Why isn't tungsten used in nuclear reactors?

[Mackenzie Cobb]

Me again, with another potentially ignorant nuclear science question:

Why isn't tungsten used to prevent meltdown in nuclear reactors?

If tungsten has a higher melting point of tungsten is almost 6200 degrees Fahrenheit, and nuclear meltdown happens when the uranium fuel is some 5200 degrees, why not line the bottom of reactors and containment vessels with tungsten in order to prevent melt-through and subsequent contamination of groundwater underneath the facility?

(Unless of course the meltdown can get hotter than 5200 degrees, but I couldn't find the actual highest temperature of a nuclear meltdown; just the melting point of the uranium fuel. Second question, what is the highest temperature nuclear materials used in reactors can reach?)

 

[anorlunda]

There are many properties of materials that make them suitable or not suitable for use in reactors. Strength, corrosion resistance, soluability ductility/brittleness, and more. High energy radiation also alters material properties, especially embrittlement. Trace amounts of natural contaminants in materials can also disqualify them.

Rest assured that tungsten and every other potentially interesting material was considered. If it was rejected, then other materials were better.

 

[Rive]

...nuclear meltdown happens when the uranium fuel is some 5200 degrees...

That is still not the maximal temperature of damaged fuel. If it stays contained (without cooling to balance heat production) in narrow space the temperature will keep climbing indefinitely.
The actual approach is to dilute and scatter the molten fuel (google up 'core catcher').

 

[Mackenzie Cobb]

That is still not the maximal temperature of damaged fuel. If it stays contained (without cooling to balance heat production) in narrow space the temperature will keep climbing indefinitely.
The actual approach is to dilute and scatter the molten fuel (google up 'core catcher').

That is very interesting. Thank you! I have not seen that in any of the reading I've done so far.

 

[berkeman]

(google up 'core catcher')

Interesting, thanks @Rive -- I hadn't heard of that either.

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

 

[Mackenzie Cobb]

There are many properties of materials that make them suitable or not suitable for use in reactors. Strength, corrosion resistance, soluability ductility/brittleness, and more. High energy radiation also alters material properties, especially embrittlement. Trace amounts of natural contaminants in materials can also disqualify them.

Rest assured that tungsten and every other potentially interesting material was considered. If it was rejected, then other materials were better.

Thank you for your answer! However, I was looking for more of a specific reason why it has obviously been rejected as a material. I cannot possibly claim to know more than trained nuclear physicists, and I am aware that there must be a reason why it is not used, that has been well-researched and tested. I'm more curious about the specific reasoning as to why. I cannot find a clear answer to that question on my vague google searches, so that's why I've come here. :) Thank you!

 

[Mackenzie Cobb]

Interesting, thanks @Rive -- I hadn't heard of that either.

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

View attachment 231670

That's fascinating! Now my question is whether or not this has been tested and proven to work. How would one test a system such as this without initiating a meltdown? (Which honestly seems counterproductive, and the risks involved seem to outweigh the reward...) Is this system mandatory in 'new' nuclear reactors?

Which leads to another question, are there any instances of scientists initiating meltdowns on purpose, in order to study the process? How would such a procedure be performed? How could it be performed safely?

 

[OmCheeto]

That's fascinating! Now my question is whether or not this has been tested and proven to work. How would one test a system such as this without initiating a meltdown? (Which honestly seems counterproductive, and the risks involved seem to outweigh the reward...) Is this system mandatory in 'new' nuclear reactors?

Which leads to another question, are there any instances of scientists initiating meltdowns on purpose, in order to study the process? How would such a procedure be performed? How could it be performed safely?

Apparently, it has been done, and it doesn't look like it was done safely!

https://en.wikipedia.org/wiki/BORAX_experiments#BORAX-I_destructive_test_and_cleanup

I did not know that.

... I was looking for more of a specific reason why it has obviously been rejected as a material.

My guess is cost.
Nuclear reactors are not designed to fail.
They are VERY nasty things, when they do.

BTW, I can really appreciate your inability to find answers to these questions. Although I'm somewhat versed in nuclear technology, I'm having a very difficult time getting answers, via google, to answer your questions.

