Hacker News new | past | comments | ask | show | jobs | submit login
What Is Called Nuclear Waste Is Mostly Fuel for Molten Salt and Fast Reactors (nextbigfuture.com)
79 points by rbanffy 32 days ago | hide | past | web | favorite | 49 comments



This is true only if you ignore costs. It falls apart as soon as you include market pricing or any other reasonable resource allocation mechanism.

It's like pretending that the liabilities represented by a huge coal ash pond are actually an asset, because if all the aluminum and silicon in the ash were purified and sold it would be worth a lot at present prices. The problem is that the cost of turning the waste into a salable product is higher than the market price. So it is still in fact a liability.

Under present and foreseeable economic and technical conditions, the asset-value of spent nuclear fuel does not exceed its liabilities.


But the "spent" fuel is already a liability that no one has figured out how to get rid of.


Trying to permanently get rid of spent nuclear fuel under present and foreseeable conditions costs more than monitored dry cask storage for the indefinite future. A determination to permanently get rid of it despite the lower cost of once-through fuel cycles is a determination to take a modest liability and make it worse.


...unless you count the risks of something bad happening.

That’s always the unspoken actual debate when it comes to nuclear energy: what’s the probability of the worst case, and how does it affect the amortized cost?

We’re terrible at doing this calculation in other areas (e.g. terrorism vs. the cost of security theater), so it’s not especially informative to talk about the status quo. Clearly, we believe it’s cheaper to just store the stuff, or we wouldn’t be doing it. With nuclear energy, the exceptions matter a lot.


Both present once-through fuel cycles and "burner" reactor concepts produce fission products in proportion to energy generated. The short and medium lived fission products are the most intensely radioactive and dangerous byproduct of nuclear fission for the first few centuries of waste storage. The advantage of a burner reactor from a waste perspective is that it can consume longer lived actinide waste products like neptunium, plutonium, and americium. A burner reactor converts these actinide wastes into more intensely radioactive but less-long-lived fission products. (Additionally, some much longer lived and less intensely radioactive fission products like technetium 99 are also formed in burners and conventional reactors.)

This is to say that even if you place a high premium on avoiding environmental releases of radioactive materials, there is no appreciable benefit from burner reactors on a human time scale. The overall radiotoxicity inventory that needs ongoing management is nearly the same for both burners and once-through reactors for 200+ years after the fuel is done with. Further, fuel reprocessing facilities are themselves more likely to leak fission products into the environment than dry cask storage is.


> Further, fuel reprocessing facilities are themselves more likely to leak fission products into the environment than dry cask storage is.

But these fuel reprocessing already exist and the waste are currently already reprocessed to make MOX fuel, at least in some countries (France, UK, India, Russia and Japan).


Yes, and those plants release more becquerels of radioactive material into the environment when they reprocess fuel than would be released if the fuel elements were left intact in dry storage. I'm not saying that reprocessing plants release a dangerous amount of radioactive material. I'm saying that reprocessing makes things worse if your top priority is to prevent releases of radioactive materials into the environment. It also makes things worse if your top priority is to make nuclear-generated electricity affordable.


The first storage is actually opening in Finland. Its not that difficult to find a place technically. Its mostly a political problem.


I am often tempted to frame it as, "mostly an adulting problem".


"This is true only if you ignore costs. "

I think it would be better to say: "This is not true in many scenarios, depending on how costs are derived"

Liabilities are very hard to determine.

We are not yet to the point of operationalizing much of this stuff very well.

Given the vast amount of latent energy in such things, it's probably worthwhile looking at how we can effectively streamline and enable safety, fully accepting the fact may be 'it's impossible'.


Right, the liabilities are hard to determine and those liabilities may increase substantially if simply buried.

This is because most types of spent nuclear fuel slowly become "weapons-grade" as the short half-life isotopes decay.


The grade of plutonium is determined by the isotopic composition of the plutonium itself, not the short-lived fission products found around it in spent fuel. Weapons grade plutonium contains a high fraction of Pu-239 and a low fraction of Pu-240. The plutonium in spent fuel from commercial power reactors contains much more Pu-240 than weapons grade plutonium does.

Since Pu-240 has a shorter half life than Pu-239, it is technically true that the isotopic composition in spent fuel becomes more weapon-like as the fuel ages. But it is very slow because Pu-240 still has a half life in excess of 6500 years. It would take more than 13000 years for the decay of Pu-241 to improve the isotopic profile of today's power reactor plutonium to something near "weapons grade."


