By John Krzyzaniak, Nicholas R. Brown, December 18, 2019
Andrew Yang at the 2019 Iowa State Fair. Photo credit: Gage Skidmore
Andrew Yang, like many of the 2020 Democratic presidential hopefuls, has an ambitious plan to wean America off of fossil fuels. Unlike many of the other candidates, however, a key piece of his plan to address climate change involves harnessing nuclear power—in particular thorium. According to Yang, thorium is “superior to uranium on many levels.” But Yang isn’t alone; thorium boosters have been extolling its supposed virtues for years.
Do the claims about thorium actually hold up? The Bulletin reached out to Nicholas R. Brown, an associate professor in the department of Nuclear Engineering at the University of Tennessee, to examine five common claims about thorium and next-generation nuclear reactors. Brown’s responses are below.
Overall, although existing and new nuclear reactors may indeed be part of a long-term carbon-free energy mix in the United States, the public has good reason to be skeptical that thorium can or should play any role in the future.
Claim: Thorium reactors would be more economical than traditional uranium reactors, particularly because thorium is more abundant than uranium, has more energy potential than uranium, and doesn’t have to be enriched.
False. Although thorium is more abundant than uranium, the cost of uranium is a small fraction of the overall cost of nuclear energy. Nuclear energy economics are driven by the capital cost of the plant, and building a power plant with a thorium reactor is no cheaper than building a power plant with a uranium reactor. Further, using thorium in existing reactors is technically possible, but it would not provide any clear commercial benefit and would require other new infrastructure.
Additionally, there is technically no such thing as a thorium reactor. Thorium has no isotopes that readily fission to produce energy. So thorium is not usable as a fuel directly, but is instead a fertile nucleus that can be converted to uranium in a reactor. Only after conversion to uranium does thorium become useful as a nuclear fuel. So, even for a reactor that would use thorium within its fuel cycle, most energy produced would actually come from uranium fissions.
Claim: Next generation thorium reactors would be safer than current reactors.
True but misleading. Nuclear energy is already very safe, and Yang is correct about that. The current US nuclear fleet generates about 20 percent of all electricity in the United States and has an excellent safety record—despite accidents such as Three Mile Island. When it comes to new reactors, although some next-generation designs offer potential safety benefits relative to current reactors, they could be operated in either thorium-uranium or uranium-plutonium fuel cycles. Consequently, the benefits are a function of the inherent safety in the next-generation designs, not the utilization of thorium.
Claim: The waste from thorium reactors would be easier to deal with than waste from today’s uranium reactors.
False. A comprehensive study from the US Energy Department in 2014 found that waste from thorium-uranium fuel cycles has similar radioactivity at 100 years to uranium-plutonium fuel cycles, and actually has higher waste radioactivity at 100,000 years.
Claim: Thorium would be more proliferation-resistant than current reactors—you can’t make nuclear weapons out of it.
False. A 2012 study funded by the National Nuclear Security Administration found that the byproducts of a thorium fuel cycle, in particular uranium 233, can potentially be attractive material for making nuclear weapons. A 2012 study published in Nature from the University of Cambridge also concluded that thorium fuel cycles pose significant proliferation risks.
Claim: Building new nuclear reactors will likely be necessary if the United States wants to achieve net-zero emissions by 2049.
True. Nuclear energy is already the primary ultra-low carbon energy source for base-load electricity generation. Although solar and wind have their place in the energy mix, the primary benefit of nuclear energy is that it is not intermittent, as solar and wind are, so it is almost always available without needing energy storage. So new nuclear reactors will be necessary to both replace aging ones and to meet a net-zero carbon emissions goal. But thorium-uranium fuel cycles provide no inherent benefits relative to uranium-plutonium fuel cycles, so the new reactors need not be thorium-powered.
The Bulletin elevates expert voices above the noise. But as an independent, nonprofit media organization, our operations depend on the support of readers like you. Help us continue to deliver quality journalism that holds leaders accountable. Your support of our work at any level is important. In return, we promise our coverage will be understandable, influential, vigilant, solution-oriented, and fair-minded. Together we can make a difference.
View Comments
The statement about weapons proliferation needs more detail. Commercial LWR power plants are not proliferant. Their spent fuel does contain plutonium, but the isotopic composition of spent fuel is not suitable for military weapons. The statement does not take account of the specific design of a power plant using thorium. In the case of the ThorCon liquid fission plant, the proliferation resistance is better than that of a standard LWR. I recommend the authors review the document at http://thorconpower.com/docs/docs_safeguards_pub.pdf and explore others on the website ThorConPower.com.