One thing I haven't seen mentioned is something called "decay heat". Though, Rive kind of inferred it:

...the temperature will keep climbing indefinitely

Same thing.

From some back of napkins maths I just did, a small 1000 Mw nuclear reactor core, with a solid volume of about 1/2 m³ would still be generating 30 million watts, an hour after it had been shut down.

Image I interpolated the volume from:

1/2 m³ is about the size of a tiny refrigerator.
30 million watts, is, a lot.
I'll let you do the maths, if the core had the thermal properties of say, steel, as to what the temperature would be, after an hour.
Too much maths for me right now.
Just had lunch, and it's time for my nap.

ps. I considered the original question to have been mostly answered in post #2. ie. It's VERY complicated.

 

[Rive]

My guess is cost.

I think you are right with this. As it seems, tungsten can be considered a proper 'nuclear material' since it is in use for long (google up 'demon core'). I could not find any negative effects, like -for example - for Cobalt, which is practically banned from reactors, despite being a well known useful component of many alloys. So, I too think it'll be about the high cost for not enough benefit.

If there is no cooling, then a meltdown cannot be contained in the RPV, regardless the materials used. If there is cooling, then any decent steel can can do the trick. So let's go with the steel can, and add more concrete to the primary containment... I guess.

 

[Denis Khliustin]

Why isn't tungsten used to prevent meltdown in nuclear reactors?

Good idea, it's possible to invent the system of tungsten tubes under reactor, which will separate the melted core into several fluxes, subcritical each of them, and direct them into different cooled places under reactor.
World production of tungsten, about 40000 tons/year, is not very big but probably will not make a big bareer for such project.

Another potentially possible way of using tungsten is using W184 isotope for shells of fuel rods in the core. It has capture cross section about 0.5 barns compared to ~20 barns of natural tungsten isotopes mixture.

 

[OmCheeto]

I think you are right with this.

Yay!

As it seems, tungsten can be considered a proper 'nuclear material' since it is in use for long (google up 'demon core').

Wow! I read the wiki entry on it. Hilarious and sad, at the same time.

I could not find any negative effects, like -for example - for Cobalt, which is practically banned from reactors, despite being a well known useful component of many alloys. So, I too think it'll be about the high cost for not enough benefit.

If there is no cooling, then a meltdown cannot be contained in the RPV, regardless the materials used.

What about silver? It has the best thermal conductivity I could find. You just need a whole lot more money, and assume that there is going to be an accident.

If there is cooling, then any decent steel can can do the trick. So let's go with the steel can, and add more concrete to the primary containment... I guess.

After my nap yesterday, watching 6 hours of "Rick and Morty" reruns, and a full nights sleep, I researched how many reactor vessels were breached. I only came up with Chernobyl and 3 of the Fukushima Daiichi vessels.
I was surprised to see that the Three Mile Island vessel was never breached, even though 19,000 kg of the core melted! [ref]
Good lord!

One strange thing I found out about the Three Mile Island "Corium" was that it was 70% Uranium by weight. [same ref as above]
I guess I've been out of the industry too long, as that sounded like quite a healthy concentration to not be critical.

I was going to research that some more at our Japan Earthquake: nuclear plants thread to find what I was not understanding, but it's over 7 years old, and has nearly 16,000 comments.

Complicated, is an understatement, IMHO.

 

[keithdow]

If you want to consider why or why not something was used, the first consideration often is money.

 

[gmax137]

If there is no cooling, then a meltdown cannot be contained in the RPV, regardless the materials used. If there is cooling, then any decent steel can can do the trick. So let's go with the steel can, and add more concrete to the primary containment

I think this is it. The vessel is designed to maintain the coolant pressure, and keep the fuel submerged in the coolant. While it is much nicer post-accident cleanup if the fuel remains in the vessel (as TMI) it wasn't a design criterion for the vessels.

