13 ms·

ITER is a showcase for the drawbacks of fusion energy

88 points by okket 9y ago

IntronExon 9y ago
It’s worth making a few notes: 1. ITER is a political beast, not a practical one. 2. These are problems with D-T fusion, not all fusion, although... 3. Aneutronic fusion isn’t even on the remote horizon. Fusion has incredible potential, but probably not for anyone alive today.
- ZenoArrow 9y ago
  Regarding point 2, the only drawback I'm aware of with D-T fusion is that it still produces nuclear waste (whereas aneutronic fusion produces zero/close to zero nuclear waste), however the nuclear waste produced by D-T fusion is said to be: * Smaller in size than the nuclear waste produced by the current approaches to nuclear fission. * Has a shorter half life than spent nuclear fission fuel, meaning that it doesn't take as long before it's safe to take it out of storage. Aside from nuclear waste, are there other drawbacks to D-T fusion that people should be aware of?
  - DennisP 9y ago
    Concern about tritium leaks, the need for tritium breeding which limits how fast you can build new reactors, hard neutron radiation that theoretically could be used to breed fissiles (but only if you have a high-enough tritium breeding ratio to keep the reactor fueled, despite not using all your neutrons for that), and the need for a steam turbine, which means you're not going to achieve the extremely low costs that may be achievable with aneutronic fusion. Which doesn't mean D-T is hopeless, just not as good as aneutronic. I've seen fusion scientists say the waste from D-T reactors would only need to be contained for several decades.
    - ZenoArrow 9y ago
      > "the need for a steam turbine" Where's the steam turbine in the following fusion approach? https://en.wikipedia.org/wiki/Dense_plasma_focus https://en.wikipedia.org/wiki/Dense_plasma_focus https://lppfusion.com/fusion-power/dpf-device/ https://lppfusion.com/fusion-power/dpf-device/
      - DennisP 9y ago
        Exactly, that one's aneutronic.
        ZenoArrow 9y ago
        It can use both D-T and pB11. The majority of tests to date have been with D-T. It's best not to confuse the reactor design with the target fuel.
        DennisP 9y ago
        Actually pure deuterium in the tests so far. But the design explicitly targets boron fusion. If it were adapted to D-T, then it would in fact need a steam turbine. The great thing about pB11 is that most of the energy output is fast-moving charged particles, and that's not true of D-T.
  - IntronExon 9y ago
    Sputtering of the metal in the reactor due to fast neutrons, leading to embrittlement, large amounts of neutron activated waste, and an unreasonable maintenance schedule. Lost energy due to said neutrons. Plus all of what has been said before me by DennisP.
- simias 9y ago
  >1. ITER is a political beast, not a practical one. Can you expand on that? Whose politics and for what purpose?
  - IntronExon 9y ago
    Sure, these are two good sources for thst. In depth: http://ffden-2.phys.uaf.edu/feb.7.03.pdf http://ffden-2.phys.uaf.edu/feb.7.03.pdf Light: https://www.nytimes.com/2017/03/27/science/fusion-power-plant-iter-france.html https://www.nytimes.com/2017/03/27/science/fusion-power-plan... In short it was originally Cold War politics to get some international cooperation going around energy. Over time it morphed into a pissing/$$$ contest between nations involved.
  - loopbit 9y ago
    I'm not an expert and all of this happened almost 20 years ago, but I do remember avidly reading about ITER around 2001-2003, when the project was more or less agreed between all parties. Wikipedia[0] has a short paragraph on the history and I'd say that any project involving so many different countries and governments (and bureaucracies!) is a red-tape hell. Just choosing the location was a long and difficult process[1]. Canada proposed a location but later pulled out, Japan put a lot of pressure to get it built there. Even within Europe it caused a political "war" between France and Spain, both proposing different sites. First the EU decided to back France and it was finally agreed upon as the final location. From what I remember, many people complained about the decision. Japan felt left out and was offered 20% of the research personnel. The infrastructures in France had to be built from the ground up even before the actual work to build the plant could start, the area didn't even have big enough roads to bring all the materials, which in turn delayed the works and increased the costs (something that wouldn't have happened in Spain or Japan, no idea about Canada). [0] https://en.wikipedia.org/wiki/ITER#Organization_history https://en.wikipedia.org/wiki/ITER#Organization_history [1] https://en.wikipedia.org/wiki/ITER#Location https://en.wikipedia.org/wiki/ITER#Location
- DennisP 9y ago