True that LWRs are not a major proliferation problem, as long as the spent fuel is not reprocessed. That is because you can't steal spent fuel and separate the plutonium. It is too radioactive, at least for the next several hundred years.
The isotopic composition has nothing to do with it.
If LWR fuel is reprocessed, however, the plutonium can be used for bombs by governments or terrorists, irrespective of isotopic composition.
This is publicly known since the 1970's. See for example https://fas.org/rlg/980826-pu.htm
Thorium needs reprocessing in any case.
For that reason it does not solve the (not-so-severe) problem of LWR proliferation. It makes it far worse.
Yes. You can make a bomb from Pu240 and Pu 242, extracted from LWR waste but spontaneous fissioning makes the yield terrible, and risks a fizzle.
Its possible, but barely plausible. LWR reactor waste diversion, for weapons, is senseless.
These arguments are in reference to solid-fuel thorium reactors, when the design being most discussed for the past decade is a liquid-fueled thorium reactor (LFTR). The latter is greatly superior regarding resistance to weaponization.
There are two options with conventional solid-uranium fueled LWR waste: store or re-process. Neither of these is an issue with an LFTR.
If the waste is stored, it eventually cools (by radioactive decay) to the point it could be conveniently and safely handled. Eventually the heavier plutonium isotopes eventually decay, leaving the much longer-lived, weaponizable isotope Pu-239. It may be thousands of years in the future, but it is simply irresponsible to say the future is not our problem!
If the waste is to be recycled (used for fuel in a different kind of reactor), it is re-processed, at which point the plutonium is separated. This plutonium is typically not of weapons grade, but it is still nasty stuff that would be very scary in the wrong hands.
In contrast, a liquid-fueled reactor continuously cycles its fuel, with non-useful components continuously being removed. In a LFTR, particularly, the uranium and plutonium isotopes that arise are useful for energy production, so they are not removed, and are burned down for energy. Thus they do not form a major part of the waste.
Although the normal waste from a LFTR is not weaponizable, it would be possible for the operator of an LFTR to re-engineer the reactor to produce whatever isotopes they want -- a possibility that exists with any reactor. But to do so would require top-level management, and even top political direction. It couldn't be done without top-level detection.
But if a country wants plutonium, an LFTR would be a very expensive and inefficient way of producing it, compared to conventional reactors that were designed for that purpose.
It is true that the *wrong kind* of thorium-based reactor would have its own weapons proliferation problem. This is not an argument against an LFTR.
There are a lot of flaws in this analysis of Andrew Yang's claims. Admittedly Yang does not clarify what type of thorium reactor he is referring to, but from his comments it is likely that he is referring to the liquid flouride throrium reactor (LFTR) or some similar variant, which uses liquid fluoride as the coolant rather than water, and does not use solid fuel, but rather adds the fuel (unanium 233) to the liquid coolant that runs through the system.
COST:
The first claim from Andrew Yang is that a thorium reactor would be cheaper. The article indicates that this is a false claim. However, the cost savings would not primarily come from the cost of fuel, but rather from the cost of construction. Constructing traditional reactors is very expensive due to the fact that traditional reactors run at high pressure. Safety is considered a much higher priority in the Nuclear industry than in other energy sectors such as coal. A great deal of the cost of building a plant is tied to the safety requirements. And many of the safety requirements are tied to the fact that a traditional plant runs at high pressure.
The LFTR is designed to run a atmospheric pressure, which eliminates one major factor that has to be addressed in traditional nuclear power plants. Of course, one of the large parts of the cost of building a plant is certification of the design, and right now no one knows how much that will cost since the standard process for qualifying a plant design approved by the nuclear regulatory commission (NRC) is based on traditional designs, and the NRC has not defined how nuclear plants will be qualified for designs that differ significantly from the traditional. The expectation is that once a LFTR design has been approved, it will be much less expensive to build since the factors that drive up the cost of traditional plants are not so much an issue with the LFTR design.
A related design that uses a liquid coolant and includes the fuel in the coolant is Terrestrial energy's integral molten salt reactor (IMSR). This has similar safety properties to the LFTR, but does not use thorium. The company is currently in the process of qualifying this design in Canada. They claim that it will be much cheaper to build than traditional plants, and once they get their design qualified, we will see if they are right. If they are prove to be correct, it will be a real boost for proponents of LFTR designs.
WASTE:
Another claim from Yang is that waste from the thorium reactor would be easier to deal with. The article claims this is false. However, if you are talking about a plant that uses a liquid fuel, this is almost certainly true. Note that this does not apply to thorium reactors, per se, but is applicable to reactors like the IMSR that use uranium fuel.
