Seven years ago, CleanTechnica published its policy position to not cover thorium nuclear reactors. Today, the United States has a Democratic presidential candidate in the top 10 who loves thorium, yet CleanTechnica still ignores it. Why is that?
Let’s start with the candidate, Andrew Yang.
Yes, the Andrew Yang campaign is promising next generation reactors on the grid in eight years. This was covered as part of the And then there’s the bad section of CleanTechnica‘s review of his climate action plan (tl;dr: good on carbon tax, meh on other things, really bad on energy).
What did CleanTechnica say about thorium and other next-generation nuclear in our review of his plan?
“the most realistic timeframe for fusion in actual utility-scale generation is 2050 at the earliest (and more likely much later), and there are exactly zero thorium nuclear plants operating in the world. The history of building nuclear that we know how to build today indicates a 10–15 year timeframe for the known technology. […]
The thorium crowd likes to point at India and China, but China has committed to only build a couple of molten salt reactors with a 12 MW capacity that might use thorium at some future date, and India has even less ambitious plans. […]
As Mark Z. Jacobson said in our email conversation, […]:
“I think the more wasteful parts of his proposal are spending on thorium, nuclear fusion, and geoengineering. None of these has a chance to help solve the problem, which we need implemented now. They are all opportunity costs.”
There is no empirical evidence suggesting that a 2027 time frame is remotely likely, and Top 100 Climate Influencing academics think it’s wasteful and not helpful.
So why did CleanTechnica stop bothering with this in 2012? At the time, the evidence was just as clear. CleanTechnica published a piece duplicating a fact sheet on the subject from 2009, a decade ago. The fact sheet was as follows.
By Arjun Makhijani and Michele Boyd
A Fact Sheet Produced by the Institute for Energy and Environmental Research and Physicians for Social Responsibility
Thorium “fuel” has been proposed as an alternative to uranium fuel in nuclear reactors. There are not “thorium reactors,” but rather proposals to use thorium as a “fuel” in different types of reactors, including existing light-water reactors and various fast breeder reactor designs.
Thorium, which refers to thorium-232, is a radioactive metal that is about three times more abundant than uranium in the natural environment. Large known deposits are in Australia, India, and Norway. Some of the largest reserves are found in Idaho in the U.S. The primary U.S. company advocating for thorium fuel is Thorium Power (www.thoriumpower.com). Contrary to the claims made or implied by thorium proponents, however, thorium doesn’t solve the proliferation, waste, safety, or cost problems of nuclear power, and it still faces major technical hurdles for commercialization.
Not a Proliferation Solution
Thorium is not actually a “fuel” because it is not fissile and therefore cannot be used to start or sustain a nuclear chain reaction. A fissile material, such as uranium-235 (U-235) or plutonium-239 (which is made in reactors from uranium-238), is required to kick-start the reaction. The enriched uranium fuel or plutonium fuel also maintains the chain reaction until enough of the thorium target material has been converted into fissile uranium-233 (U-233) to take over much or most of the job. An advantage of thorium is that it absorbs slow neutrons relatively efficiently (compared to uranium-238) to produce fissile uranium-233.
The use of enriched uranium or plutonium in thorium fuel has proliferation implications. Although U-235 is found in nature, it is only 0.7 percent of natural uranium, so the proportion of U-235 must be industrially increased to make “enriched uranium” for use in reactors. Highly enriched uranium and separated plutonium are nuclear weapons materials.
In addition, U-233 is as effective as plutonium-239 for making nuclear bombs. In most proposed thorium fuel cycles, reprocessing is required to separate out the U-233 for use in fresh fuel. This means that, like uranium fuel with reprocessing, bomb-making material is separated out, making it vulnerable to theft or diversion. Some proposed thorium fuel cycles even require 20% enriched uranium in order to get the chain reaction started in existing reactors using thorium fuel. It takes 90% enrichment to make weapons-usable uranium, but very little additional work is needed to move from 20% enrichment to 90% enrichment. Most of the separative work is needed to go from natural uranium, which has 0.7% uranium-235, to 20% U-235.
It has been claimed that thorium fuel cycles with reprocessing would be much less of a proliferation risk because the thorium can be mixed with uranium-238. In this case, fissile uranium-233 is also mixed with non-fissile uranium-238. The claim is that if the uranium-238 content is high enough, the mixture cannot be used to make bombs without a complex uranium enrichment plant. This is misleading. More uranium-238 does dilute the uranium-233, but it also results in the production of more plutonium-239 as the reactor operates. So the proliferation problem remains — either bomb-usable uranium-233 or bomb-useable plutonium is created and can be separated out by reprocessing.
