrong, and had to take all the fuel rods out, and grind a sliver of aluminium off each by hand, before the thing would work. At least in this case the cure would be simpler – use nuclear grade zirconium for the coolant, with no hafnium in it, and put neutron shielding round the tank instead.

  21. As usual you’ve provided a discussion FAR more interesting than the 50 comments trying to connect the reactor choice back to nefarious dealings by Russians/Chinese/Nixon/Rickover/Kennedy/aliens/lizardmen/…

  22. in the fuel rods is comparable to BWRs. The volume of the core is huge and the power density low on a total volume basis, but the power density is high on a channel and fuel rod basis. The RBMK is like a CANDU where the graphite replaces the heavy water. CANDU would be low power density on a core basis but rather high power density on a channel and fuel rod basis. These MSR with the graphite core can be run at arbitrary power level – whatever the components and humor can support with reasonable lifetime. Seems that all the MSR guys want to use the graphite core for 4 years and then toss it – at least until they get some operating experience. Makes sense. All that I have been saying is that I am tempted to start looking at calculations for a MSR configuration where insulated water rods replace the graphite. I’m very certain that the neutronics would work out nicely using H2O or D2O and I’m surprised that it’s not discussed. It’s just a devils advocate thing… people think this MSR is different from that MSR and this one uses thorium and this one needs low enrichment that one high enrichment – they are all the same just driven by whatever assumptions they felt they should take. 18 months ago I was playing with a cylindrical configuration of UF4 with central region of H2 at like 0.6g/cc – it lit up just fine – just trying to keep from getting rusty with things I used to know.

  23. I see how I led you to see an apparent contradiction. One of my pet ideas IS a very large ‘undermoderated’ LWR that operates at a low power density. See, I work with LWRs that operate between 60-110 kw/liter with a refueling interval of 18 to 24 months. It should be clear that these reactors therefore have enough excess reactivity to go double the interval at half the power density. That is a fact; a 4GW PWR on an 18 month fuel cycle would could easily operate at 2GW for 36 months with the same exact core loading. Nobody would do that. The particular recipe of the pet idea is a huge low pressure LWR with a higher fuel to moderator ratio than seen in power reactors, (like the Shippingport PWR) for a desalination plant. Similar to the Chinese district heater reactors – LWR can be made big, low pressure, and with a low heat rate, low enrichment, and if undermoderated they can be pushed towards breeding (like Shippingport). Higher fuel to moderator ratio reduces the need for poisons (conserves neutrons), large size conserves reduces leakage (conserves neutrons) and pushes towards breeding. This is all to get exceedingly long refueling intervals – like 5 years to a decade. The breeding or conversion helps with that. The main thing is low power density lenghtens the refueling interval – relative to current commercial reactors because the fuel is similarly dense. It’s basically a solution looking for a problem – massive scale desalination. Low power density doesn’t play out the same in solid fuel graphite moderated reactors. The atom density of the fuel atoms in solid fueled graphite moderated reactors is very low compared to oxide fueled LWRs. The bulk of their volume is graphite and not fuel; the pebble as 10 grams of uranium in a tennis ball sized graphite pebble. All the high temp gas cooled reactors that use TRISO and graphite fuel have very bulky fuel elements/pebbles with little fuel mass. At ‘high’ burnup (energy per kilogram of fuel) they still ha

  24. Weren’t the Japanese going in that direction with their Reduced Moderation Reactor proposals – tighter lattice with more uranium and less water ? On another tack you answered a comment of mine about the Moltex paper reactor ( I can’t find the original thread.) I read somebody criticizing their design saying they should get rid of the fluoride coolant as it eats too many neutrons and use sodium metal instead. Reckoned that would get it up to breeding. Ian Scott’s reasoning was for ease of licensing. I’m not sure how much difference running on a fast spectrum would make to the hafnium’s high neutron cross-section or to the effects of fuel shuffling. LWR fuel rods are over 1000 C at the centre tapering off to 500 or so near the cladding. Moltex claim their plutonium chloride would be at average 760 C across all fuel tubes with a centreline peak of 1075. That should give them a more even and faster negative temperature feedback compared to solid fuel and maybe level the heat output. Ian Scott says the hafnium in the coolant will reduce neutron flux at the tank wall by four orders of magnitude so if it’s about a metre from the fuel rods to the edge and half a centimetre between rods how many would you lose from the adjacent rod ? It wouldn’t be the first British reactor to fail to start up. When they were trying to make plutonium for their initial bomb they got the neutron calculations sightly wrong and had to take all the fuel rods out and grind a sliver of aluminium off each by hand before the thing would work. At least in this case the cure would be simpler – use nuclear grade zirconium for the coolant with no hafnium in it and put neutron shielding round the tank instead.