The reason that they are much better in terms of waste is that a traditional nuclear plant has to remove fuel after a small fraction of the fuel has been used up (maybe 2%). The rest of the 98% of the fuel is stored as waste. With a plant that uses liquid fuel, there is no need to remove the fuel when 2% of the fuel is used up. You just add more fuel. There is actually talk of using the waste fuel from traditional designs as the fuel these liquid fuel designs so that we can get rid of the waste that has already accumulated. So, yes, the waste is much easier to deal with because in the end, there is expected to be much less of it.
I think the claim that thorium reactors are more proliferation resistant has to do with the fact that the process of converting thorium to uranium 233 also results in generating uranium 232. Uranium 232 is quite easy to detect from large distances, which makes it undesirable for nuclear weapons (i.e., it is too easy to find them). Purifying uranium 233 by removing uranium 232 is very difficult (expensive) and is not 100% effective, which makes uranium 233 a poor choice for building weapons.
You wouldn't let U232 accumulate in the U233, in the first place. U233 can be sorted very cleanly.
The proliferation risk is not zero.
On the other hand, neutron efficiency in a Th MSR is so tight, extracting U233 out of the system would have you loading in new Th at a relatively fast clip. Simple accounting input methods would be enough to spot a weapons grade U233 plant.
Loading in...? Are you talking about a solid-fueled reactor?
In an LFTR, fuel is loaded and re-processed continuously.
We do not need it, and we cannot afford more nuclear power adventures. The VC Summer plant was cancelled after the waste of $8,000,000,000, and the Vogtle plants in Georgia have such high construction costs their cost of power will be over 15 cents/kWh.
By contrast, the City of Los Angeles just contracted for PV power at under 2 cents/kWh daytime and 3.3 cents/kWh at night from utility-scale battery storage.
My own household and at least one of our two electric cars are powered by the PV system on our roof, which lets us put power into the grid in the daytime and take it out at night. Our battery storage is for the event of earthquakes.
The future got here a few years ago, . . why are many of us still stuck with 20th Century ideas?
Except thorium reactors *will* cost a lot less to build than a standard Light Water Reactor. There's no need for that kind of investment in the first place. They don't need containment dome and all kinds of scram features. Molten thorium is both the coolant and a fuel source. It is kept circulating by a pump and held inside the reactor by a freeze plug. It's walk away safe. Power goes down, the freeze plug melts, the thorium drains out simply by gravity.
If you took every battery in the US, even the ones in cars and trucks, they could store enough power to run the country for .... 30 minutes. They need to hold many days, even weeks worth of energy to function in an industrial society. Batteries aren't cheap and they're really hard to recycle.
That analysis isn't quite honest, is it?
MSRs may not need certain expensive features of conventional LWR plants but they will need other expensive features instead such as a chemical processing plant for their fuel salt for recycling of the salt and removal of fission products. Corrosion and how to control is still not fully solved. Heck, it's not even clear yet what chemical composition the salt should be - but we do know that if your salt contains Lithium, you'd have to use enriched Li-7 which is another expensive feature over LWRs.
At this early stage I find such definitive claims about the costs of MSRs rather questionable.
Either way, at least batteries are actually existing and reliable technology unlike molten salt reactors which are still unproven, experimental technology.
On the other hand, renewable energy doesn't only consist of PV backed up by batteries. For example, solar thermal power plants with molten salt energy storage can deliver both high temperature heat for industrial processes as well as energy production well past sundown. Hydrogen from water splitting has the potential to replace coal and fossil fuels as a clean energy carrier.
Speaking of honesty...
In a comparison of conventional LWRs with LFTRs, both of them require extensive chemical facilities. One of the huge advantages of MSRs over solid-fuel reactors is that chemical processing can be performed on the fly in MSRs -- your tone is that this was somehow a bad thing...
As to expense -- all power generation solutions have their costs. For example, you have simply chosen disregard the cost of the presence of tens of thousands of windmills, or square miles of solar cells, and huge battery installations.
The point being made was that LFTR reduces both capital cost and operating cost considerably over conventional designs. But if you want to compare nuclear against renewable, you should consider that the ecological footprint of nuclear installations is tiny compared with any renewable power solution.
As to the engineering details -- this is why test facilities are needed. You are arguing that we shouldn't research a question because we don't know all the answers.
These are not either-or options, you know. Of course in the near future there will be some mix of renewable and nuclear power, which both serve to reduce the amount of geological carbon we pump into the biosphere.
We have a big problem coming up, and we need to look at all the solutions, renewable and nuclear alike. To deny either one is counterproductive.
MSRs can be used to make U233/U232, which is easily fashioned into low-yield nuclear weapons. Bomb experts from Los Alamos and Lawrence LIvermore have already described how easy it is to do this.