Regarding ex-vessel "core catcher" this was considered by the US regulators in the 1960s. David Okrent's book goes into this in some detail. See NRC ADAMS website, search for accession number ML090630275, "On The History of the Evolution of Light Water Reactor Safety in the United States".

https://www.nrc.gov/docs/ML0906/ML090630275.html

By the way, this document is fascinating if you have an interest in the history of nuclear regulations in the US. But be warned it is over 1000 pages.

 

[Mark Harder]

I was thinking about using W for small rocket nozzles. In researching its properties I found that W is very brittle and therefore difficult to work unless it is very pure. In order to spin, bend & forge tungsten it needs to be something like 99.99% pure. In contrast to most industrial metals, that's a pretty unforgiving specification. Making a vessel out of many, many tons of a 99.99% (again, that's just an order of magnitude guess) elemental metal has got to be a very expensive proposition. Also, W has a density of 19.35 g/cm^3, slightly greater than gold, so unless a thinner W PV is sufficiently strong compared with steel, etc., you would also need stronger materials for all the structures that support the PV. There are alloys of W that melt at higher temperatures and resist the tendency to creep under tensile force at high temperatures. They are used to make impeller blades in military jet engines. The alloy is a mixture of tungsten and rhenium. The latter is one of the rarest natural materials in the periodic table, so unless you have a DOD contract, you're not going to be using rhenium to build big pressure vessels.

 

[Astronuc]

Me again, with another potentially ignorant nuclear science question:

Why isn't tungsten used to prevent meltdown in nuclear reactors?

If tungsten has a higher melting point of tungsten is almost 6200 degrees Fahrenheit, and nuclear meltdown happens when the uranium fuel is some 5200 degrees, why not line the bottom of reactors and containment vessels with tungsten in order to prevent melt-through and subsequent contamination of groundwater underneath the facility?

(Unless of course the meltdown can get hotter than 5200 degrees, but I couldn't find the actual highest temperature of a nuclear meltdown; just the melting point of the uranium fuel. Second question, what is the highest temperature nuclear materials used in reactors can reach?)

Tungsten isn't used for design reasons and cost. While tungsten has a high melting point, it is subject to corrosion and embrittlement.

PWR and BWR pressure vessels have numerous penetrations in the bottom part of the pressure vessel. PWR have instrumentation systems, e.g., thermocouple tubes and flux thimbles that allow thermocouples and neutron detectors to travel into the core. Some systems are designed to insert instruments from the top, which reduces or eliminates need for penetrations. BWRs on the other hand have control rod and detectors inserted from the bottom of the core. The penetration tubes and lower core structures are composed of stainless steel, usually a type of 304 or 316, but some could be 347 or 348. The pressure vessels are type SA 508 Cl 2 or SA 533 B, and the inner surface is clad with an austenitic stainless steel. Adding a layer of W-alloy (e.g., alloyed with Re or some other elements) to the pressure vessel inner surface would be rather impractical. One would have to protect the tungsten alloy from corrosion by the coolant and interaction with the other structural materials, and deal with differences in thermal expansion as the reactor vessel expands to operating temperature and contracts to cold shutdown conditions.

The core support structures are cast stainless steel. The upper and lower nozzles of PWR fuel assemblies and upper and lower tie plates of BWR fuel assemblies are made of wrought or cast stainless steel, usually a 300 series, e.g., 304/316/347 or derivatives. Casting equivalent is often CF3/CF3M.

The actual maximum temperature of corium is complicated and can only be estimated on a case-by-case basis. The Wikipedia article makes a good estimate, but it's up to the melting point of UO₂. The maximum temperature will depend on the decay heat, composition of the melt, porosity and how much coolant is available. Furthermore, since the melting point of stainless steel is ~ 1375-1400 C and Zr-alloys ~ 1850 C, that would pretty much limit the temperature of the melt, although it could go higher to melting point of metal oxides if steel and Zr-alloys react (oxidize/corrode) in high temperature water.

The melt temperature does not increase indefinitely, but is limited by what supports and/or interacts with the melt. One can find many simulations and experiments by searching on "Corium simulations" or "corium experiments", or "reactor severe accident analysis" or "experiments".

The original design of LWRs calls for emergency core cooling systems. Modern reactor system designs have more passive features.