  There are several aneutronic fusion projects. The biggest is Tri Alpha, with over $500 million invested. They proved stable plasma in 2015 at 10 million degrees, and just completed a new reactor they plan to take to 100 million degrees; according to their model, the plasma should get more stable at higher temperatures. If it turns out that way, pumping up the heating to boron fusion temperatures is relatively simple, and they have a straightforward path to a practical reactor in 2025 or so. (Source: recent articles plus a presentation I saw from one of their people in 2016) There's a group that thinks a sufficiently powerful petawatt picosecond laser could ignite fusion in a block of boron fuel. This will be testable with an off-the-shelf laser before long; these lasers are improving by a factor of ten every three years. YCombinator has an investment in Helion, which is working on a hybrid D-D/D-He3 reactor (the He3 comes from the D-D reactions), saying only 6% of the energy would be released as neutron radiation. Plus there's LPP, a tiny dark-horse project that will finally get a decent test of their idea this year.
  - IntronExon 9y ago
    Like you, I’m hoping for an unforeseen breakthrough, but it seems very unlikely anytime soon. Just containing such an energetic plasma for any length of time past ignition is going to require it’s own set of breakthroughs in materials and manipulation of magnetic fields.
    - DennisP 9y ago
      The temperature would be very high but the amount of energy wouldn't necessarily be remarkable. And for the pulsed designs (laser and LPP), containment time isn't a concern.
maxharris 9y ago
Thanks to some brilliant people, as well as advances in superconducting materials, MIT has a better plan than ITER, for a fraction of the budget: https://www.youtube.com/watch?v=KkpqA8yG9T4 https://www.youtube.com/watch?v=KkpqA8yG9T4
- ZenoArrow 9y ago
  It should be noted that all non-tokamak approaches to nuclear fusion offer cheaper costs than ITER. Tokamaks like ITER are an outlier when it comes to projected/actual costs. I'm going to avoid listing my personal favourite fusion project yet again on HN, but what I will say is ITER is not the only game in town, and there's lots of promising research being done into non-tokamak designs.
  - DennisP 9y ago
    I agree, but will mention that MIT's plan is a tokamak. But it's ten times smaller, because they use new, commercially-available high temperature superconducting tapes that weren't around when ITER was designed. This lets them use a much stronger magnetic field. (MIT doesn't have funding for it but the private company Tokamak Energy is doing something similar.) MIT's design is also easy to take apart, has a 3-d printed reactor core that gets replaced once a year, and surrounds that with FLiBe, the same molten salt used in some fission designs, for cooling and tritium breeding.
    - XorNot 9y ago
      It's also all a design, and you'd be a fool to believe any upfront cost-projections are going to be hit. ITER was not as expensive as it is back when it was designed either, but building things which have never been built before is full of risk. In a sane world, we'd be funding both and probably more, seeing as how fusion power is basically one of the single most important technologies we could perfect in the modern age.
      - DennisP 9y ago
        True, but we have built working tokamaks like JET, which is about the same size as MIT's design, so it's not like the numbers come out of thin air either.
pablobaz 9y ago
ITER is not a power station it is a research project. In other news Wright brothers first plane unsuitable for transatlantic business travel. Colossus struggles to run Call of Duty.
- majewsky 9y ago
  I had that same criticism in my head while reading the first half of the article that mainly discusses the energy use of the project. That criticism was dispelled in the second half where fundamental problems with fusion reactors are discussed. This article will certainly not cause me to dismiss fusion as a potential energy source outright, but it makes it very clear that even with ITER up and running, we are nowhere near being able to use fusion for energy generation. [1] To pick up your analogy: It took only 1-2 decades to get from the Wright brothers to commercial aircrafts. In that sense, ITER looks more akin to one of Da Vinci's flying machines. [1] Which makes me wonder about the buzz about Lockheed-Martin's skunkworks fusion projects.
  - igravious 9y ago
