Slashdot Mirror
A Nuclear Startup Will Fold After Failing To Deliver Reactors That Run on Spent Fuel (technologyreview.com)
Transatomic Power, an MIT spinout that drew wide attention and millions in funding, is shutting down almost two years after the firm backtracked on bold claims for its design of a molten-salt reactor. From a report: The company, founded in 2011, plans to announce later today that it's winding down. Transatomic had claimed its technology could generate electricity 75 times more efficiently than conventional light-water reactors, and run on their spent nuclear fuel. But in a white paper published in late 2016, it backed off the latter claim entirely and revised the 75 times figure to "more than twice," a development first reported by MIT Technology Review. Those downgrades forced the company to redesign its system. That delayed plans to develop a demonstration reactor, pushing the company behind rival upstarts like TerraPower and Terrestrial Energy, says Leslie Dewan, the company's cofounder and chief executive. The longer timeline and reduced performance advantage made it harder to raise the necessary additional funding, which was around $15 million. "We weren't able to scale up the company rapidly enough to build a reactor in a reasonable time frame," Dewan says.
Processing spent fuel as you suggest is extremely dirty and generates about 10x the amount of original waste, most of it highly radioactive.
People forget the US tried to reprocess fuel for a while, the location is a radioactive superfund site.
I'd rather that money be spent/wasted on failed attempts at getting us off of fossil fuels when the math indicates that the idea might work, than to continue to burn hydrocarbon fuel for stationary power plants.
I'd rather not nurture your false dichotomy, and suggest instead using the money to improve and implement already-proven technologies like wind and solar.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
But I thought there were already "breeder reactors" that did this. How was this design different from them?
Old fashioned breeder reactors turn U-238 into plutonium. Although plutonium can be used as reactor fuel, it can also be used to make bombs. Furthermore, these reactors use fuel rods, and pressurized containers, and have the same complexity and safety problems as LWRs.
What makes this reactor different is that it doesn't make plutonium, it burns the fuel that in breeds in situ so no extra expensive reprocessing is needed, and it is an inherently safe design: It can't have a "meltdown" since it is already liquid, and it is not pressurized.
That is the theory. In practice, molten salt reactors don't have a very good track record.
Molten salt reactor
Running a little experimental reactor for a couple years tells you nothing about the commercial viability let alone safety.
Pardon my French: horseshit. Obviously we build smaller ones on the way toward building bigger ones -- whether it be nukes or jets or whatever -- it's called engineering.
One of the biggest problems with "molten salt" or liquid sodium reactors is that if the reaction vessel holding this mix of highly radioactive sodium and uranium mixture is every directly exposed to water or oxygen it will explode, burn and fill the atmosphere with a highly radioactive cloud of burning sodium which will then rain down on the surrounding countryside
You seem to be conflating molten salt with molten sodium. They are completely different. Sodium by itself is highly reactive, whereas sodium chloride (though somewhat corrosive) is quite stable. If you hit a LFTR with a bunker-buster bomb, it would indeed spray radioactive molten salt around the countryside. But it would rapidly solidify and fall to the ground, where it would be easy to find with a geiger counter. (Unlike radioactive steam which just floats away...)
you have to do stuff like the current Georgia reactor under construction and spend $20 Billion building a pressure vessel that can survey tidal waves and earthquakes
The reason for that pressure vessel is because water boils at 100C, and nuclear reactors are just getting warmed up around 400C. So a water-cooled reactor needs plumbing that can handle 150 atmospheres of pressure, just so they can run the reactor at the barely efficient temperature of 300C. But since FLiBe doesn't even melt until 360C (and doesn't boil until well over 1400C) you have a very heat-dense material that can both transport your fuel (enabling on-the-fly reprocessing) and cool your reactor over a broad range of heat regimes... and it does all this at ambient pressure. So you don't need that $20B pressure vessel in the first place.
Watch the video I linked above, it will explain all this in greater detail, and save me the trouble of writing it.
XML is like violence. If it doesn't solve your problem, you're not using enough of it. --AC