I ask that amateurs should stop pumping Thorium molten salt reactors. Let the experts and the scientists, who understand the science, handle this. Please.
The last claim, that nuclear is necessary for zero emissions, is open to doubt.
Wind and solar power will soon overtake nuclear, in annual terawatt-hours, in the world.
It has already done so in China, India, the UK, Germany, Italy, Indonesia, Brazil, Mexico, Spain and in fact most nations of the world, according to data for 2018 in the BP statistical review of world energy.
It has been done, it can be done.
There is no reason why wind, photovoltaics, thermal solar, geothermal, existing hydro and biomass could not supply 100 per cent of the global electricity, together with demand side management, batteries and possibly with some part for hydrogen storage.
You can argue that nuclear is better for some reason or other, not that it is necessary.
The problem with your statement is not in the actual production of the energy, but in the space required to produce energy from those sources, the toxicity of the materials used, versus the after effects of placement.
Hydroelectric Dams while close to carbon neutral after production requires large bodies of water to generate minimal amounts of electricity. To produce said dams, a large area has to be flooded before energy production can take place. While this method does not release carbon into the atmosphere, it still causes large area scale environment destruction.
Wind turbines are not as clean as one would expect either. Again, large areas of land need to be cleared just to erect the turbine assembly and drivetrain to include the logistics of transporting the massive components. The amount of energy produced is also very low. Wind farms produce close to half of a Mega Watt Hour at peak, but only runs at peak performance for a third of its life span.
Solar by far is the dirtiest renewable on the market. Solar panel contain three extremely toxic materials; Lithium, Cadmium, and Chromium. The same goes for many "high-capacity" batteries. Cadmium, and Chromium are known carcinogens with Chromium being also more radioactive than plutonium. Lithium is highly reactive with water and carbon, making all three elements highly toxic to almost all forms of life. In addition the average output for a full scale solar plant is about .7Mwh per acre. Not very efficient when compared to nuclear and fossil fuel plants.
There is a place in the power grid for renewable energy, but not as a primary source. Take a look at France (whom has the lowest carbon emissions in the world), 73 percent of their power comes from Nuclear Fission and they have yet to have a serious nuclear accident since the 60s. I like that you have quoted Germany as having a massive renewable energy infrastructure. While it is true that close to 40 percent of Germany's grid is renewable, they have double the carbon footprint of France and the footprint is increasing.
My final point is that while everyone is seeking clean energy through renewables, they are ignoring the cleanest source of energy of all. Nuclear Fusion. Nuclear Fusion chain reactions are easy to produce, (Example: Hydrogen Bomb) but difficult to contain. Electromagnets have been proven capable of containing the hot plasma required for the reaction,but require massive amount of electricity to get the reactor running. An energy demand that renewables cannot produce all at once, unlike a Fission Reactor
Regarding your last remark about fusion: it is false that fusion has been ignored. It has been heavily funded for many decades. There are U.S. congresspersons currently pushing it.
And for fifty years, we have been told of fusion break-throughs, and solutions around the corner. The truth is, more break-throughs will be required to make it practical; there must be a solution, but it would be naive to believe it's around the next corner or the one after.
Fusion is indeed the dream energy source. We must continue to pursue it -- but it would be irrational to rely on it to solve immediate problems.
It appears the fact checking was only for solid fuel reactors. In that sense they are correct. But Wang has also noted that GenIV reactors including Molten Salt Reactors (for thorium as well as uranium) which are much cheaper to build since, at least at one level, the reactor vessel and balance of plant, has only to be built for atmospheric pressure, unlike in today's GenIII pressurized water reactors. The only area where the fact checking is quite wrong is in acquiring thorium. Thorium only has to be milled to purity using existing industrial technology. There is no enrichment. Many designs include reprocessing integral to the reactor and thus can remove unfissionables far easier producing far less waste. The bottom line that thorium is 4 times more abundant than uranium and is easier to process for fuel. Molten Salt Reactors be they running on uranium or thorium will be cheaper to build and safer to run than even today's very safe reactors.
You are right that Thorium is 4 times more abundant than Uranium. But note that the Uranium 235, which is directly usable in reactors, is less than 1% of the uranium that is mined. So Thorium is really much easier to access by a factor of hundreds than the typical uranium used in reactors.
Lance, I think you are comparing the claims of LFTR reactors with conventional uranium LWR reactors. Liquid salt uranium reactors have some of the same efficiency advantages as LFTR -- in particular, they can be designed to operate with different neutron energy spectrums, so that they could in principle burn more of the input fuel than a slow-neutron LWR.