    Hi friend. If I may? If you want to do footnotes on HN you can use the dagger symbol, like so†. Dagger is Unicode U+2020 which you can read about here[1]. Asterisk is out of bounds because it is typographically significant in Markdown and dagger is the next in the order[2]. It's just that conventionally numbers enclosed by [] are used for non-inline hyperlinks (denoting cited references) in order to not clutter up the text with incomprehensible and long URLs. I think this is from text email culture but it could go back even further, I am not sure about that. Apologies for the persnicketiness. [1] http://www.fileformat.info/info/unicode/char/2020/index.htm http://www.fileformat.info/info/unicode/char/2020/index.htm [2] https://en.wikipedia.org/wiki/Note_(typography)#Numbering_and_symbols https://en.wikipedia.org/wiki/Note_(typography)#Numbering_an... † I am a footnote indexed by a dagger. On Linux at least you can enter Unicode by using the key combo Ctlr-Shift-u and keying in the appropriate number. Dagger is easy to remember because it is 2020.
    - Y_Y 9y ago
sddfd 9y ago
ITER is a research project. As with all research, its main purpose is discovery of "incidental knowledge" , i.e. things you did not set out to research, but are important to know for future projects in the direction. ITER has between 1000 and 5000 highly qualified workers and scientists on site. The experience these people gain during the project can hardly be expressed in USD. Furthermore, many companies gain experience in manufacturing parts for ITER - experience that might be useful for other projects in the future too. Lastly, many different players collaborate in ITER so the whole of things is also a management experiment.
- eesmith 9y ago
  This article is a "critique of fusion as an energy source", not a critique of ITER as a research project. Part of its complaint is about the marketing behind ITER, including the "misguided motto" and misleading information about the ITER power balance. It brings up some of the same benefits you did. "[ITER's] most favorable legacy is that, like the International Space Station, it will have set an impressive example of decades-long international cooperation among nations both friendly and semi-hostile." and "A second invaluable role of ITER will be its definitive influence on energy-supply planning."
- bambax 9y ago
  > a management experiment Really? Billions and billions of Euros, hundreds of thousand of metric tons of concrete, hundreds of MW of power, thousands of gallons of fresh water-per-minute, tends of thousands of nuclear waste, to run a "management experiment"?? This has to be the most ludicrous, expensive and ill-designed case study in the history of mankind. Also, it seems the word "experiment" is inadequate. An experiment requires 1/ a theory and an expectation of what should happen 2/ a rigorous, controlled and replicable setup, where you test the theory and compare the actual result to the expected one Having many people from different nations working together does not qualify as a "experiment"; maybe it's an experience, or a happening (like a concert). It's the Woodstock of science; it must certainly be fun for all involved but it's unlikely humanity will make much progress because of it.
  - Jedd 9y ago
PhantomGremlin 9y ago
I'd love to read equivalent commentary about the Lockheed Martin Skunk Works Compact Fusion project. https://www.lockheedmartin.com/us/products/compact-fusion.html https://www.lockheedmartin.com/us/products/compact-fusion.ht... If Compact Fusion is crank science, why is it on Lockheed's website? Potential Applications: safe power for safe seas planes with unlimited range and unmatched endurance power a city of 50,000 people speed up space travel Etc. Is that just bullshit to make shareholders feel better? Or are there realistic alternatives to grandiose megaprojects like ITER?
- simias 9y ago
  Has there been any update since 2014? I'm not holding by breath.
- Nihilartikel 9y ago
  I think there are quite a few credible alt-fusion projects out there. I have no particular animus towards ITER but I am disappointed that its multi-decade scope, megabudget, and sunk cost, are starving research allocation into other approaches. Some of which seem to have clearer paths to energy production than magnetic confinement fusion. My favorite dark horse is the inertial electrostatic confinement Polywell design, which is currently being funded (crumbs in comparison to ITER) by the Navy, who would of course love a submarine sized fusion reactor. https://en.wikipedia.org/wiki/Polywell https://en.wikipedia.org/wiki/Polywell Cool thing with the Polywell - if the physics actually pan out, the full size one should be able to fuse hydrogen and boron, which won't produce neutron radiation, just alpha particles that can be directly converted to electricity.
physicsguy 9y ago
I went to a talk by one of the materials scientists at JET a few years ago, and he stressed that in terms of that part, the biggest challenge is the neutron bombardment of the steel reactor vessel. He was saying that the biggest problem is that current calculations mean you'd need to replace the entire highly irradiated reactor vessel after 3/4/5 years or so because of structural defects and helium bubbles forming in the cracks.