  25. As usual you’ve provided a discussion FAR more interesting than the 50 comments trying to connect the reactor choice back to nefarious dealings by Russians/Chinese/Nixon/Rickover/Kennedy/aliens/lizardmen/…

  26. in the fuel rods is comparable to BWRs. The volume of the core is huge and the power density low on a total volume basis but the power density is high on a channel and fuel rod basis. The RBMK is like a CANDU where the graphite replaces the heavy water. CANDU would be low power density on a core basis but rather high power density on a channel and fuel rod basis.These MSR with the graphite core can be run at arbitrary power level – whatever the components and humor can support with reasonable lifetime. Seems that all the MSR guys want to use the graphite core for 4 years and then toss it – at least until they get some operating experience. Makes sense. All that I have been saying is that I am tempted to start looking at calculations for a MSR configuration where insulated water rods replace the graphite. I’m very certain that the neutronics would work out nicely using H2O or D2O and I’m surprised that it’s not discussed. It’s just a devils advocate thing… people think this MSR is different from that MSR and this one uses thorium and this one needs low enrichment that one high enrichment – they are all the same just driven by whatever assumptions they felt they should take. 18 months ago I was playing with a cylindrical configuration of UF4 with central region of H2 at like 0.6g/cc – it lit up just fine – just trying to keep from getting rusty with things I used to know.”

  27. I see how I led you to see an apparent contradiction. One of my pet ideas IS a very large ‘undermoderated’ LWR that operates at a low power density. See I work with LWRs that operate between 60-110 kw/liter with a refueling interval of 18 to 24 months. It should be clear that these reactors therefore have enough excess reactivity to go double the interval at half the power density. That is a fact; a 4GW PWR on an 18 month fuel cycle would could easily operate at 2GW for 36 months with the same exact core loading. Nobody would do that. The particular recipe of the pet idea is a huge low pressure LWR with a higher fuel to moderator ratio than seen in power reactors (like the Shippingport PWR) for a desalination plant. Similar to the Chinese district heater reactors – LWR can be made big low pressure and with a low heat rate low enrichment and if undermoderated they can be pushed towards breeding (like Shippingport). Higher fuel to moderator ratio reduces the need for poisons (conserves neutrons) large size conserves reduces leakage (conserves neutrons) and pushes towards breeding. This is all to get exceedingly long refueling intervals – like 5 years to a decade. The breeding or conversion helps with that. The main thing is low power density lenghtens the refueling interval – relative to current commercial reactors because the fuel is similarly dense. It’s basically a solution looking for a problem – massive scale desalination.Low power density doesn’t play out the same in solid fuel graphite moderated reactors. The atom density of the fuel atoms in solid fueled graphite moderated reactors is very low compared to oxide fueled LWRs. The bulk of their volume is graphite and not fuel; the pebble as 10 grams of uranium in a tennis ball sized graphite pebble. All the high temp gas cooled reactors that use TRISO and graphite fuel have very bulky fuel elements/pebbles with little fuel mass. At ‘high’ burnup (energy per kilogram of fuel) they still haven’

  28. Yep, reduced moderation is good for fuel utilization, you just can’t run it at the same heat rate (relative to normal commercial BWR/BWR) because it will be closer to ‘boiling crisis’ or ‘critical heat flux’ because it is relatively starved of coolant flow. Fluorine has a very low capture cross section per a google image search of capture vs. energy. It certainly doesn’t absorb thermal neutrons. There are some resonances between 10^4 and 10^6ev (per the image)…. maybe that is what you are talking about, but still – not a strong absorber. Sodium thermal cross section is 50x fluorine and it too has high energy resonances so I don’t know what the guy was saying about getting rid of the fluorine. Sure the fuel temperature profile in the Moltex ‘rod’ is going to be more flat than an oxide pellet – there will be convection in the pin and the thermal conductivity is prolly significantly better than ceramic. About the centerline fuel temperature in oxide fuel rods – it is high and the fuel stores a lot of energy compared to metallic fuel. This stored energy is bad in a LOCA. That said, we have spots in the core today running at 43kw/m in steady-state. Isn’t that incredible? That means that a 3.5cm segment of clad rod, at that peak location, is making 1.5kw by itself – it could heat a large room in the NJ winter! I don’t see why anybody would put hafnium bearing fluid in any reactor except to shut it down. Like I said, I consider them to be a ‘garage band’ to illustrate my point – not to be mean. I don’t see any real argument for Moltex – it might just be making some thesis for grad students. Yes, the British have had mixed luck forging their own path in nuclear. Rickover had to shove the SW1 reactor down their throats. Magnox and ACR were not very cheap to operate or fuel, and had to be de-tuned, but they have hundreds of successful reactor years of operating experience. I’m absolutely positive the RR could make a fine SMR and that the ACR technology c