It's sunken cost. Anything you can do to extract value out of them offsets the cost of maintaining the SNF repos.

90% percent of uranium is not burnt, and it's fuel for MSR. Subtract the costs of movilizing the fuel, and tell me it's not worth dumping it in a MSR to be burtn almost to the 100%.


This would be a good argument if we already had industrial scale molten salt reactors just waiting to burn up spent fuel from light water reactors. But we don't, so anyone wishing to burn up spent fuel this way has to pay additional costs for researching and constructing the new reactors.


Nuclear waste is also mostly fuel for hybrid fusion-fission reactors. From Wikipedia[1]:

> Hybrid nuclear fusion–fission (hybrid nuclear power) is a proposed means of generating power by use of a combination of nuclear fusion and fission processes. The basic idea is to use high-energy fast neutrons from a fusion reactor to trigger fission in otherwise nonfissile fuels like U-238 or Th-232. Each neutron can trigger several fission events, multiplying the energy released by each fusion reaction hundreds of times. This would not only make fusion designs more economical in power terms, but also be able to burn fuels that were not suitable for use in conventional fission plants, even their nuclear waste.

As an extra advantage, such a reactor would be subcritical (similar to a Accelerator-driven subcritical reactor [2]) so it would never be at risk of a melt-down.

On the other hand, I'm pretty sure that such a reactor would still be a proliferation risk like fission breeder reactors.

[1] https://en.wikipedia.org/wiki/Nuclear_fusion%E2%80%93fission...

[2] https://en.wikipedia.org/wiki/Accelerator-driven_subcritical...


> so it would never be at risk of a melt-down

This is a misunderstanding of what a melt-down is. In every case we know of involving commercial reactors, fuel melt-down happened when the reactor was non-critical, completely devoid of fission.

Meltdown happens when the heat generated by decay of products produced when the reactor was critical cannot be removed fast enough, such as when a cooling system fails or coolant does not cover the fuel modules.

Any reactor with solid fuel and a high enough energy density will have a meltdown possibility, unless it is somehow using a reaction chain that precludes the production of high-activity fission products.


I wanted to make almost the same comment, but this part is not 100% accurate:

> In every case we know of involving commercial reactors, fuel melt-down happened when the reactor was non-critical, completely devoid of fission.

This is indeed true for Three Mile Island, Fukushima and Saint-Laurent-des-Eaux in 1969, but IIRC not for Chernobyl, and this was clearly not the case in Saint-Laurent-des-Eaux in 1980. (Yes, Saint-Laurent-des-Eaux faced two indendant meltdown events…).

More generally, as long as you're using light water as a coolant and moderator (PWR and BWR), you're likely to have negative void coefficient, which prevents fission-induced meltdown, but for other technologies, the risk exists.


I'll accept the correction regarding Saint-Laurent-des-Eaux, thank you.

However the steam explosion(s) that blew apart the reactor assembly at Chernobyl did not melt the fuel rods there although of course there may have been some localized areas of melting. The melt that created the tons of corium was due to decay heat. It occurred many hours after the power excursion and explosion(s).


Regardless of precisely what a “melt down” is or whether this was the appropriate term here, the Chernobyl explosion was caused by an excessive amount of fission output in a super-critical reactor. As I understand it, in a fission- or accelerator-driven reactor, if you turn it off, fission stops rapidly. Decay heat will remain, but that’s a smaller amount of output power and is presumably a good deal less dangerous.


That's way enough to cause an accident though, see Three Mile Island and Fukushima.


Fukushima successfully SCRAMed their reactor. Previously produced fission byproducts continued to decay after criticality was lost. While the thermal power of this decay process was an order of magnitude below peak, it greatly exceeded passive cooling capability. Without backup power to actively cool the core, meltdown is unavoidable: this decay heat is far more than enough to destroy the core.

This is a real risk for current reactor designs. Loss of backup power means a loss of any safe shutdown option.


"Mostly"? It's not like spent fuel is the only kind of nuclear waste...there's also tons of radioactive building materials, contaminated water and whatnot. Probably not as dangerous as (spent) fuel, but radioctive and voluminous enough to be a problem.


> there's also tons of radioactive building materials

It's a rare case where this is an accurate statement, because a typical nuclear power plant produces a dozen or two tons of waste per year. Take a look at this observation until it sinks in just how little waste that is.


How heavy are the buildings housing the reactor? How much of that material is radioactive/unsafe after a couple of years of operation?