Further, while an enrichment plant is needed to separate U-233 from U-238, it would take less separative work to do so than enriching natural uranium. This is because U-233 is five atomic weight units lighter than U-238, compared to only three for U-235. It is true that such enrichment would not be a straightforward matter because the U-233 is contaminated with U-232, which is highly radioactive and has very radioactive radionuclides in its decay chain. The radiation-dose-related problems associated with separating U-233 from U-238 and then handling the U-233 would be considerable and more complex than enriching natural uranium for the purpose of bomb making. But in principle, the separation can be done, especially if worker safety is not a primary concern; the resulting U-233 can be used to make bombs. There is just no way to avoid proliferation problems associated with thorium fuel cycles that involve reprocessing. Thorium fuel cycles without reprocessing would offer the same temptation to reprocess as today’s once-through uranium fuel cycles.
Not a Waste Solution
Proponents claim that thorium fuel significantly reduces the volume, weight, and long-term radiotoxicity of spent fuel. Using thorium in a nuclear reactor creates radioactive waste that proponents claim would only have to be isolated from the environment for 500 years, as opposed to the irradiated uranium-only fuel that remains dangerous for hundreds of thousands of years. This claim is wrong. The fission of thorium creates long-lived fission products like technetium-99 (half-life over 200,000 years). While the mix of fission products is somewhat different than with uranium fuel, the same range of fission products is created. With or without reprocessing, these fission products have to be disposed of in a geologic repository.
If the spent fuel is not reprocessed, thorium-232 is very-long lived (half-life:14 billion years) and its decay products will build up over time in the spent fuel. This will make the spent fuel quite radiotoxic, in addition to all the fission products in it. It should also be noted that inhalation of a unit of radioactivity of thorium-232 or thorium-228 (which is also present as a decay product of thorium-232) produces a far higher dose, especially to certain organs, than the inhalation of uranium containing the same amount of radioactivity. For instance, the bone surface dose from breathing an amount (mass) of insoluble thorium is about 200 times that of breathing the same mass of uranium.
Finally, the use of thorium also creates waste at the front end of the fuel cycle. The radioactivity associated with these is expected to be considerably less than that associated with a comparable amount of uranium milling. However, mine wastes will pose long-term hazards, as in the case of uranium mining. There are also often hazardous non-radioactive metals in both thorium and uranium mill tailings.
Ongoing Technical Problems
Research and development of thorium fuel has been undertaken in Germany, India, Japan, Russia, the UK, and the U.S. for more than half a century. Besides remote fuel fabrication and issues at the front end of the fuel cycle, thorium-U-233 breeder reactors produce fuel (“breed”) much more slowly than uranium-plutonium-239 breeders. This leads to technical complications. India is sometimes cited as the country that has successfully developed thorium fuel. In fact, India has been trying to develop a thorium breeder fuel cycle for decades but has not yet done so commercially.
One reason reprocessing thorium fuel cycles haven’t been successful is that uranium-232 (U-232) is created along with uranium-233. U-232, which has a half-life of about 70 years, is extremely radioactive and is therefore very dangerous in small quantities: a single small particle in a lung would exceed legal radiation standards for the general public. U-232 also has highly radioactive decay products. Therefore, fabricating fuel with U-233 is very
expensive and difficult.
Not an Economic Solution
Thorium may be abundant and possess certain technical advantages, but it does not mean that it is economical. Compared to uranium, the thorium fuel cycle is likely to be even more costly. In a once-through mode, it will need both uranium enrichment (or plutonium separation) and thorium target rod production. In a breeder configuration, it will need reprocessing, which is costly. In addition, as noted, inhalation of thorium-232 produces a higher dose than the same amount of uranium-238 (either by radioactivity or by weight).
Reprocessed thorium creates even more risks due to the highly radioactive U-232 created in the reactor. This makes worker protection more difficult and expensive for a given level of annual dose.
Fact sheet completed in January 2009
Updated July 2009
There were some criticisms of this fact sheet at the time, and CleanTechnica published them too. But seven years later, there’s still no commercial product, no commercial siting approved, no regulatory approval, and in fact not much of any movement.
As I said to a commenter elsewhere recently, I have a simple rule of thumb for assessing technologies (in addition to my often exceeding deep ways of assessing them):
If a technology has been in existence for decades and yet there are no commercial installations of it anywhere in the world, there is very little likelihood of it becoming viable.