epistasis 9y ago
When you start comparing the $20B cost for ITER versus "big" projects like the human genome, which took $1B over a similar time scale, ITER really does seem like a boondoggle. It's sucking up massive amounts of research dollars and attention on scientific/engineering goals that don't necessarily even move in the right direction! Suppose they succeed; then what? What has been learned that can spur the next $20B? Will the next tokamok take an order of magnitude less money to produce, like the next genome did the day after the first was released? Big projects always use ridiculous and deceptive hype, but "unlimited energy" takes that to a new level, IMHO.
- Jedd 9y ago
  It's possible that your assessment is predicated on the idea that all advancement can be reduced to $-values, normalised over large time spans. My feeling -- undeniably hard to substantiate -- is that this is a difficult position to defend or attack.
  - epistasis 9y ago
    I'm certainly not reducing the advancement to dollar values. In contrast, the expenditure necessarily must be evaluated in money and time, and it's an age old problem: who gets the research allocations? The advancements are often unknown, but even if the genome project hadn't had the side effect of spinning off a massive amount of technology, the end product would be a significant advancement that would have accelerated all manner of biological research. This was biology's biggest project ever, and it is massively dwarfed by the boondoggles of physics. In contrast, what can we learn form ITER? The opportunity for learning seems awfully minimal, and if it does complete its goal, as the article points out, it's not terribly desirable. And if the end goal isn't desirable, where's the side-effects of tech development? There appears to be none. However many contractors get rich. Potential for advancements can be difficult to judge, and the biggest advancements are almost never expected. But when judging must be done about potential, it's typically best done by those close to the field rather than by political means. ITER seems to be more of a political beast [1] than a scientific one. A political beast that has now robbed the field of fusion of $20B of dollars that could have gone to other research projects. (There's significant intra-science politics of course, that greatly steer the course of research as much as scientific goals. However I'd much rather go with those political winds than the politics of nations when deciding research.) [1] https://news.ycombinator.com/item?id=16380390 https://news.ycombinator.com/item?id=16380390
anovikov 9y ago
Who cares? How much were the solar cells on the market recently? 22 cents a watt? Any chance fusion could beat that price, even taken into account solar's dismal load factors?
- okket 9y ago
  But you need access to sunlight for solar panels to work. You can't have that constantly due to the rotation of the earth. Or if you move further away from the sun. Or into the constant shadow of a planet or moon, to shield your scientific equipment from sun radiation. Or even underwater/underground on earth. Etc. So there will always be a use case for fusion energy, if it works.
  - anovikov 9y ago
    True, there may be uses cases for it. But with renewables price collapse in the last 10 years, and very likely further at least 2x price reduction of solar and slight on wind, this is no longer a make or break thing for humanity. More research and commercialization of advanced batteries would further reduce importance of fusion. I am getting to think that fusion has nearly ran out of time. 10 years down the road, potential market for it may become too small to justify the immense expenses - and by that point, 'mainstream' projects are not expected to bear fruit yet.
    - wiz21c 9y ago
      Problem is : even they're cheap, you have to build massive amounts of them... So fusion energy (or anything nuclear) is still required I think. But well, I'm no expert, just repeating what I've heard (check Jean-Marc Jancovici, french guy)
    - acidburnNSA 9y ago
      It's easy to make the leap you're making but while intermittent renewable costs have indeed been falling, they're doing it in an environment with cheap dispatchable backup natural gas. Once you get to around 50% intermittent renewable penetration the storage costs get crazy. In a four-state area in the Pacific Northwest from December 5 to December 15th, all 4 billion watts of wind generation sat dormant because of a wind lull. Batteries for that would cost $90 billion and take up a football field 100 stories tall.
Gravityloss 9y ago