It is clear that conventional uranium LWRs are very inefficient in converting the input fuel to energy, compared to a LFTR. Breeder reactors can burn U-238, however, so this isn't so much a distinction between uranium and thorium as fuel, but rather, one of reactor design.
Each reactor design has its advantages and disadvantages, its own engineering challenges. But LFTR does indeed seem to be a sweet spot for our current needs.
This article completely overlooks the current outstanding issues presented by thorium reactors. First among many issues is the lack of metal alloys that can resist the extremely corrosive molten salt fuel mixtures.
Having materials that can withstand the corrosive salts is indeed a challenge. This was considered one of the great challenges to solve by the engineers and scientists that built the molten salt reactor experiment at Oak Ridge National Laboratories and ran it in the 60s. Significant progress was made at the time to find materials that could withstand the corrosive salts. Hastelloy-N was found to do a reasonable job. It is unfortunate that the research was halted by lack of support from the Nixon administration since this issue might have been resolved by now. The chinese appear to be spending quite a lot of research dollars to resolve problems like this for molten salt reactors.
The Fact Check’s refutation of the claim that one can’t make nuclear weapons from the thorium fuel cycle, can be strengthened by noting that the US has successfully tested nuclear weapons using U-233 from thorium -- see U.S. Congress Office of Technology Assessment 1994. Technical Options for the Advanced Liquid Metal Reactor---Background Paper. OTA-BP-ENV-126. Washington, DC: U.S. Government Printing Office, May.
Yes, Yang was incorrect to state that Thorium (or at least Uranium 233 is part of the Thorium decay chain) can't be used to to make nuclear weapons. It can. However, the decay chain also include Uranium 232, so you always get a mix of Uranium 233 and 232 when using thorium to generate fuel. Uranium 232 is a gamma emitter, and gamma radiation is easy to detect from long distances. Uranium 232 is also very difficult to handle because it is a gamma emitter. Separating the two completely is not practical. This makes uranium 233 much less desirable for nuclear weapons than other materials such as Uranium 235 or Plutonium. The fact that it is easier to monitor is the primary reason it is considered to be less of a proliferation concern.
Not true. By extracting the Pa233 as quickly as it's made, the U233 that decays from it is very clean.
Any reactor can produce pretty much any isotope -- that's kind of what they do. The questions here are: who is doing it, and would it be practical for them? Let's run through some scenarios.
Could a country misuse an LFTR to produce weapons-grade plutonium? I know of no reason why they couldn't, in principle. But would they go to the trouble and expense to use an LFTR to do something it wasn't designed to do, when there are well-known reactor designs that can produce plutonium very efficiently?
If a country wants plutonium, it is not that hard to make, but an LFTR would be a very poor way to go about it. I'm going to say, no.
Could a terrorist group take over a LFTR reactor, and re-engineeer the fuel cycle to produce enough weapons-grade plutonium to make a bomb? I'm going to just say, again, no.
Could a rogue administartor of an LFTR twist it to carry out a fiendish plot? In real life, any such activity would be detected and stopped. But let's say the administrator zaps all the workers with a mind-control ray to make them do their bidding and tell nobody ... it probably already happened in a sci-fi movie. But in real life -- no.
My imagination is used up. I would like to hear any other ideas.
In normal operation, an LFTR would not produce practically weaponizable materials, and it would not be practical to re-engineer it to do so -- outside of action-adventure movies.
The Molten Salt Reactor, which does not have to be thorium fuelled does not need expensive steam containment structure, therefore, it is mass producible. It is based on molten salts and there are a variety of designs that MUST be approved FOR TESTING (see the best one). One uses molten salt in tubes surrounded by non fissionable fuel salts. Others are just like the old MSRE that Nixon nixed for reasons of??? Still others are like LFTR, using thorium.
This article totally forgot the heart of "thorium", the MSR!
These points are not helpful to the larger discussion. The type of reactor Yang references, a molten salt reactor, is still very much worth pursuing because 1: It is not water-cooled, and 2: It does not operate under pressure. This means that the reactors can be deployed anywhere, not just near a body of water, and that in the event of an emergency they shut down on their own without human guidance. This isn't USA warming, it's GLOBAL warming. We need an energy-dense technology that can be deployed in remote regions of the world without a massive team of nuclear specialists standing by 24/7. This technology has the potential to partially decarbonize manufacturing and heating, which together account for more carbon than electricity. These reactors are also physically scalable, which means smaller reactors can be made on an assembly line, drastically reducing cost. I'm assuming that if this author is a nuclear scientist, they are probably pro-nuclear since they have devoted their life to studying the technology. To the expert who provided this article: is this article useful?