4 ms·

Have we ever thought about building a nuclear reactor underground before? Seems a little insane, but I'm trying to think through the "why" part. It would be aw

Have we ever thought about building a nuclear reactor underground before? Seems a little insane, but I'm trying to think through the "why" part. It would be awfully convenient if the way you "store" the nuclear waste is to just leave it in the reactor underground and then build another one! Immediate problem that comes to mind: Groundwater contamination. Also it being a pain to build (probably?). Anybody aware of if anybody has explored this before?

jmyeet 2y ago
So there are people working on that [1]. It's basically geothermal where the heat is generated by a nuclear reactor. If anything goes wrong, you just bury it. If it's geologically stable, not near a water table and a mile deep, then there's a certain logic to that. [1]: https://eepower.com/news/underground-energy-how-this-mile-deep-nuclear-reactor-works/# https://eepower.com/news/underground-energy-how-this-mile-de...
- Eric_WVGG 2y ago
  Sounds neat, but isn't like the number 1 most important thing at a reactor keeping it cooled? You can do that underground in Klei’s Oxygen Not Included — synthesize some supercoolant, run it through an aquatuner, eat up the waste heat with steam turbines, maybe throw a wheezewort someplace for good measure… but I might be missing some steps.
  - jmyeet 2y ago
    Water is generally used to cool nuclear reactors. And you're heating water so it's kind of a 2-for-1. You could even have a separate cooling loop with a heat exchange with the water you're boiling to avoid any radioactive containment, I guess? I don't know what the failure modes for this look like. These are small reactors. Is there enough fuel for it go critical? I would assume not. What happens if there's a containment breach? Will it start releasing radioactive steam? The whole thing is still at a conceptual stage.
- credit_guy 2y ago
  That idea was totally crazy. In order to save on pouring lots of concrete for a containment dome, you dig one mile in bedrock. But one thing that everyone missed when the story came out was what was so special about 1 mile. Why not half a mile or 1.5 miles? The neat thing (which they did not bother to publicize) was that the pressure in a pressurized water reactor is exactly the pressure at the bottom of a column of water one mile high. So you do not need pumps for your reactors. You just let your water fall to the reactor, the reactor heats it and sends it back. It's quite elegant. Of course, the remaining engineering details are left to the reader. For example, what happens if some valve breaks 1 mile deep underground?
  - kragen 2y ago
    dimensionally it's hugely advantageous. a 1600-meter-thick containment dome (to use modern units rather than medieval ones) would be 17 billion cubic meters of concrete, which would cost about 700 billion dollars, at the price of 40 dollars per cubic meter i happened to see in the newspaper today. but a 300-mm-diameter hole 1600 meters into the bedrock only requires breaking up 110 cubic meters of bedrock. all things being equal, that's about 150 million times easier (in practice people skimp on the containment instead of trying to build trillion-dollar containment domes) as for water pressure, you can design pwrs to work over a significant range of water pressures
SoftTalker 2y ago
Reactors have to be refueled periodically. This might be much more difficult if the reactor is deep underground, or at least it might complicate the design. Unless the design and construction can be simple enough that it's cheaper to make a new one than refuel an existing one.
orwin 2y ago
The plan for old HI nuclear material in USSR in a decommissioned plant was to "settle in place" for a hundred year (after that its mostly MI/LI that you can glass easily). If you have enough space, the "return to green" that is the goal in France (10-30% of the lifetime costs depending on who you listen to) is too expensive for little benefits.

jmyeet 2y ago

So there are people working on that [1]. It's basically geothermal where the heat is generated by a nuclear reactor. If anything goes wrong, you just bury it. If it's geologically stable, not near a water table and a mile deep, then there's a certain logic to that. [1]: https://eepower.com/news/underground-energy-how-this-mile-deep-nuclear-reactor-works/# https://eepower.com/news/underground-energy-how-this-mile-de...

Eric_WVGG 2y ago
Sounds neat, but isn't like the number 1 most important thing at a reactor keeping it cooled? You can do that underground in Klei’s Oxygen Not Included — synthesize some supercoolant, run it through an aquatuner, eat up the waste heat with steam turbines, maybe throw a wheezewort someplace for good measure… but I might be missing some steps.
- jmyeet 2y ago
  Water is generally used to cool nuclear reactors. And you're heating water so it's kind of a 2-for-1. You could even have a separate cooling loop with a heat exchange with the water you're boiling to avoid any radioactive containment, I guess? I don't know what the failure modes for this look like. These are small reactors. Is there enough fuel for it go critical? I would assume not. What happens if there's a containment breach? Will it start releasing radioactive steam? The whole thing is still at a conceptual stage.
credit_guy 2y ago
That idea was totally crazy. In order to save on pouring lots of concrete for a containment dome, you dig one mile in bedrock. But one thing that everyone missed when the story came out was what was so special about 1 mile. Why not half a mile or 1.5 miles? The neat thing (which they did not bother to publicize) was that the pressure in a pressurized water reactor is exactly the pressure at the bottom of a column of water one mile high. So you do not need pumps for your reactors. You just let your water fall to the reactor, the reactor heats it and sends it back. It's quite elegant. Of course, the remaining engineering details are left to the reader. For example, what happens if some valve breaks 1 mile deep underground?
- kragen 2y ago
  dimensionally it's hugely advantageous. a 1600-meter-thick containment dome (to use modern units rather than medieval ones) would be 17 billion cubic meters of concrete, which would cost about 700 billion dollars, at the price of 40 dollars per cubic meter i happened to see in the newspaper today. but a 300-mm-diameter hole 1600 meters into the bedrock only requires breaking up 110 cubic meters of bedrock. all things being equal, that's about 150 million times easier (in practice people skimp on the containment instead of trying to build trillion-dollar containment domes) as for water pressure, you can design pwrs to work over a significant range of water pressures

SoftTalker 2y ago

Reactors have to be refueled periodically. This might be much more difficult if the reactor is deep underground, or at least it might complicate the design. Unless the design and construction can be simple enough that it's cheaper to make a new one than refuel an existing one.

orwin 2y ago

The plan for old HI nuclear material in USSR in a decommissioned plant was to "settle in place" for a hundred year (after that its mostly MI/LI that you can glass easily). If you have enough space, the "return to green" that is the goal in France (10-30% of the lifetime costs depending on who you listen to) is too expensive for little benefits.