ITER is a very large and tightly packed network of piping, wiring and components. A change to one part will cascade through the system. Having spent some time with computer assisted design systems, I suspect making these changes is a very laborious manual process. Are we software limited here too?
indescions_2018 9y ago
The contrarian opinion is to increase budget 100X, construct Helium-3 mining colonies on the lunar surface, and deploy rocket foundries for missions to Titan and beyond ;) http://www.moonexpress.com/ http://www.moonexpress.com/ Direct Fusion Drive For A Human Mars Orbital Mission https://bp.pppl.gov/pub_report/2014/PPPL-5064.pdf https://bp.pppl.gov/pub_report/2014/PPPL-5064.pdf Future Chinese Lunar Missions https://nssdc.gsfc.nasa.gov/planetary/lunar/cnsa_moon_future.html https://nssdc.gsfc.nasa.gov/planetary/lunar/cnsa_moon_future...
jlebrech 9y ago
Do we really need to build something anymore if we can simulate it in a computer first? How did the simulation go and is it a net gain? I guess if the data varies that it will also help us learn more about physics.
- misnome 9y ago
  Question: How do you know if your simulation is correct?
- xyzzyz 9y ago
  Fluid dynamics is notoriously hard to simulate.
gwbas1c 9y ago
With all the waste heat, you'd think some kind of district heating would be attempted?
- pfdietz 9y ago
  ITER will only be able to run at full power for a few weeks before its materials reach their radiation damage limits, so trying to use its heat output would not make any sense. And unless they can get disruptions well under control on non-tritium shots, it will never be allowed to do any DT runs.
pfdietz 9y ago
ITER is a dead end. It would require multiple miracles to get anything remotely competitive out of it, and that's just not going to happen. First, realize what's holding back nuclear fission. It's not nuclear waste, or safety, or Greenpeace, or the cost of uranium. It's the capital cost of the powerplants, and to a lesser extent the non-fuel operating cost of the plants. Fusion would make these two main problems worse, not better. The fusion power density of ITER's plasma (note: just the plasma, not including the total volume of the reactor) is 0.6 MW/m^3. The power density of the core of a pressurized water reactor, on the other hand, is 100 MW/m^3. And the machinery around a PWR core is much simpler than all the magnets and such around a fusion core. (The ARC reactor has somewhat higher power density, but still sucks compared to a PWR.) That fusion power density sucks has been known for decades. It follows from basic physical considerations (square-cube law) and is largely independent of the details of the reactor design. See Lawrence Lidsky's famous 1983 article in MIT's Technology Review, "The Trouble With Fusion". The inevitable result of this is that fusion power will be more expensive than fission power. The capital cost will be much higher. ITER's high cost has been explained as due to international coordination, but actually the cost was lowballed from the start. If they had gone with cost estimates of previous reactors, and scaled them on the cost/size plot, ITER should have been estimated at 3-4x what they said it would cost. Now, operating cost... Fusion reactors are extremely complex, much more so than fission reactors. This means they will break more often. And all the parts inboard of a DT (or DD, or D3He) reactor's biological shield (in ARC, this includes the magnets) will be too hot for hands-on maintenance. So all the repairs will have to be done by robots. Good luck with that! In a reactor the size of ITER, all the armor on the first wall will have to be water cooled. Each component slab of this armor cannot fail. A single leak in any of them will render the reactor incapable of sustaining a plasma. I understand that when conventional reliability estimation techniques are applied to these things, one concludes the reactor will be able to operate only a few percent of the time. This isn't even good enough for a research reactor, never mind a commercial reactor. And what will fusion face if, by some miracle, a "working" reactor is finally made? In the last 40 years photovoltaic modules have declined in cost by about a factor of 200. What will happen in the next 40 years? ITER, if all its fusion output were converted to electrical power at 40% efficiency (and none needed to be fed back) would cost > $100/W. PV modules are now about a factor o 300x less than that. Even if we divide the PV number by four, ITER is so far out of the running it's absurd. To finish, I'll add that to truly replace fossil fuels, electrical power has to come in cheap enough that resistive heat displaces natural gas. This requires electricity around $0.01/kWh. There appears to be no chance ITER, ARC, or any of the other fusion schemes could approach this, even with hugely optimistic assumptions.