Molten Salt reactors would also be much, much smaller decreasing that problem significantly as well.


This article is misrepresenting the truth. Maybe the majority of the waste mass is really uranium (i doubt it, but i'm no expert), but from what i heard from an EPR engineer, the volume of this waste is in fact tools, clothing, used part that have been replaced and stuff like that. Not hard to take care of, and waste that contain very low level of radioactivity, but this should be the majority of nuclear waste, and this is not "mostly fuel" in any case.

Please be precise.


A more precise version of the title would refer to "high-level nuclear waste":

https://en.wikipedia.org/wiki/High-level_waste


The majority of the waste mass, is, indeed, unburnt uranium.


Which is why the UK was importing nuclear waste for a while. Whether it is economic is another thing. Is hard to tell from the UK experience, since that collapsed due to faking data for fuel pellets that were being sold to Tepco.


What is the relation of cladding vs. content in a fuel rod? AFAIU any used fuel rod needs to be disassembled to be freed from its cladding, and only then can be reprocessed by whichever means to make new fuel rods with new cladding, which leaves the old cladding as irradiated trash. Am i popping a bubble there? Or is there a way to somehow melt that stuff into new cladding. Anybody thought about that? I mean, just look at the pictures in

https://duckduckgo.com/?q=Fuel+Rod&t=ffsb&iax=images&ia=imag...

or

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

Anybody thougt about that?

Edit: I mean even IF used in molten salt reactors where there are no rods and therefore no cladding, what happens with the old cladding? That's not just some tinfoil where you could make hats out of it!


Well the cladding is radioactive but to a far lesser degree then the fuel itself. Its waste but lower level waste as it doesn't contain the most long lived elements. I mean, they will be their anyway, so your not creating more waste buy having the old rods.

Also, most molten salt reactors wouldn't use fuel rod. And the ones that do would be different compared to those in PWR so reuse is unlikely.

Edit: See here https://youtu.be/TvXcoSdXYlk?t=1828


Its fuel for any kind of reactor. The process is a bit easier if you make salt fuel.

Its makes sense if you also get money to consume the waste otherwise its cheaper to go with freshly mined uranium. Or if you are interested in a fuel mix that is easier to obtain in spent fuel.

Edit: For those interested, one of the company engaged in the Canadian regulator process is proposing this:

https://youtu.be/TvXcoSdXYlk?t=1828

MSR Reprocessing - Sylvie Delpech: https://www.youtube.com/watch?v=fBEjys1SKCQ

This is a talk by the company that is furthest along the timeline of bringing molten salt to the market and he talks a little about their fuel as well (they don't do reprocessing): https://www.youtube.com/watch?v=l8QVxYrQRxA


How much waste do the molten salt and fast reactors create?


It depends. There are a variety of differences.

In terms of volume its far less then with a PWR as there you have lots of other stuff in there as well. For the same amount of power, the waste from a MSR will much less and would have a better waste profile.

However beyond that, it all depends. There are many different designs, different inputs, different burn rates and so on and so on.


And monkeys might come flying out of my butt except in our post enlightenment world we need facts, observations, and reasoned evidence to support claims. Molten salt reactors have so far proven extremely difficult to build and operate and tend to leak molten salt. Fast reactors are mostly not a thing at this point. Making complex technologies work means not only overcoming technical challenges but doing so in a way that constrains both costs and risks which so far has proven to be extremely difficult.


Nuclear waste produced per person is on the order of kilograms versus the volume and mass of carbon dioxide which is in the tonnes.

As long as the volume of nuclear waste is easy to manage, it is more likely we will bury nuclear waste than use it, as opposed to carbon dioxide which is far less likely to be injected into the Earth.


We'll soon find out. China is reportedly spending quite a lot of money trying out LFTR reactors.


> proven extremely difficult to build and operate and tend to leak molten salt

The only are 2 molten salt reactor ever built and neither one had a leaking problem. In fact, they would even be self sealing as the salt will cool down rapidly and seal the hole. If you have a rupture in the whole core, the spilled out salt would be captured in a pool and harden, it would be contained in the nuclear site boundary. This is the advantage of high temperature and salt. The temperature differential to the environment that the cooldown is very rapid.

Compared to PWR, the danger of a a molten salt reactor is incredibly small. There is no high pressure and steam, thus no explosion that can transport stuff like zirconium outside of the nuclear boundary.