  29. Yep reduced moderation is good for fuel utilization you just can’t run it at the same heat rate (relative to normal commercial BWR/BWR) because it will be closer to ‘boiling crisis’ or ‘critical heat flux’ because it is relatively starved of coolant flow.Fluorine has a very low capture cross section per a google image search of capture vs. energy. It certainly doesn’t absorb thermal neutrons. There are some resonances between 10^4 and 10^6ev (per the image)…. maybe that is what you are talking about but still – not a strong absorber. Sodium thermal cross section is 50x fluorine and it too has high energy resonances so I don’t know what the guy was saying about getting rid of the fluorine. Sure the fuel temperature profile in the Moltex ‘rod’ is going to be more flat than an oxide pellet – there will be convection in the pin and the thermal conductivity is prolly significantly better than ceramic. About the centerline fuel temperature in oxide fuel rods – it is high and the fuel stores a lot of energy compared to metallic fuel. This stored energy is bad in a LOCA. That said we have spots in the core today running at 43kw/m in steady-state. Isn’t that incredible? That means that a 3.5cm segment of clad rod at that peak location is making 1.5kw by itself – it could heat a large room in the NJ winter!I don’t see why anybody would put hafnium bearing fluid in any reactor except to shut it down. Like I said I consider them to be a ‘garage band’ to illustrate my point – not to be mean. I don’t see any real argument for Moltex – it might just be making some thesis for grad students.Yes the British have had mixed luck forging their own path in nuclear. Rickover had to shove the SW1 reactor down their throats. Magnox and ACR were not very cheap to operate or fuel and had to be de-tuned but they have hundreds of successful reactor years of operating experience. I’m absolutely positive the RR could make a fine SMR and that the AC

  30. Yep, reduced moderation is good for fuel utilization, you just can’t run it at the same heat rate (relative to normal commercial BWR/BWR) because it will be closer to ‘boiling crisis’ or ‘critical heat flux’ because it is relatively starved of coolant flow. Fluorine has a very low capture cross section per a google image search of capture vs. energy. It certainly doesn’t absorb thermal neutrons. There are some resonances between 10^4 and 10^6ev (per the image)…. maybe that is what you are talking about, but still – not a strong absorber. Sodium thermal cross section is 50x fluorine and it too has high energy resonances so I don’t know what the guy was saying about getting rid of the fluorine. Sure the fuel temperature profile in the Moltex ‘rod’ is going to be more flat than an oxide pellet – there will be convection in the pin and the thermal conductivity is prolly significantly better than ceramic. About the centerline fuel temperature in oxide fuel rods – it is high and the fuel stores a lot of energy compared to metallic fuel. This stored energy is bad in a LOCA. That said, we have spots in the core today running at 43kw/m in steady-state. Isn’t that incredible? That means that a 3.5cm segment of clad rod, at that peak location, is making 1.5kw by itself – it could heat a large room in the NJ winter! I don’t see why anybody would put hafnium bearing fluid in any reactor except to shut it down. Like I said, I consider them to be a ‘garage band’ to illustrate my point – not to be mean. I don’t see any real argument for Moltex – it might just be making some thesis for grad students. Yes, the British have had mixed luck forging their own path in nuclear. Rickover had to shove the SW1 reactor down their throats. Magnox and ACR were not very cheap to operate or fuel, and had to be de-tuned, but they have hundreds of successful reactor years of operating experience. I’m absolutely positive the RR could make a fine SMR and that the ACR technology c

  31. Yep reduced moderation is good for fuel utilization you just can’t run it at the same heat rate (relative to normal commercial BWR/BWR) because it will be closer to ‘boiling crisis’ or ‘critical heat flux’ because it is relatively starved of coolant flow.Fluorine has a very low capture cross section per a google image search of capture vs. energy. It certainly doesn’t absorb thermal neutrons. There are some resonances between 10^4 and 10^6ev (per the image)…. maybe that is what you are talking about but still – not a strong absorber. Sodium thermal cross section is 50x fluorine and it too has high energy resonances so I don’t know what the guy was saying about getting rid of the fluorine. Sure the fuel temperature profile in the Moltex ‘rod’ is going to be more flat than an oxide pellet – there will be convection in the pin and the thermal conductivity is prolly significantly better than ceramic. About the centerline fuel temperature in oxide fuel rods – it is high and the fuel stores a lot of energy compared to metallic fuel. This stored energy is bad in a LOCA. That said we have spots in the core today running at 43kw/m in steady-state. Isn’t that incredible? That means that a 3.5cm segment of clad rod at that peak location is making 1.5kw by itself – it could heat a large room in the NJ winter!I don’t see why anybody would put hafnium bearing fluid in any reactor except to shut it down. Like I said I consider them to be a ‘garage band’ to illustrate my point – not to be mean. I don’t see any real argument for Moltex – it might just be making some thesis for grad students.Yes the British have had mixed luck forging their own path in nuclear. Rickover had to shove the SW1 reactor down their throats. Magnox and ACR were not very cheap to operate or fuel and had to be de-tuned but they have hundreds of successful reactor years of operating experience. I’m absolutely positive the RR could make a fine SMR and that the AC