This has proven true for airborne wind energy and many other technologies I’ve assessed, including next-generation nuclear.
Of course, I’ve looked at the likelihood of next-generation nuclear, including the molten-salt reactors necessary for thorium nuclear generation, in depth as well.
None are likely to be in commercial operation before 2040 at the earliest. By the time any get to commercial market availability, the cost of renewables will have only fallen further, making any nuclear even less competitive. How do we know? Let’s start with what the World Nuclear Association has to say about Gen IV reactors.
“After some two years’ deliberation and review of about one hundred concepts, late in 2002 GIF (then representing ten countries) announced the selection of six reactor technologies which they believe represent the future shape of nuclear energy. […] At least four of the systems have significant operating experience already in most respects of their design, which provides a good basis for further R&D and is likely to mean that they can be in commercial operation before 2030.”
2030 isn’t bad. But there’s more.
What about the Gen IV International Forum?
“Some of these reactor designs could be demonstrated within the next decade, with commercial deployment beginning in 2030.”
That’s an even less enthralling statement. Demonstrated sometime within ten years. Maybe commercial deployment beginning in 12 years. And given that the average nuclear plant takes 15 years for a commercial deployment, and that’s for known, established technologies, it could easily be 2045 by the time Gen IV reactors are pumping electricity into the grid commercially.
So in 2002, six technologies were chosen and it’s possible that a couple of them might be generating electricity commercially sometime after 2040. That’s 38+ years to commercialization of a technology. I’m trying to think of something equivalently slow-paced in the world of modern generation technology that isn’t fusion generation and mostly failing.
Maybe in 2002 the thought of competitive nuclear generation was reasonable. But the chart above highlights the precipitous drop in the price of both wind and solar technologies. Those renewable technologies have all of the advantages of nuclear — very low carbon emissions per MWh, very low pollution per MWh, low direct or secondary mortality impacts per MWh — a bunch of additional advantages — cheaper, faster, viable in all countries — with none of the disadvantages of nuclear — nuclear waste, terrorism concerns, inflexibility, low social license.
The low modern cost of wind and solar along with the low cost of natural gas generation and the inherent challenges of nuclear has seen a bunch of reactors shut down and fleet reduction announcements globally. There was a blip of increase due to China finally getting some of its reactors going, but that’s ended as China has delayed most new starts while simultaneously increasing wind and solar.
As the World Nuclear Association documents, nuclear started its decline in 2006, long before Fukushima. Most of the reactors in Japan that were shut down won’t be coming back on line. France is on track to reduce its nuclear capacity. The nuclear industry in the US is lobbying for subsidies and tax breaks to allow existing reactors to stay open, mostly without success. Toshiba Westinghouse is bankrupt and was bought by Brookfield for the 40 to 80 years of decommissioning service revenue, not to build new reactors. That’s the same reason SNC Lavalin bought CANDU.
It’s worth looking at the unsubsidized LCOE for different forms of generation, per Lazard. You’ll note what the cheapest forms of generation are in 2018: utility-scale wind and solar. You’ll note what isn’t cheapest: nuclear. And that isn’t commercially unproven technologies, that’s technology which has been being commissioned for 50 years.
Hinkley in the UK is a Gen III reactor design and continues to be pushed forward by the conservative government there despite a price tag of 15 cents USD per kWh guaranteed for 35 years. Similarly, other Gen III reactors in France and Finland are years and billions over budget. That’s the reality of nuclear technology advances. Nuclear doesn’t get cheaper. Unlike almost every other technology we’ve invented, it just gets more expensive.
There’s no reason to think that Gen IV reactors, whenever they actually get to a stage where they might be deployed commercially, will end up being cheaper than alternatives. The opposite is likely to be true, that they will be much more expensive than the alternatives and with other unappealing characteristics. They’ll have to fight for commercial share in a space where they are very much the unknown, unproven, risky alternative against very well known, widely deployed, effective and cheap competitors.
There’s little reason to consider thorium, molten salt reactors and Gates’ “traveling wave” TerraPower technology when considering the future of energy. We have solutions today. They may be boring and low-tech, but they are cheap, fast to build, reliable, predictable, and have incredibly low negative externalities. By the time any of these technologies actually see the market, they’ll be like the Christian concept of a god in a world of science, with nowhere to stand and nothing to do.
As a result, CleanTechnica‘s policy will be to continue to ignore them in favor of the actually transformative technologies reshaping our world for the better.