In the case of Fukushima, we're still trying to learn what happened there. Clearly there was an chemical oxidation reaction with whatever water was present, from which produced the hydrogen that exploded. Clearly the cores had insufficient water in the system, or the core was so hot in stagnant steam/hydrogen that the metals simply reacted and disintegrated (some believe melting).

Getting back to a layer of tungsten, with a mass of molten core sitting on top, the heat would transfer to the softer steel under the W-alloy layer, or melt the stainless steel penetrations, and possibly the W-alloy would react with the steels. So some layer would have to be present, e.g., ceramic. So one is still left with differential thermal expansion and other issues.

BTW, tungsten is used in some control elements. See US patent 8537962 B1.

W-184 would make an interesting reflector.

 

[Mackenzie Cobb]

Tungsten isn't used for design reasons and cost. While tungsten has a high melting point, it is subject to corrosion and embrittlement.

PWR and BWR pressure vessels have numerous penetrations in the bottom part of the pressure vessel. PWR have instrumentation systems, e.g., thermocouple tubes and flux thimbles that allow thermocouples and neutron detectors to travel into the core. Some systems are designed to insert instruments from the top, which reduces or eliminates need for penetrations. BWRs on the other hand have control rod and detectors inserted from the bottom of the core. The penetration tubes and lower core structures are composed of stainless steel, usually a type of 304 or 316, but some could be 347 or 348. The pressure vessels are type SA 508 Cl 2 or SA 533 B, and the inner surface is clad with an austenitic stainless steel. Adding a layer of W-alloy (e.g., alloyed with Re or some other elements) to the pressure vessel inner surface would be rather impractical. One would have to protect the tungsten alloy from corrosion by the coolant and interaction with the other structural materials, and deal with differences in thermal expansion as the reactor vessel expands to operating temperature and contracts to cold shutdown conditions.

The core support structures are cast stainless steel. The upper and lower nozzles of PWR fuel assemblies and upper and lower tie plates of BWR fuel assemblies are made of wrought or cast stainless steel, usually a 300 series, e.g., 304/316/347 or derivatives. Casting equivalent is often CF3/CF3M.

The actual maximum temperature of corium is complicated and can only be estimated on a case-by-case basis. The Wikipedia article makes a good estimate, but it's up to the melting point of UO₂. The maximum temperature will depend on the decay heat, composition of the melt, porosity and how much coolant is available. Furthermore, since the melting point of stainless steel is ~ 1375-1400 C and Zr-alloys ~ 1850 C, that would pretty much limit the temperature of the melt, although it could go higher to melting point of metal oxides if steel and Zr-alloys react (oxidize/corrode) in high temperature water.

The melt temperature does not increase indefinitely, but is limited by what supports and/or interacts with the melt. One can find many simulations and experiments by searching on "Corium simulations" or "corium experiments", or "reactor severe accident analysis" or "experiments".

The original design of LWRs calls for emergency core cooling systems. Modern reactor system designs have more passive features.

In the case of Fukushima, we're still trying to learn what happened there. Clearly there was an chemical oxidation reaction with whatever water was present, from which produced the hydrogen that exploded. Clearly the cores had insufficient water in the system, or the core was so hot in stagnant steam/hydrogen that the metals simply reacted and disintegrated (some believe melting).

Getting back to a layer of tungsten, with a mass of molten core sitting on top, the heat would transfer to the softer steel under the W-alloy layer, or melt the stainless steel penetrations, and possibly the W-alloy would react with the steels. So some layer would have to be present, e.g., ceramic. So one is still left with differential thermal expansion and other issues.

BTW, tungsten is used in some control elements. See US patent 8537962 B1.

W-184 would make an interesting reflector.

Thank you so much! This is very informative. Also thank you for the terms to search. I think half of the research problems that I am having are because I'm not sure how to adequately word my searches to get the best answers. Thank you!