  32. Weren’t the Japanese going in that direction with their Reduced Moderation Reactor proposals – tighter lattice, with more uranium and less water ? On another tack, you answered a comment of mine about the Moltex paper reactor ( I can’t find the original thread.) I read somebody criticizing their design, saying they should get rid of the fluoride coolant as it eats too many neutrons, and use sodium metal instead. Reckoned that would get it up to breeding. Ian Scott’s reasoning was for ease of licensing. I’m not sure how much difference running on a fast spectrum would make to the hafnium’s high neutron cross-section, or to the effects of fuel shuffling. LWR fuel rods are over 1000 C at the centre, tapering off to 500 or so near the cladding. Moltex claim their plutonium chloride would be at average 760 C across all fuel tubes, with a centreline peak of 1075. That should give them a more even, and faster, negative temperature feedback compared to solid fuel, and maybe level the heat output. Ian Scott says the hafnium in the coolant will reduce neutron flux at the tank wall by four orders of magnitude, so if it’s about a metre from the fuel rods to the edge, and half a centimetre between rods, how many would you lose from the adjacent rod ? It wouldn’t be the first British reactor to fail to start up. When they were trying to make plutonium for their initial bomb, they got the neutron calculations sightly wrong, and had to take all the fuel rods out, and grind a sliver of aluminium off each by hand, before the thing would work. At least in this case the cure would be simpler – use nuclear grade zirconium for the coolant, with no hafnium in it, and put neutron shielding round the tank instead.

  33. Weren’t the Japanese going in that direction with their Reduced Moderation Reactor proposals – tighter lattice with more uranium and less water ? On another tack you answered a comment of mine about the Moltex paper reactor ( I can’t find the original thread.) I read somebody criticizing their design saying they should get rid of the fluoride coolant as it eats too many neutrons and use sodium metal instead. Reckoned that would get it up to breeding. Ian Scott’s reasoning was for ease of licensing. I’m not sure how much difference running on a fast spectrum would make to the hafnium’s high neutron cross-section or to the effects of fuel shuffling. LWR fuel rods are over 1000 C at the centre tapering off to 500 or so near the cladding. Moltex claim their plutonium chloride would be at average 760 C across all fuel tubes with a centreline peak of 1075. That should give them a more even and faster negative temperature feedback compared to solid fuel and maybe level the heat output. Ian Scott says the hafnium in the coolant will reduce neutron flux at the tank wall by four orders of magnitude so if it’s about a metre from the fuel rods to the edge and half a centimetre between rods how many would you lose from the adjacent rod ? It wouldn’t be the first British reactor to fail to start up. When they were trying to make plutonium for their initial bomb they got the neutron calculations sightly wrong and had to take all the fuel rods out and grind a sliver of aluminium off each by hand before the thing would work. At least in this case the cure would be simpler – use nuclear grade zirconium for the coolant with no hafnium in it and put neutron shielding round the tank instead.

  34. Yep, reduced moderation is good for fuel utilization, you just can’t run it at the same heat rate (relative to normal commercial BWR/BWR) because it will be closer to ‘boiling crisis’ or ‘critical heat flux’ because it is relatively starved of coolant flow.

    Fluorine has a very low capture cross section per a google image search of capture vs. energy. It certainly doesn’t absorb thermal neutrons. There are some resonances between 10^4 and 10^6ev (per the image)…. maybe that is what you are talking about, but still – not a strong absorber. Sodium thermal cross section is 50x fluorine and it too has high energy resonances so I don’t know what the guy was saying about getting rid of the fluorine.

    Sure the fuel temperature profile in the Moltex ‘rod’ is going to be more flat than an oxide pellet – there will be convection in the pin and the thermal conductivity is prolly significantly better than ceramic. About the centerline fuel temperature in oxide fuel rods – it is high and the fuel stores a lot of energy compared to metallic fuel. This stored energy is bad in a LOCA. That said, we have spots in the core today running at 43kw/m in steady-state. Isn’t that incredible? That means that a 3.5cm segment of clad rod, at that peak location, is making 1.5kw by itself – it could heat a large room in the NJ winter!

    I don’t see why anybody would put hafnium bearing fluid in any reactor except to shut it down. Like I said, I consider them to be a ‘garage band’ to illustrate my point – not to be mean. I don’t see any real argument for Moltex – it might just be making some thesis for grad students.

    Yes, the British have had mixed luck forging their own path in nuclear. Rickover had to shove the SW1 reactor down their throats. Magnox and ACR were not very cheap to operate or fuel, and had to be de-tuned, but they have hundreds of successful reactor years of operating experience. I’m absolutely positive the RR could make a fine SMR and that the ACR technology could be re-vamped into Gen4, but the Brits aren’t as keen on forging their own path any longer – better to work with the global consensus methinks. It’s not a competition – all nonmilitary operating experience is shared and should continue to be shared.