88 points by okket 9y ago

It’s worth making a few notes: 1. ITER is a political beast, not a practical one. 2. These are problems with D-T fusion, not all fusion, although... 3. Aneutronic fusion isn’t even on the remote horizon. Fusion has incredible potential, but probably not for anyone alive today.

ZenoArrow 9y ago
Regarding point 2, the only drawback I'm aware of with D-T fusion is that it still produces nuclear waste (whereas aneutronic fusion produces zero/close to zero nuclear waste), however the nuclear waste produced by D-T fusion is said to be: * Smaller in size than the nuclear waste produced by the current approaches to nuclear fission. * Has a shorter half life than spent nuclear fission fuel, meaning that it doesn't take as long before it's safe to take it out of storage. Aside from nuclear waste, are there other drawbacks to D-T fusion that people should be aware of?
- DennisP 9y ago
  Concern about tritium leaks, the need for tritium breeding which limits how fast you can build new reactors, hard neutron radiation that theoretically could be used to breed fissiles (but only if you have a high-enough tritium breeding ratio to keep the reactor fueled, despite not using all your neutrons for that), and the need for a steam turbine, which means you're not going to achieve the extremely low costs that may be achievable with aneutronic fusion. Which doesn't mean D-T is hopeless, just not as good as aneutronic. I've seen fusion scientists say the waste from D-T reactors would only need to be contained for several decades.
  - ZenoArrow 9y ago
    > "the need for a steam turbine" Where's the steam turbine in the following fusion approach? https://en.wikipedia.org/wiki/Dense_plasma_focus https://en.wikipedia.org/wiki/Dense_plasma_focus https://lppfusion.com/fusion-power/dpf-device/ https://lppfusion.com/fusion-power/dpf-device/
    - DennisP 9y ago
      Exactly, that one's aneutronic.
      - ZenoArrow 9y ago
        It can use both D-T and pB11. The majority of tests to date have been with D-T. It's best not to confuse the reactor design with the target fuel.
        DennisP 9y ago
        Actually pure deuterium in the tests so far. But the design explicitly targets boron fusion. If it were adapted to D-T, then it would in fact need a steam turbine. The great thing about pB11 is that most of the energy output is fast-moving charged particles, and that's not true of D-T.
- IntronExon 9y ago
  Sputtering of the metal in the reactor due to fast neutrons, leading to embrittlement, large amounts of neutron activated waste, and an unreasonable maintenance schedule. Lost energy due to said neutrons. Plus all of what has been said before me by DennisP.
simias 9y ago
>1. ITER is a political beast, not a practical one. Can you expand on that? Whose politics and for what purpose?
- IntronExon 9y ago
  Sure, these are two good sources for thst. In depth: http://ffden-2.phys.uaf.edu/feb.7.03.pdf http://ffden-2.phys.uaf.edu/feb.7.03.pdf Light: https://www.nytimes.com/2017/03/27/science/fusion-power-plant-iter-france.html https://www.nytimes.com/2017/03/27/science/fusion-power-plan... In short it was originally Cold War politics to get some international cooperation going around energy. Over time it morphed into a pissing/$$$ contest between nations involved.
- loopbit 9y ago
  I'm not an expert and all of this happened almost 20 years ago, but I do remember avidly reading about ITER around 2001-2003, when the project was more or less agreed between all parties. Wikipedia[0] has a short paragraph on the history and I'd say that any project involving so many different countries and governments (and bureaucracies!) is a red-tape hell. Just choosing the location was a long and difficult process[1]. Canada proposed a location but later pulled out, Japan put a lot of pressure to get it built there. Even within Europe it caused a political "war" between France and Spain, both proposing different sites. First the EU decided to back France and it was finally agreed upon as the final location. From what I remember, many people complained about the decision. Japan felt left out and was offered 20% of the research personnel. The infrastructures in France had to be built from the ground up even before the actual work to build the plant could start, the area didn't even have big enough roads to bring all the materials, which in turn delayed the works and increased the costs (something that wouldn't have happened in Spain or Japan, no idea about Canada). [0] https://en.wikipedia.org/wiki/ITER#Organization_history https://en.wikipedia.org/wiki/ITER#Organization_history [1] https://en.wikipedia.org/wiki/ITER#Location https://en.wikipedia.org/wiki/ITER#Location
DennisP 9y ago
There are several aneutronic fusion projects. The biggest is Tri Alpha, with over $500 million invested. They proved stable plasma in 2015 at 10 million degrees, and just completed a new reactor they plan to take to 100 million degrees; according to their model, the plasma should get more stable at higher temperatures. If it turns out that way, pumping up the heating to boron fusion temperatures is relatively simple, and they have a straightforward path to a practical reactor in 2025 or so. (Source: recent articles plus a presentation I saw from one of their people in 2016) There's a group that thinks a sufficiently powerful petawatt picosecond laser could ignite fusion in a block of boron fuel. This will be testable with an off-the-shelf laser before long; these lasers are improving by a factor of ten every three years. YCombinator has an investment in Helion, which is working on a hybrid D-D/D-He3 reactor (the He3 comes from the D-D reactions), saying only 6% of the energy would be released as neutron radiation. Plus there's LPP, a tiny dark-horse project that will finally get a decent test of their idea this year.
- IntronExon 9y ago
  Like you, I’m hoping for an unforeseen breakthrough, but it seems very unlikely anytime soon. Just containing such an energetic plasma for any length of time past ignition is going to require it’s own set of breakthroughs in materials and manipulation of magnetic fields.
  - DennisP 9y ago
    The temperature would be very high but the amount of energy wouldn't necessarily be remarkable. And for the pulsed designs (laser and LPP), containment time isn't a concern.

