Hacker News Reader

Comment by quasse | Hacker News Reader

In case anyone else was wondering how having the molten fuel pour out of the reactor is safe - it is designed to pour into a container with criticality safe geometry to stop fission:

"in an emergency situation [the fuel] can be quickly drained out of the reactor into a passively cooled dump tank. MSRs designs have a freeze plug at the bottom of the core—a plug of salt, cooled by a fan to keep it at a temperature below the freezing point of the salt. If temperature rises beyond a critical point, the plug melts, and the liquid fuel in the core is immediately evacuated, pouring into a sub-critical geometry in a catch basin. This formidable safety tactic is only possible if the fuel is a liquid." [1]

[1] https://www.sciencedirect.com/science/article/pii/S168785071...

View on HN

> This formidable safety tactic is only possible if the fuel is a liquid." [1]

Isn't something similar possible with pebble reactors, at least in principle?

Possible, and implemented in recent reactors - for the "corium". Though in that case it's for the extreme situations, since after that the reactor is completely done for.

I would love to see what a sub-critical geometry looks like. I also wonder what kind of math and engineering they did to come up with the shape they ended up with.

http://thorconpower.com/wp-content/uploads/2019/03/ColdWallL...

This one is just a bunch of separate tanks spaced as far away from each other as possible.

I've also seen a "Bed of nails" design where the liquid salt would flow over a bunch of moderating spikes, seep down between them and freeze.

In some contexts, a subcritical geometry is simply a plastic jug that's too small to hold a critical mass. Obviously they're not using plastic jugs for this, but the principle of the thing is pretty straight forward. They could create a wide 'dish' under the reactor which, if the liquid fuel were drained into, would spread the fuel out into a shallow puddle using the self-leveling property of fluids.

> I also wonder what kind of math and engineering they did to come up with the shape they ended up with.

I assume this is why some Finite Element Analysis packages come with warnings along the lines of "You must not use this to help Kim Jong-Un do you know what"

Feynman wrote of this when he was working at Los Alamos, i.e. the labourers weren't informed as to what they were really working with, so they could come very close to criticality accidents.

A pyramid may work with a wide flat basin at edge, especially if core material is moderating itself (vs just separating).

a pancake ?

So instead of melting down by accident it melts down deliberately, controlled by a fan?

Somehow I'm not as reassured as I might be.

Well, no. A meltdown is when the stuff escapes containment and then continues undergoing fission, heating up and melting out of containment and into groundwater/the environment.

The drain tanks are shaped specifically so if the (liquid) fuel flows in, it will stop reacting. And presumably the whole apparatus (drain tanks included) is inside of a biological shield.

So a "meltdown" here just means a big hunk of solid nuclear salt inside the bio-shield. I think the idea being that you could just bury the whole thing in case of failure.

IIRC draining the fuel salt into the drain tanks was standard operating procedure for one of the early MSR experimental reactors.

Next morning when they came in to continue experiments, started the (electrical?) heaters to liquify the salt and pump it back into the reactor core.

Wouldn't surprise me if the modern designs are similar.

Also, there is no pressure build up, which is what caused the big explosion in Fukushima. All the coolant starts boiling, turning into gas, and well the rest was in the news

Just for clarification, the big explosion at Fukushima was a hydrogen explosion. At high temperatures, fuel cladding breaks water into hydrogen and oxygen. Hydrogen collected at the highest point in the structure and later exploded.

This would not be melting down. Melting down is melting out of the containment device, which this is not doing. True it is melting INTO a secondary containment device but unless it continues to melt through that it would not be a 'melt down'

Reply on news.ycombinator.com