 

[trurle]

Thank you for your answer! However, I was looking for more of a specific reason why it has obviously been rejected as a material. I cannot possibly claim to know more than trained nuclear physicists, and I am aware that there must be a reason why it is not used, that has been well-researched and tested. I'm more curious about the specific reasoning as to why. I cannot find a clear answer to that question on my vague google searches, so that's why I've come here. :) Thank you!

I think the most critical problem with tungsten in reactor would be corrosion resistance. Tungsten form oxide which is powdery and do not offer corrosion resistance above 200C temperature.

 

[Unobtanium]

Lots of good ideas here.
Using W (or some Rh alloy suggested here) as a corium separator could be a possibility, but perhaps also Ultra High Temp Ceramics?
If the reactor vessel is breached then the "core catcher" should have both be able to reduce critically and dissipate heat, by separating the molten material into streams, where I would guess both a moderating material and some neutron absorbing material is used. I've heard of ideas of having some neutron absorbing core killer that could drop in and stop any reaction in case of a severe failure.

Retrofitting core catchers into old installations would be expensive, or impossible, but could reduce public perception of LWR safety.

The public fear of nuclear power got a big boost from Japan. The glaring example of Fukushima comes to mind. Placement of several backup systems below grade, their wiring location, lack of hook-ups for tertiary power, site selection, lack of suitable tsunami walls, lack of water barriers around reactor building, emergency operating procedures, and likely a host of other design and planning misses speaks to the real issues at hand. It would not have taken much effort to have prevented this mishap. The public's fear nuclear power is in part based on the belief that a nuclear explosion could happen, which of course is not possible with LEU, hence some educational effort would be good. The failure of Fukushima (and the two others) was not really a nuclear power failure, but that of planning and implementation. Cars are generally safe, until they are not used properly. Don't blame the car. Fear is often irrational and has a long memory.

“The hydrology of the Fukushima site is very complicated and thus the exact water flow is hard to predict,” ... “especially during heavy rains.”

If the hydrology was unknown under the site, one wonder why it was ever used. A cultural climate of being "yes-men" would likely also be at fault, where saying "no" is impolite and avoided.

"Saying ‘no’ in Japan is a tricky subject. The Japanese will rarely give a direct no to an answer, preferring instead to give an indirect answer that conveys the message of no.
Giving someone a direct no is too disruptive in a society that values keeping the harmony at all cost. As a result the Japanese will usually choose their words carefully, especially in business related situations."

The concept of "harmony" in Japan would be shunned in the western world where a healthy discourse is expected and needed. This would mean any of their installations should be heavily scrutinized.

As their 30m "ice-wall" does not seem to work, maybe its time to dig a trench to isolate the site?

The (co)inventor of the LWR, Alvin M. Weinberg did not favor the scaling of a small naval vessel reactor into the sizes we now have, but also invented a Fluid Fuel Reactor, where there a "melt down" scenario does not exist. These reactor types could be slow spectrum (low cost) Thorium breeder or fast spectrum nuclear waste eating machines with negative temp coefficients and very low pressures. All of, if not most of the concerns of nuclear safety is related to the use of water, disappears with Fluid Fuel Reactor, and was successfully demonstrated (MSRE) back in the 1970's.
More work is needed to develop these reactor types for longevity of materials used, which has been started, but perhaps needs a good boost to get results.
Regulations concerning nuclear reactors are focused on the problems of using H2O, at 2000 psi, and regulations are not applicable to most if not all of the concerns with MSR's. A great regulatory overhaul would be in order to make a transition to MSR's.

 

[Mark Harder]

Your description of the cultural factors at work at Fukushima reminds me of the recent docudrama, "Chernobyl". According to my interpretation, the Soviet system was responsible for the resistance to accepting the truth about the accident, right down to fundamentals like, Did the reactor explode? (Angry boss sends a man to look into the pit and he returns with radiation sunburned face to report.) Comrade director, the reactor has exploded, it's mostly gone. NO! The reactor couldn't have exploded! The culture there seems to have been that what your superiors tell you is the truth is the truth, no matter what the evidences of your senses are telling you. Instead of an excess of politeness and face-saving, it bred a system of bullies who forced their version of the truth on their subordinates. Unfortunately, nuclear reactors do not surrender to bullies.