  35. As usual you’ve provided a discussion FAR more interesting than the 50 comments trying to connect the reactor choice back to nefarious dealings by Russians/Chinese/Nixon/Rickover/Kennedy/aliens/lizardmen/…

  36. As usual you’ve provided a discussion FAR more interesting than the 50 comments trying to connect the reactor choice back to nefarious dealings by Russians/Chinese/Nixon/Rickover/Kennedy/aliens/lizardmen/…

  37. in the fuel rods is comparable to BWRs. The volume of the core is huge and the power density low on a total volume basis, but the power density is high on a channel and fuel rod basis. The RBMK is like a CANDU where the graphite replaces the heavy water. CANDU would be low power density on a core basis but rather high power density on a channel and fuel rod basis. These MSR with the graphite core can be run at arbitrary power level – whatever the components and humor can support with reasonable lifetime. Seems that all the MSR guys want to use the graphite core for 4 years and then toss it – at least until they get some operating experience. Makes sense. All that I have been saying is that I am tempted to start looking at calculations for a MSR configuration where insulated water rods replace the graphite. I’m very certain that the neutronics would work out nicely using H2O or D2O and I’m surprised that it’s not discussed. It’s just a devils advocate thing… people think this MSR is different from that MSR and this one uses thorium and this one needs low enrichment that one high enrichment – they are all the same just driven by whatever assumptions they felt they should take. 18 months ago I was playing with a cylindrical configuration of UF4 with central region of H2 at like 0.6g/cc – it lit up just fine – just trying to keep from getting rusty with things I used to know.

  38. in the fuel rods is comparable to BWRs. The volume of the core is huge and the power density low on a total volume basis but the power density is high on a channel and fuel rod basis. The RBMK is like a CANDU where the graphite replaces the heavy water. CANDU would be low power density on a core basis but rather high power density on a channel and fuel rod basis.These MSR with the graphite core can be run at arbitrary power level – whatever the components and humor can support with reasonable lifetime. Seems that all the MSR guys want to use the graphite core for 4 years and then toss it – at least until they get some operating experience. Makes sense. All that I have been saying is that I am tempted to start looking at calculations for a MSR configuration where insulated water rods replace the graphite. I’m very certain that the neutronics would work out nicely using H2O or D2O and I’m surprised that it’s not discussed. It’s just a devils advocate thing… people think this MSR is different from that MSR and this one uses thorium and this one needs low enrichment that one high enrichment – they are all the same just driven by whatever assumptions they felt they should take. 18 months ago I was playing with a cylindrical configuration of UF4 with central region of H2 at like 0.6g/cc – it lit up just fine – just trying to keep from getting rusty with things I used to know.”

  39. I see how I led you to see an apparent contradiction. One of my pet ideas IS a very large ‘undermoderated’ LWR that operates at a low power density. See, I work with LWRs that operate between 60-110 kw/liter with a refueling interval of 18 to 24 months. It should be clear that these reactors therefore have enough excess reactivity to go double the interval at half the power density. That is a fact; a 4GW PWR on an 18 month fuel cycle would could easily operate at 2GW for 36 months with the same exact core loading. Nobody would do that. The particular recipe of the pet idea is a huge low pressure LWR with a higher fuel to moderator ratio than seen in power reactors, (like the Shippingport PWR) for a desalination plant. Similar to the Chinese district heater reactors – LWR can be made big, low pressure, and with a low heat rate, low enrichment, and if undermoderated they can be pushed towards breeding (like Shippingport). Higher fuel to moderator ratio reduces the need for poisons (conserves neutrons), large size conserves reduces leakage (conserves neutrons) and pushes towards breeding. This is all to get exceedingly long refueling intervals – like 5 years to a decade. The breeding or conversion helps with that. The main thing is low power density lenghtens the refueling interval – relative to current commercial reactors because the fuel is similarly dense. It’s basically a solution looking for a problem – massive scale desalination. Low power density doesn’t play out the same in solid fuel graphite moderated reactors. The atom density of the fuel atoms in solid fueled graphite moderated reactors is very low compared to oxide fueled LWRs. The bulk of their volume is graphite and not fuel; the pebble as 10 grams of uranium in a tennis ball sized graphite pebble. All the high temp gas cooled reactors that use TRISO and graphite fuel have very bulky fuel elements/pebbles with little fuel mass. At ‘high’ burnup (energy per kilogram of fuel) they still ha

  40. I see how I led you to see an apparent contradiction. One of my pet ideas IS a very large ‘undermoderated’ LWR that operates at a low power density. See I work with LWRs that operate between 60-110 kw/liter with a refueling interval of 18 to 24 months. It should be clear that these reactors therefore have enough excess reactivity to go double the interval at half the power density. That is a fact; a 4GW PWR on an 18 month fuel cycle would could easily operate at 2GW for 36 months with the same exact core loading. Nobody would do that. The particular recipe of the pet idea is a huge low pressure LWR with a higher fuel to moderator ratio than seen in power reactors (like the Shippingport PWR) for a desalination plant. Similar to the Chinese district heater reactors – LWR can be made big low pressure and with a low heat rate low enrichment and if undermoderated they can be pushed towards breeding (like Shippingport). Higher fuel to moderator ratio reduces the need for poisons (conserves neutrons) large size conserves reduces leakage (conserves neutrons) and pushes towards breeding. This is all to get exceedingly long refueling intervals – like 5 years to a decade. The breeding or conversion helps with that. The main thing is low power density lenghtens the refueling interval – relative to current commercial reactors because the fuel is similarly dense. It’s basically a solution looking for a problem – massive scale desalination.Low power density doesn’t play out the same in solid fuel graphite moderated reactors. The atom density of the fuel atoms in solid fueled graphite moderated reactors is very low compared to oxide fueled LWRs. The bulk of their volume is graphite and not fuel; the pebble as 10 grams of uranium in a tennis ball sized graphite pebble. All the high temp gas cooled reactors that use TRISO and graphite fuel have very bulky fuel elements/pebbles with little fuel mass. At ‘high’ burnup (energy per kilogram of fuel) they still haven’

  41. Weren’t the Japanese going in that direction with their Reduced Moderation Reactor proposals – tighter lattice, with more uranium and less water ?
    On another tack, you answered a comment of mine about the Moltex paper reactor ( I can’t find the original thread.) I read somebody criticizing their design, saying they should get rid of the fluoride coolant as it eats too many neutrons, and use sodium metal instead. Reckoned that would get it up to breeding. Ian Scott’s reasoning was for ease of licensing.
    I’m not sure how much difference running on a fast spectrum would make to the hafnium’s high neutron cross-section, or to the effects of fuel shuffling. LWR fuel rods are over 1000 C at the centre, tapering off to 500 or so near the cladding. Moltex claim their plutonium chloride would be at average 760 C across all fuel tubes, with a centreline peak of 1075. That should give them a more even, and faster, negative temperature feedback compared to solid fuel, and maybe level the heat output.
    Ian Scott says the hafnium in the coolant will reduce neutron flux at the tank wall by four orders of magnitude, so if it’s about a metre from the fuel rods to the edge, and half a centimetre between rods, how many would you lose from the adjacent rod ? It wouldn’t be the first British reactor to fail to start up. When they were trying to make plutonium for their initial bomb, they got the neutron calculations sightly wrong, and had to take all the fuel rods out, and grind a sliver of aluminium off each by hand, before the thing would work. At least in this case the cure would be simpler – use nuclear grade zirconium for the coolant, with no hafnium in it, and put neutron shielding round the tank instead.

  42. Actually, the tube would just boil dry and thereby give negative reactivity feedback. Remember, it isn’t pressurized – none of it is. Try again. Graphite makes reactors BIG and of low power density. If you don’t want it to be BIG then you have to raise the enrichment BIG.

  43. Actually the tube would just boil dry and thereby give negative reactivity feedback. Remember it isn’t pressurized – none of it is. Try again.Graphite makes reactors BIG and of low power density. If you don’t want it to be BIG then you have to raise the enrichment BIG.

  44. As usual you’ve provided a discussion FAR more interesting than the 50 comments trying to connect the reactor choice back to nefarious dealings by Russians/Chinese/Nixon/Rickover/Kennedy/aliens/lizardmen/…

  45. in the fuel rods is comparable to BWRs. The volume of the core is huge and the power density low on a total volume basis, but the power density is high on a channel and fuel rod basis. The RBMK is like a CANDU where the graphite replaces the heavy water. CANDU would be low power density on a core basis but rather high power density on a channel and fuel rod basis.

    These MSR with the graphite core can be run at arbitrary power level – whatever the components and humor can support with reasonable lifetime. Seems that all the MSR guys want to use the graphite core for 4 years and then toss it – at least until they get some operating experience. Makes sense. All that I have been saying is that I am tempted to start looking at calculations for a MSR configuration where insulated water rods replace the graphite. I’m very certain that the neutronics would work out nicely using H2O or D2O and I’m surprised that it’s not discussed. It’s just a devils advocate thing… people think this MSR is different from that MSR and this one uses thorium and this one needs low enrichment that one high enrichment – they are all the same just driven by whatever assumptions they felt they should take. 18 months ago I was playing with a cylindrical configuration of UF4 with central region of H2 at like 0.6g/cc – it lit up just fine – just trying to keep from getting rusty with things I used to know.

  46. I see how I led you to see an apparent contradiction. One of my pet ideas IS a very large ‘undermoderated’ LWR that operates at a low power density. See, I work with LWRs that operate between 60-110 kw/liter with a refueling interval of 18 to 24 months. It should be clear that these reactors therefore have enough excess reactivity to go double the interval at half the power density. That is a fact; a 4GW PWR on an 18 month fuel cycle would could easily operate at 2GW for 36 months with the same exact core loading. Nobody would do that. The particular recipe of the pet idea is a huge low pressure LWR with a higher fuel to moderator ratio than seen in power reactors, (like the Shippingport PWR) for a desalination plant. Similar to the Chinese district heater reactors – LWR can be made big, low pressure, and with a low heat rate, low enrichment, and if undermoderated they can be pushed towards breeding (like Shippingport). Higher fuel to moderator ratio reduces the need for poisons (conserves neutrons), large size conserves reduces leakage (conserves neutrons) and pushes towards breeding. This is all to get exceedingly long refueling intervals – like 5 years to a decade. The breeding or conversion helps with that. The main thing is low power density lenghtens the refueling interval – relative to current commercial reactors because the fuel is similarly dense. It’s basically a solution looking for a problem – massive scale desalination.

    Low power density doesn’t play out the same in solid fuel graphite moderated reactors. The atom density of the fuel atoms in solid fueled graphite moderated reactors is very low compared to oxide fueled LWRs. The bulk of their volume is graphite and not fuel; the pebble as 10 grams of uranium in a tennis ball sized graphite pebble. All the high temp gas cooled reactors that use TRISO and graphite fuel have very bulky fuel elements/pebbles with little fuel mass. At ‘high’ burnup (energy per kilogram of fuel) they still haven’t produced much energy on a core volume basis compared to LWR. These reactors (like the Chinese HTGRs) are small diameter in order to allow radial heat conduction when isolated from the normal heat sink and they necessarily use high enrichment (10-15%) because the small diameter means they leak 10% of the neutron population in the radial direction. So, they are low power density, and use high enrichment because the diameter is small. The fuel is expensive because it is built up by physical vapor deposition. Sintered oxide fuel in clad tubes is very cheap in comparision and it packs a lot of fuel atoms into the volume. Oxide fuel pellets pack ~11g of uranium into a cubic centimeter as opposed to 10g in a tennis ball of graphite for the pebble beds. Reactors should contain very dense fuel; it’s fundamental – it also minimizes the volume of waste (besides the point).

    The RBMK is huge and powerful but the power is made in fuel assemblies in channels and the linear heat rate in t

  47. Actually, the tube would just boil dry and thereby give negative reactivity feedback. Remember, it isn’t pressurized – none of it is. Try again.

    Graphite makes reactors BIG and of low power density. If you don’t want it to be BIG then you have to raise the enrichment BIG.

  48. ‘why graphite?’ Because if the heat ever equalises across your fuel/moderator wall, the water will explode, and the salt will freeze. You might be able to figure out how to ensure that never happens, and what effects phase changes might start to have on the neutronics and thermohydraulics, but why bother ? Graphite isn’t perfect as a moderator, but it works ok in eleven GW of Russian reactors and eight of British ones, with another starting up in China this year.

  49. ‘why graphite?’ Because if the heat ever equalises across your fuel/moderator wall the water will explode and the salt will freeze. You might be able to figure out how to ensure that never happens and what effects phase changes might start to have on the neutronics and thermohydraulics but why bother ? Graphite isn’t perfect as a moderator but it works ok in eleven GW of Russian reactors and eight of British ones with another starting up in China this year.

  50. As Huawei and half a dozen other Chinese companies demonstrate, copying an iPhone is not difficult either.

  51. As Huawei and half a dozen other Chinese companies demonstrate copying an iPhone is not difficult either.

  52. ‘why graphite?’ Because if the heat ever equalises across your fuel/moderator wall, the water will explode, and the salt will freeze. You might be able to figure out how to ensure that never happens, and what effects phase changes might start to have on the neutronics and thermohydraulics, but why bother ? Graphite isn’t perfect as a moderator, but it works ok in eleven GW of Russian reactors and eight of British ones, with another starting up in China this year.

  53. It’s a throw-away nuclear battery, where a Thorium cycle plant can keep feeding the fertile material into the outer jacket of fluid, converted to U233 in a month. A LFTR reactor can run a lot longer than these seven-year throwaways.

  54. It’s a throw-away nuclear battery where a Thorium cycle plant can keep feeding the fertile material into the outer jacket of fluid converted to U233 in a month. A LFTR reactor can run a lot longer than these seven-year throwaways.

  55. But no Thorium cycle reactor. How lame. India, China, and the Netherlands are going to hand us our lunch with LFTR’s, a design that America pioneered.

  56. But no Thorium cycle reactor. How lame. India China and the Netherlands are going to hand us our lunch with LFTR’s a design that America pioneered.

  57. It’s a throw-away nuclear battery, where a Thorium cycle plant can keep feeding the fertile material into the outer jacket of fluid, converted to U233 in a month. A LFTR reactor can run a lot longer than these seven-year throwaways.

  58. In a TEI IMSR? No. The only thing that a thorium breeder has in common with the IMSR is that they both use molten salt.

  59. In a TEI IMSR? No. The only thing that a thorium breeder has in common with the IMSR is that they both use molten salt.

  60. Thorium isn’t fissionable. It’s bred into U-233, which is what actually fissions. The breeder versions of MSRs are considerably more complex than what TEI is proposing. The best way to think of the IMSR is as if it’s just an alternative to an LEU fuel rod, but with the reactor built into the fuel unit. When it reaches the point where the U-235 is burned up, a crane pulls the whole unit out and replaces it with a fresh one, similar to the way that a light-water reactor swaps out fuel rods when they’re burned up.