maxharris 9y ago

Thanks to some brilliant people, as well as advances in superconducting materials, MIT has a better plan than ITER, for a fraction of the budget: https://www.youtube.com/watch?v=KkpqA8yG9T4 https://www.youtube.com/watch?v=KkpqA8yG9T4

ZenoArrow 9y ago
It should be noted that all non-tokamak approaches to nuclear fusion offer cheaper costs than ITER. Tokamaks like ITER are an outlier when it comes to projected/actual costs. I'm going to avoid listing my personal favourite fusion project yet again on HN, but what I will say is ITER is not the only game in town, and there's lots of promising research being done into non-tokamak designs.
- DennisP 9y ago
  I agree, but will mention that MIT's plan is a tokamak. But it's ten times smaller, because they use new, commercially-available high temperature superconducting tapes that weren't around when ITER was designed. This lets them use a much stronger magnetic field. (MIT doesn't have funding for it but the private company Tokamak Energy is doing something similar.) MIT's design is also easy to take apart, has a 3-d printed reactor core that gets replaced once a year, and surrounds that with FLiBe, the same molten salt used in some fission designs, for cooling and tritium breeding.
  - XorNot 9y ago
    It's also all a design, and you'd be a fool to believe any upfront cost-projections are going to be hit. ITER was not as expensive as it is back when it was designed either, but building things which have never been built before is full of risk. In a sane world, we'd be funding both and probably more, seeing as how fusion power is basically one of the single most important technologies we could perfect in the modern age.
    - DennisP 9y ago
      True, but we have built working tokamaks like JET, which is about the same size as MIT's design, so it's not like the numbers come out of thin air either.

pablobaz 9y ago

ITER is not a power station it is a research project. In other news Wright brothers first plane unsuitable for transatlantic business travel. Colossus struggles to run Call of Duty.

