[m0llusk]

This boldly claims that molten salt reactors are safe, but my understanding is that these systems require carefully engineered coatings to function and that in practice every such system has leaked liquid sodium. The statement may be essentially true kind of like saying elevators are safe, but there is still room for risk analysis and expectation of occasional failures.

[ortusdux]

I've been working on molten salt reactor claddings for a few years now. MSRs are very safe.

First, there are two main groups of molten salt systems. One group uses generic salt to transport heat. This usually provides efficiency gains over water based systems. These are the systems that might leak sodium and it's no big deal.

The 2nd kind of system uses a salt with the fissile material dissolved in it. This is what the article is about. Leaks in these systems are not an option. The molten salt in this case slowly attacks any metal it comes in contact with. The tanks, tubes, etc. are designed with a large safety margin and constantly monitored. The salt chemistry is also monitored for signs of the materials the storage vessel is made of.

[irjustin]

> These are the systems that might leak sodium and it's no big deal.

Sorry can you back this statement up with more than "it's safe, trust me"?

I'm not knowledgeable enough to say if this statement can be accepted at face value. History has shown that's not enough.

But to me, any system that leaks by accident is a disaster waiting to happen on time scales that power plants are expected to operate on.

[jiggawatts]

It's a relative statement.

In this case, leaking molten sodium merely causes a nontoxic metal fire. Definitely in the category of "not fun", but not the end of the world either. Just dump some sand on it and wait for it to cool down.

Leaks of molten salt nuclear fuels are an entirely different category of industrial disaster. A nuclear accident at best, and a large-scale disaster at worst.

It's like the difference between a water truck having an accident and spilling water on the road and a damn bursting and flooding a city.

One makes for a funny picture on Reddit, the other gets the national guard called up to deal with the emergency. Both involve spilled water.

[Gravityloss]

Salt can be extremely chemically inert. Very different from sodium. Also it has very low vapor pressure so different from water.

[jiggawatts]

Not the salts used in reactors. Toxic element are used like fluorine, not to mention the fuel that’s dissolved into it.

[Gravityloss]

I use fluoride tooth paste every day. I wouldn't put sodium in my mouth...

[ortusdux]

I think that you are conflating two different types of systems.

There are systems that use molten salt to transport heat. Several types of power generation systems are more efficient above the boiling point of water. The most famous of these would probably be solar collectors. The efficiency gains are higher at increased operating temperatures, so these systems are usually pushed to their limits. These can leak without causing an international incident.

Molten salt reactors, MSRs, are nuclear reactors that use a liquid fuel in place of a solid one. As far as I know, only a handful have ever been built and reached criticality. These were all done in the 50's and 60's. These systems have secondary and probably tertiary containment vessels in the case of leaks.

These are two very different, but often conflated, systems. Part of the reason for this is that many MSR designs use a secondary molten salt loop as a temperature step-down.

Here is the Wikipedia diagram for a MSR. Note the fuel salt loop, the molten salt coolant loop, and the steam turbine.

https://en.wikipedia.org/wiki/Molten_salt_reactor#/media/Fil...

[ncmncm]

If you find "materials the storage vessel is made of" in the salt, do you shut it down and abandon it?

Alternative would be to drain it into a holding tank and, what, send somebody in to fix it? Who?

[mlyle]

> If you find "materials the storage vessel is made of" in the salt, do you shut it down and abandon it?

You identify the operational error in salt chemistry that has lead to losing some of the storage vessel material, and then you perform engineering analysis on the new resultant factor of safety. If necessary, next fueling cycle you shut down the reactor and perform remediation actions.

[ncmncm]

Assume you need "remediation actions", and that would have to mean sending somebody to climb inside and weld something. Do you shut it down and abandon it, instead? Who goes in?

[jabl]

You seem to be desperately angling for somebody to say that the evil reactor management orders one of the workers to be sacrificed by going into the highly radioactive reactor vessel. Why?

[ncmncm]

Just trying to get answers to simple questions. If there aren't any good answers, that is an answer.

[ortusdux]

Keep in mind that only a few MSRs have reached criticality, and they were experimental units operated in the 50's and 60's. This article and all of this talk is about "Gen IV" MSRs, which, AFIAK, are all currently in the design an testing phase.

Salt chemistry monitoring is a critical part of a MSR operation. The fissile material is slowly exhausted in the reaction and needs to be replaced, while the byproducts need to be removed. The nuclear reaction causes some transmutation of every element it irradiates, which means you get a predictable but diverse set of contaminants that need to also be removed.

The molten salt corrodes everything it comes in contact with, and the radiation also degrades the materials in wild ways. One study I read estimated that directly exposed tungsten would transmute into rhenium at the rate of 1% per year, and that the newly created rhenium would transmute into osmium at the same rate. Basically, it will take an army of people much much smarter than me to plan for all the issues that might come up during the operation of a MSR.

There are many startups backed by billions of dollars trying to find a good solution. One of my favorites simply designs in 8 containment vessels and a crane into the reactor chamber. They plan to just pick up the top of the reactor and move it into a new vessel every 7 years or so.

[tinco]

Literally anyone or anything? Repairing vessels is hardly an unsolved engineering challenge.

[ncmncm]

Repairing vessels with a slime of hard-radioactive salt over it is rather different from what people need to do in PWRs.

[noneeeed]

My physics is very rusty. Why is molton salt more efficient than water for heat transfer? Is it simply that you can get it up to higher temperatures and heat systems are more efficient at high temperatures?

[Gravityloss]

High temperature, low pressure.

You can use water, sodium, lead, fluoride salts. All have their different issues.

[abracadaniel]

Storing and piping molten salt proved too unreliable for the Crescent Dunes solar collector. Swapping out the simple and free heat source with a much more complex nuclear reactor doesn’t seem like a good idea.

https://en.wikipedia.org/wiki/Crescent_Dunes_Solar_Energy_Pr...

[mlyle]

Mostly because of problems with foundations and the power-tower architecture. And note that other molten salt concentrated solar plants have done just fine, and that nuclear power is much more highly regulated and requires a higher degree of engineering analysis than a concentrated solar plant (that's why it's so expensive, in large part!)

[Stevvo]

There are multiple other solar plants operating with molten salts without issue, including the biggest in Morocco.

[jeffbee]

I don't know if they are safe or not but one thing that disturbs me about neo-fission propaganda is how they so nonchalantly dismiss what seem to me to be threats worth discussing. For example, MSR fans will say that they operate at standard pressure, and therefore can't rupture. Also, gaseous fission products are removed continuously. OK but what about the gaseous fission products if the removal system ceases to function? Then what about the pressure? Same question about water. MSRs are safer because they contain no water, which can't flash to steam and explode. OK, but what about water intrusion on the surprisingly wet planet Earth? What if the reactor is flooded? Would a flooded reactor stop the passive safety plug from melting?

I always had the same suspicions about the pebble bed reactor boosters. That was a design that was always sold as being inherently safe and self-moderating, with the huge, huge asterisk that if any oxygen or water ever gets anywhere near the fuel the whole thing will explode and then catch on fire, which seems like a pretty severe drawback. Any time you see a claim that a reactor vessel "does not contain" either water or oxygen, just mentally translate that to "is not designed to contain" or "does not initially contain" before reading further.

[jabl]

> in practice every such system has leaked liquid sodium

Just to clarify, reactors cooled by liquid sodium are not molten salt reactors but rather belong to the class of reactors often called "metal-cooled" (sodium is a metal, not a salt, see).