Its also false that they have been extremely difficult to build. The Molten Salt Reactor experiment was operated and built by small time at a tiny fraction of the budget that were spent on many of the other nuclear reactors built at that time. They are far easier to build then a PWR in terms of complexity of and scale involved. The Molten Salt reactor experiment (the first ever molten salt reactor) was one of the fastest nuclear reactor was ever built.

The reason we don't have them has more to do with the history of nuclear technology. The Navy focused on PWR for ships (as makes sense=, those were commercialized with big investment of government and industry and were the most mature by far. Molten Salt reactors were in development by the Navy and were discontinued when rocket technology made nuclear bombers unneeded. After that the inventor of molten salt reactors only got a tiny amount of money to continue his research for a civilian reactor.

The big nuclear players didn't really care about building new reactor types as they had lots and lots of orders of the production designs they already had. There were only a very small number of people who were educated about what a molten salt reactor or how it worked. It required a somewhat different skill set (the inventor was a chemist) and neither government nor industry had a taste for commercialization at that moment. It didn't help that the inventor basically made himself unpopular with the regulatory agency when he claimed PWR were unsuited for civilian use.

When regulation changed because of TMI and Chernoyl the nuclear companies had their hands full with rebuilding all the plants according to new regulation to keeping the existing fleet ruining (and finishing project that were being built) and by that time the regulation had changed so that it was basically impossible for any new type of reactor to be licensed. In the US the technology was hard coded in to the regulation and made it impossible to deploy a new type of reactor in a commercial way. Since then hardly any new reactor type has been built or tested as you can not get fuel for anything beyond a tiny research scale reactor without passing full commercial regulation.

Development of pretty much all serious molten salt fuel has now moved to Canada as the regulators there have not hard coded specific technologies and the regulator investing its own resources to get threw regulation of such a reactor. In the US, you would have to pay the regulator to evaluate your design with no timeline or guarantee that would even allow it at all. Meaning you would have to design a full commercial reactor before the regulators even tell what you would need to do to convince them about the reactor. Meaning you potentially spend 10s of millions on paying the regulator before the actual regulatory process could even start, as you would have to finance the development of a regulatory framework for molten salt reactors.

Thankfully the DoE has realized this now and they are working hard at chaining their approach and congress is for once actually engaging in a bipartisan effort to improve the situation. See this talk for more information about government efforts: https://www.youtube.com/watch?v=p1lkDRX2huM


You can't wave away the remaining challenges of a molten salt reactor. It is a legitimately difficult materials science problem to operate a liquid salt pump loop operating at ~1000K and under high neutron flux.

It's also not a super-great explanation that a major accident will just cause the salt to flow into the holding tank. What if the holding tank is full of water? Why would it be full of water? I don't know but the planet is absolutely filthy with water and "the basement is unexpectedly full of water" has happened to almost every building humans have ever built. These are supposed to be installed _entirely_ underground, enhancing the danger of water intrusion.

Finally there is still the challenge of continuous processing of the liquid fuel, to remove ash, add fissile materials, and separate gases. Nobody knows how to do this. Nobody has explained what to do with all the radioactive krypton etc.


Nobody has explained what to do with all the radioactive krypton etc.

But they have. There's dozens of youtube videos on MSR.


It is a legitimately difficult materials science problem to operate a liquid salt pump loop operating at ~1000K and under high neutron flux.

And so it's containing the fusion plasma, yet billions are poured into material research. And hastelloy N is already invented.


Hastelloy N will not be used by any of the companies building commercial reactor. Hastelloy N would have to go threw a whole lot of regulation to allowed to use in modern reactor.

The data gathered on its properties would not be sufficient for a modern regulator, and such a qualification would take way to long. Also, there is not really a supplier for commercial Hastelloy N.

Their focus is on using 'regular' nuclear grade materials that have already been approved. This is THE single most important focus of those reactor design. In the choice of containment, reactor type, salt solution and so on, they all picked sub-optimal things in order to make it threw the regulator and have a existing supply chain.


I don't wave anything away. My point is there are not fundamental challenges. Its a large expensive engineering project, but you don't need any miracles. We built a working molten salt reactor with 9m and 30 people and they built it in 4 years and then operated it for 10-15 years in the 60-70s. Not Magic.

> It is a legitimately difficult materials science problem to operate a liquid salt pump loop operating at ~1000K and under high neutron flux.

