4 ms·

Number of electricity-generating fusion reactors built: 0 Number of electricity-generating molten salt reactors built: 0 I agree that the challenges of scalin

Number of electricity-generating fusion reactors built: 0 Number of electricity-generating molten salt reactors built: 0 I agree that the challenges of scaling up and industrializing molten salt fission reactors appear significantly less daunting than those of industrializing fusion power. But there is still R&D required. The last molten salt reactor to actually attain criticality shut down in 1969 (ORNL Molten Salt Reactor Experiment). That experiment was too small for direct translation to industrial practice (7.4 MW-thermal, e.g. 3.3 MW-electrical with 45% efficient conversion). It didn't generate electricity, didn't operate long enough to demonstrate industrial durability (only 13,172 hours (1.5 years) of full-power-equivalent operation), and didn't demonstrate the neutron economy of a breeder or even a near-steady-state burner reactor. There are a lot of steps remaining before any molten salt reactor becomes a commercially available product that a utility company can buy from a nuclear supplier. It would take many fewer steps than for a fusion reactor, but saying that molten salt reactors can replace all carbon we use now with current technology and current manufacturing is way overstating their technological maturity and availability.

Animats 9y ago
Molten salt reactors require an on-site chemical plant that operates on radioactive salt. Radioactive chemical plants are a pain. Lots of people, lots of plumbing, highly reactive and corrosive materials, risks of leaks. The history of nuclear fuel reprocessing plants isn't good. Most of them are now Superfund toxic waste sites. The good thing about PWR and BWR reactors is that the radioactive portion of the system is mechanically simple. All the complexity is outside. The working fluid is water, which is easy to handle and, other than the production of tritium, doesn't become radioactive. Every reactor design which had complicated stuff happening in the radioactive portion of the system has been a commercial failure.
- DennisP 9y ago
  LFTRs require on on-site chemical plants but that's not the case for some of the MSRs fueled by uranium.
- philipkglass 9y ago
  Very true, though fuels that start as water-soluble halide salts should theoretically produce somewhat less complicated flowsheets. You have elegantly summarized why adding a fuel reprocessing step to commercial BWR/PWR fleets makes the systemic costs of nuclear power higher and the waste problems worse. EDIT: oh, uranium and thorium halides hydrolyze except at very low pH. And so will a number of fission-product-halides. So maybe not simpler to start with halide salts instead of oxide fuels.
  - mjevans 9y ago
    Am I correct in understanding that "waste reduction" (breeder) reactors essentially burn more (most?) the fuel in the core, without that external re-processing?
    - philipkglass 9y ago
      Breeder reactors still require a chemical re-processing step to separate fission products (the lighter atoms produced when heavy actinides like uranium split and release energy) from the fuel. There are two main theoretical advantages for breeder reactors. One advantage is that the total energy extractable from terrestrial mineral resources is much greater with dedicated breeder reactors than with reliance on naturally fissile U-235 in simpler non-breeding reactors. In practice, nuclear power production has not grown fast enough to cause uranium shortages. Until there is a significant uranium shortage, reflected in much higher prices for freshly mined uranium, the higher complexity/cost of a breeder reactor cannot be offset by cost savings on fresh uranium fuel. The other theoretical advantage is a reduction in actinide waste production (plutonium, americium, curium). These heavy waste atoms dominate the radioactivity of spent nuclear fuel in the medium term, thousands to tens of thousands of years from now. But in the short term (years, decades, and centuries), the radioactivity of spent nuclear fuel is dominated by fission products rather than actinides. Breeder reactors do not notable reduce production of fission products compared to conventional reactors. Further, the reprocessing of nuclear fuel to separate the fission products has a tendency to turn potent but small, easily-handled waste problems into less potent but much larger and more unwieldy waste problems. Even though there are some theoretical waste advantages to reprocessing fuel instead of using it once and sealing it up forever, in practice waste reprocessing has always led to higher expenses and more radionuclides actually released into the wider environment.
- nickik 9y ago
  It does not need to be on-site. In a liquid fuel reactor one could do batch processing. Some startups actually propose to produce reactors with the fuel inside and then recycle the hole reactor. Chemical plants might be a pain but its far easier then doing isotopic separation.
nickik 9y ago
There is some R&D required, but nothing very fundamental. My point is that the technical challenges are pretty minor compared to the rest. Nobody believes what I am proposing is impossible. There might be a discussion if it can be built in 2, 5 to have a full scale reactor that could be commercialized. The real point is that this could have been done 30 years ago, 20 years ago, 10 years ago or at any time. The issue was never a technical one. If the government made a serious push including regulatory changes, action by national labs and copy the NASA approach to commercialization I am certain we could mass produce these things in under 10 years. Once you mass produce reactors the energy production capacity you produce is quite unbelievable. It dwarfs everything else. > current manufacturing My point about manufacturing is that reactors are not more complicated or need higher tolerances then rocket engines for example. SpaceX is producing rocket engines for 2M a pop at a rate of more then a 100 year. Boeing is producing far more complex air planes for 200M a pop at the rate of 1 per week. It is far easier then producing all these wind, solar and battery plants. The needed scales of production are multible order of magnitude different. A mass produced nuclear reactor will produce between 100-300MW, while most wind turbine produce around 3MW while being just as hard to produce and harder to install.

Animats 9y ago

Molten salt reactors require an on-site chemical plant that operates on radioactive salt. Radioactive chemical plants are a pain. Lots of people, lots of plumbing, highly reactive and corrosive materials, risks of leaks. The history of nuclear fuel reprocessing plants isn't good. Most of them are now Superfund toxic waste sites. The good thing about PWR and BWR reactors is that the radioactive portion of the system is mechanically simple. All the complexity is outside. The working fluid is water, which is easy to handle and, other than the production of tritium, doesn't become radioactive. Every reactor design which had complicated stuff happening in the radioactive portion of the system has been a commercial failure.

DennisP 9y ago
LFTRs require on on-site chemical plants but that's not the case for some of the MSRs fueled by uranium.
philipkglass 9y ago
Very true, though fuels that start as water-soluble halide salts should theoretically produce somewhat less complicated flowsheets. You have elegantly summarized why adding a fuel reprocessing step to commercial BWR/PWR fleets makes the systemic costs of nuclear power higher and the waste problems worse. EDIT: oh, uranium and thorium halides hydrolyze except at very low pH. And so will a number of fission-product-halides. So maybe not simpler to start with halide salts instead of oxide fuels.
- mjevans 9y ago
  Am I correct in understanding that "waste reduction" (breeder) reactors essentially burn more (most?) the fuel in the core, without that external re-processing?
  - philipkglass 9y ago
    Breeder reactors still require a chemical re-processing step to separate fission products (the lighter atoms produced when heavy actinides like uranium split and release energy) from the fuel. There are two main theoretical advantages for breeder reactors. One advantage is that the total energy extractable from terrestrial mineral resources is much greater with dedicated breeder reactors than with reliance on naturally fissile U-235 in simpler non-breeding reactors. In practice, nuclear power production has not grown fast enough to cause uranium shortages. Until there is a significant uranium shortage, reflected in much higher prices for freshly mined uranium, the higher complexity/cost of a breeder reactor cannot be offset by cost savings on fresh uranium fuel. The other theoretical advantage is a reduction in actinide waste production (plutonium, americium, curium). These heavy waste atoms dominate the radioactivity of spent nuclear fuel in the medium term, thousands to tens of thousands of years from now. But in the short term (years, decades, and centuries), the radioactivity of spent nuclear fuel is dominated by fission products rather than actinides. Breeder reactors do not notable reduce production of fission products compared to conventional reactors. Further, the reprocessing of nuclear fuel to separate the fission products has a tendency to turn potent but small, easily-handled waste problems into less potent but much larger and more unwieldy waste problems. Even though there are some theoretical waste advantages to reprocessing fuel instead of using it once and sealing it up forever, in practice waste reprocessing has always led to higher expenses and more radionuclides actually released into the wider environment.
nickik 9y ago
It does not need to be on-site. In a liquid fuel reactor one could do batch processing. Some startups actually propose to produce reactors with the fuel inside and then recycle the hole reactor. Chemical plants might be a pain but its far easier then doing isotopic separation.

nickik 9y ago

There is some R&D required, but nothing very fundamental. My point is that the technical challenges are pretty minor compared to the rest. Nobody believes what I am proposing is impossible. There might be a discussion if it can be built in 2, 5 to have a full scale reactor that could be commercialized. The real point is that this could have been done 30 years ago, 20 years ago, 10 years ago or at any time. The issue was never a technical one. If the government made a serious push including regulatory changes, action by national labs and copy the NASA approach to commercialization I am certain we could mass produce these things in under 10 years. Once you mass produce reactors the energy production capacity you produce is quite unbelievable. It dwarfs everything else. > current manufacturing My point about manufacturing is that reactors are not more complicated or need higher tolerances then rocket engines for example. SpaceX is producing rocket engines for 2M a pop at a rate of more then a 100 year. Boeing is producing far more complex air planes for 200M a pop at the rate of 1 per week. It is far easier then producing all these wind, solar and battery plants. The needed scales of production are multible order of magnitude different. A mass produced nuclear reactor will produce between 100-300MW, while most wind turbine produce around 3MW while being just as hard to produce and harder to install.