  61. Thorium isn’t fissionable. It’s bred into U-233 which is what actually fissions. The breeder versions of MSRs are considerably more complex than what TEI is proposing.The best way to think of the IMSR is as if it’s just an alternative to an LEU fuel rod but with the reactor built into the fuel unit. When it reaches the point where the U-235 is burned up a crane pulls the whole unit out and replaces it with a fresh one similar to the way that a light-water reactor swaps out fuel rods when they’re burned up.

  62. 1. Beating natural gas prices in the US is a high bar. 1a. Even if you can’t beat natural gas prices in the US you can be quite competitive elsewhere. 2. FOAK costs are much higher than nth of a kind. Nth may well be competitive with US gas. 3. OTOH there are still costs associated with transporting/handling the spent “fuel pods” but that is 13 years off from commercial use (7 years of use, 6 years of cooling down before shipping).

  63. 1. Beating natural gas prices in the US is a high bar.1a. Even if you can’t beat natural gas prices in the US you can be quite competitive elsewhere.2. FOAK costs are much higher than nth of a kind. Nth may well be competitive with US gas.3. OTOH there are still costs associated with transporting/handling the spent fuel pods”” but that is 13 years off from commercial use (7 years of use”””” 6 years of cooling down before shipping).”””

  64. Well the original idea was to create carbon free process heat to enable Canadian tar sand oil extraction.

  65. Well the original idea was to create carbon free process heat to enable Canadian tar sand oil extraction.

  66. Thorium isn’t fissionable. It’s bred into U-233, which is what actually fissions. The breeder versions of MSRs are considerably more complex than what TEI is proposing.

    The best way to think of the IMSR is as if it’s just an alternative to an LEU fuel rod, but with the reactor built into the fuel unit. When it reaches the point where the U-235 is burned up, a crane pulls the whole unit out and replaces it with a fresh one, similar to the way that a light-water reactor swaps out fuel rods when they’re burned up.

  67. If they can do that on first of a kind, nth of a kind should be competitive. Gas turbines have had decades, and billions of dollars worth, of development, much of it paid for to start with by the military. ( The US Air Force paid to develop the KC135 flying tanker, Boeing developed that into their 7n7 s, and GE adapted the engines for power plants.) Uranium is cheaper than gas per unit of heat released, even in North America.

  68. If they can do that on first of a kind nth of a kind should be competitive. Gas turbines have had decades and billions of dollars worth of development much of it paid for to start with by the military. ( The US Air Force paid to develop the KC135 flying tanker Boeing developed that into their 7n7 s and GE adapted the engines for power plants.)Uranium is cheaper than gas per unit of heat released even in North America.

  69. 1. Beating natural gas prices in the US is a high bar.
    1a. Even if you can’t beat natural gas prices in the US you can be quite competitive elsewhere.
    2. FOAK costs are much higher than nth of a kind. Nth may well be competitive with US gas.
    3. OTOH there are still costs associated with transporting/handling the spent “fuel pods” but that is 13 years off from commercial use (7 years of use, 6 years of cooling down before shipping).

  70. Will these reactors use thorium like the ones they were expirementing in the 70’s? If they are then we should start making them and soon.

  71. Will these reactors use thorium like the ones they were expirementing in the 70’s? If they are then we should start making them and soon.

  72. Terrestrial Energy states shows on their website that they aim for the cost of electricity to be cost competitive or slightly cheaper than CC natural gas https://www.terrestrialenergy.com/technology/competitive/ Besides, out of the two electricity generation is much easier than selling process heat, so if the IMSR’s not competitive on electrical generation then it probably wouldn’t be able to pull of process heat in the first place.

  73. Terrestrial Energy states shows on their website that they aim for the cost of electricity to be cost competitive or slightly cheaper than CC natural gas https://www.terrestrialenergy.com/technology/competitive/Besides out of the two electricity generation is much easier than selling process heat so if the IMSR’s not competitive on electrical generation then it probably wouldn’t be able to pull of process heat in the first place.

  74. If they can do that on first of a kind, nth of a kind should be competitive. Gas turbines have had decades, and billions of dollars worth, of development, much of it paid for to start with by the military. ( The US Air Force paid to develop the KC135 flying tanker, Boeing developed that into their 7n7 s, and GE adapted the engines for power plants.)
    Uranium is cheaper than gas per unit of heat released, even in North America.

  75. Terrestrial Energy states shows on their website that they aim for the cost of electricity to be cost competitive or slightly cheaper than CC natural gas https://www.terrestrialenergy.com/technology/competitive/

    Besides, out of the two electricity generation is much easier than selling process heat, so if the IMSR’s not competitive on electrical generation then it probably wouldn’t be able to pull of process heat in the first place.

Comments are closed.