Depending on your reactor design your not actual pumping fuel salt itself at all. Both leading contenders use a pool type system and not LFTR type system where you have to do substantial pumping, specially not of the fuel salt. Neither system in fact pumps fuel salt that is in neutron flux at all.

Have you actually looked at the real world proposed reactors we are talking about?

> It's also not a super-great explanation that a major accident will just cause the salt to flow into the holding tank. What if the holding tank is full of water? Why would it be full of water? I don't know but the planet is absolutely filthy with water and "the basement is unexpectedly full of water" has happened to almost every building humans have ever built. These are supposed to be installed _entirely_ underground, enhancing the danger of water intrusion.

In the case of IMSR, there is 1 meter nuclear grade concrete then a thick wall of steel, then a meter of a secondary cooling salt buffer, then another thick wall of steel, then the primary reactor vessel that is another massive steel cylinder. There are no opening anywhere except from the top.

That whole thing is in a hole that is topped of heavy steel protection on top and that is in building that is safe from aircraft attack. Even if you would have contact of water with the fuel salt the water would be in contact with the top layer of the fuel salt (far less surface area then fuel tubes), that would would freeze the top layer into a solid layer and shortly after the whole reactor would essentially be one block of salt in a metal concrete box.

These reactors are also not depended on water cooling at all, meaning you can build them in the desert or wherever, you don't need to be close to water in any way shape or form. Also you don't just build these anywhere, there is a citing process that looks at these like that.

Next, there is the point that in liquid salt actually does not produce many of the same hazards as fuel pellets. There are gases inside of the solid fuel pellets that get released. In a molten salt reactor the cesium and iodine are not gases, but rather they are bound in the fuel and will not get released in gas form.

This is one of the primary reasons molten salts were proposed int he first place.

So even if you flood the building and a meteor somehow made a hole in the building and reactor containment, and then you would flood it with water that would somehow show up. And all you would achieve with this is a gigantic pool heater, as a solid block of salt sits in the bottom of a hole and produces some steam (without being a system under pressure) that didn't contain the primary dangerous gases we talk about in the 'nuclear cloud'.

The amount of highly improbable events you have to chain to even achieve any danger outside of the nuclear regulated zone borders on comical. And even then its not even close to what we saw happen in Japan.

> Finally there is still the challenge of continuous processing of the liquid fuel, to remove ash, add fissile materials, and separate gases. Nobody knows how to do this. Nobody has explained what to do with all the radioactive krypton etc.

You don't actually need continuous processing and neither company that proposes to build Molten Salt reactors are proposing that.

So again, have you actually looked at the real world proposed reactors we are talking about? You seem to think all Molten Salt reactors are LFTRs.


Wikipedia reports that Beryllium salts (fluoride and chloride at least) are quite soluble in water and are fairly toxic. Do the leading designs use these salts or do they use something else that is insoluble?


Neither of the company wants to work with Beryllium. The reason being that its a incredibly limited resources (only one supplier) and the kinds of salts you would have to work with would have go threw a whole lot of analysis first. There is a talk by one of the people who worked on the original salt used in the reactor experiment you can find.

Beryllium has excellent properties but unless you are building a breeder are not really that essential. The primary concern for the companies working on this right now is commercialization, not peak performance. They are fine with loses and inefficiency.

You have to prove to the regulator that these salts will not be outside of the reactor boundary that takes care of both the radiation and the toxic aspect.


What companies are you talking about?


The furthest along is Terrestrial Energy, they already passed the first step in the Canadian review process.

Moltex Energy is still in the first stage of Canadian review process, as they had to relocate from Britain to Canada. Because the British regulatory system is not really active.

There is also Flibe Energy and they actually want to use lithium fluoride (LiF) and beryllium fluoride (BeF2) salt (hence the name). But that is a much longer term project and they are not anywhere close to commercialiastion. They are also working in the US Regulatory system that is not close to being ready to even start a serious commercionalisation process right now. They are also not targeting commercial energy as their primary market.

There is also ThorCon power I don't know as much about them and they are not engadged in any Western effort for commercialisation as far as I know.

There used to be Transatomic Power, but the never wanted to use beryllium fluoride either and have open sourced their reactor design.


10s of million dollars is a microscopic amount of money for a reactor if they can't get enough funding then they never deserved to build it.


One of the company already has a 1.4 billion loan grantee for a sighting project in the US. They are also supported by the government of Canada who intend to make molten salt reactors their next big thing as CANDU is ending. There is already a siting project in Canada as well. But you are right, if they can't get money they want do it.




Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: