Coolant Choices for Liquid-Metal Fast Reactors

[LostCosmonaut]

Metal cooled reactors have several advantages over pressurized water reactors. For one, their power density is greater, additionally, the coolant is unpressurized, improving safety.

 

However, there are some downsides. The Soviets' Project 705 class submarines were powered by liquid metal reactors utilizing a lead-bismuth alloy as coolant. This alloy had a freezing temperature of roughly 400K. As a result, the reactors had to be run constantly, even while the submarines were in port (there were facilities to provide superheated steam to the reactors while the subs were docked, but they broke down and were never repaired). This reduces the lifetime of the reactor. Another coolant choice which has been used operationally is NaK (Sodium-Potassium). This alloy is liquid at room temperature, but reacts violently with water or air. I'm not an expert, but this seems like a bad thing.

 

 

It seems to me that if appropriate coolants could be found, it seems that liquid metal fast reactors could see more widespread acceptance. To my untrained eye, gallium looks like a good choice. Its melting point is relatively close to room temperature (~303K), and the boiling point is quite high (over 2600K). Also, gallium is less reactive than sodium or other alkali metals. It appears that there has been some research on this topic: http://www.sciencedirect.com/science/article/pii/S0149197000000640(unfortunately, the article is behind a paywall), and it looks quite promising.

 

Anybody have any opinions on this, or suggestions for alternative coolants?

[Collimatrix]

Depending on the geometry of how the coolant interacts with the core, the neutron absorbtion cross section and the atomic mass of the atoms in whatever alloy you end up using may be a concern.  You want low neutron absorbtion cross section, else the coolant is going to start sucking up neutrons from the core and become radioactive.  You also want highish atomic mass, or else the coolant could start acting like a moderator, which is bad because it's a fast reactor.

 

Ideally the metal will have a high boiling point so that the hot side of your heat engine cycle can be at as high a temperature as possible.  That is one of the attractions of liquid salt reactors.

[Collimatrix]

Some more thoughts:

 

 

The way the coolant interacts with neutrons is actually pretty important.  For the coolant to remove heat efficiently from the core, it needs to contact the maximum possible surface area in the core.  This means that it is flowing in between the fuel rods (or whatever fuel elements you're using), so it's interacting with the neutrons pretty extensively.

If the coolant absorbs significant amounts of neutrons, the neutron flux in the reactor core will increase if there is a sudden excursion in core temperature.  What will happen is the excursion will heat up the coolant faster than the radiator can dump heat, so the coolant will heat up.  Depending on the coefficient of thermal expansion and phase/pressure, the coolant will either expand or boil.  This sudden drop in coolant density will mean that it's sucking up fewer neutrons, and the reactivity will go up even more.  Obviously, this sort of positive feedback loop is scary; it's one of the flaws RBMK reactors have.

 

Also, if the coolant is sucking up neutrons, it's probably becoming radioactive.

 

According to Encyclopedia Britannica, gallium has a fairly high neutron absorption cross section.

[Collimatrix]

I found a source discussing this.

 

It sounds like helium is a very good choice.  For reasons that I'm still unclear on, helium has almost no moderating effect (with an atomic mass of 4, go figure), and of course it won't corrode anything else in the reactor.  It suffers from no neutron activation whatsoever either, and it behaves itself well in turbines.

 

On the other hand, it loves to leak out and it's expensive as hell.

 

They briefly mention the Soviet fling with N₂O₄ cooled fast neutron reactors, which sound delightfully nuts.  I can imagine how that grant proposal went:

 

"Comrade, you know that evil-looking reddish-brown gas that leaks out the sides of our Satan missiles?"

"Yes?"

"I want to design a nuclear reactor that's cooled with that stuff."

[Vanagandr]

So, I played music at the Iowa State Fair recently, the guy who I carpooled with is an electrician of some sort at the local nuclear power plant. Something interesting I learned from him is that reactors that use sodium for heat exchange actually use induction to cycle the sodium (or at least the one he used to work at did).

[Collimatrix]

Both mechanical and electromagnetic pumps are used, here's a presentation on them.