majewsky 9y ago
I had that same criticism in my head while reading the first half of the article that mainly discusses the energy use of the project. That criticism was dispelled in the second half where fundamental problems with fusion reactors are discussed. This article will certainly not cause me to dismiss fusion as a potential energy source outright, but it makes it very clear that even with ITER up and running, we are nowhere near being able to use fusion for energy generation. [1] To pick up your analogy: It took only 1-2 decades to get from the Wright brothers to commercial aircrafts. In that sense, ITER looks more akin to one of Da Vinci's flying machines. [1] Which makes me wonder about the buzz about Lockheed-Martin's skunkworks fusion projects.
- igravious 9y ago
  Hi friend. If I may? If you want to do footnotes on HN you can use the dagger symbol, like so†. Dagger is Unicode U+2020 which you can read about here[1]. Asterisk is out of bounds because it is typographically significant in Markdown and dagger is the next in the order[2]. It's just that conventionally numbers enclosed by [] are used for non-inline hyperlinks (denoting cited references) in order to not clutter up the text with incomprehensible and long URLs. I think this is from text email culture but it could go back even further, I am not sure about that. Apologies for the persnicketiness. [1] http://www.fileformat.info/info/unicode/char/2020/index.htm http://www.fileformat.info/info/unicode/char/2020/index.htm [2] https://en.wikipedia.org/wiki/Note_(typography)#Numbering_and_symbols https://en.wikipedia.org/wiki/Note_(typography)#Numbering_an... † I am a footnote indexed by a dagger. On Linux at least you can enter Unicode by using the key combo Ctlr-Shift-u and keying in the appropriate number. Dagger is easy to remember because it is 2020.
  - Y_Y 9y ago

sddfd 9y ago

ITER is a research project. As with all research, its main purpose is discovery of "incidental knowledge" , i.e. things you did not set out to research, but are important to know for future projects in the direction. ITER has between 1000 and 5000 highly qualified workers and scientists on site. The experience these people gain during the project can hardly be expressed in USD. Furthermore, many companies gain experience in manufacturing parts for ITER - experience that might be useful for other projects in the future too. Lastly, many different players collaborate in ITER so the whole of things is also a management experiment.

eesmith 9y ago
This article is a "critique of fusion as an energy source", not a critique of ITER as a research project. Part of its complaint is about the marketing behind ITER, including the "misguided motto" and misleading information about the ITER power balance. It brings up some of the same benefits you did. "[ITER's] most favorable legacy is that, like the International Space Station, it will have set an impressive example of decades-long international cooperation among nations both friendly and semi-hostile." and "A second invaluable role of ITER will be its definitive influence on energy-supply planning."
bambax 9y ago
> a management experiment Really? Billions and billions of Euros, hundreds of thousand of metric tons of concrete, hundreds of MW of power, thousands of gallons of fresh water-per-minute, tends of thousands of nuclear waste, to run a "management experiment"?? This has to be the most ludicrous, expensive and ill-designed case study in the history of mankind. Also, it seems the word "experiment" is inadequate. An experiment requires 1/ a theory and an expectation of what should happen 2/ a rigorous, controlled and replicable setup, where you test the theory and compare the actual result to the expected one Having many people from different nations working together does not qualify as a "experiment"; maybe it's an experience, or a happening (like a concert). It's the Woodstock of science; it must certainly be fun for all involved but it's unlikely humanity will make much progress because of it.
- Jedd 9y ago

PhantomGremlin 9y ago

I'd love to read equivalent commentary about the Lockheed Martin Skunk Works Compact Fusion project. https://www.lockheedmartin.com/us/products/compact-fusion.html https://www.lockheedmartin.com/us/products/compact-fusion.ht... If Compact Fusion is crank science, why is it on Lockheed's website? Potential Applications: safe power for safe seas planes with unlimited range and unmatched endurance power a city of 50,000 people speed up space travel Etc. Is that just bullshit to make shareholders feel better? Or are there realistic alternatives to grandiose megaprojects like ITER?

simias 9y ago
Has there been any update since 2014? I'm not holding by breath.
Nihilartikel 9y ago
I think there are quite a few credible alt-fusion projects out there. I have no particular animus towards ITER but I am disappointed that its multi-decade scope, megabudget, and sunk cost, are starving research allocation into other approaches. Some of which seem to have clearer paths to energy production than magnetic confinement fusion. My favorite dark horse is the inertial electrostatic confinement Polywell design, which is currently being funded (crumbs in comparison to ITER) by the Navy, who would of course love a submarine sized fusion reactor. https://en.wikipedia.org/wiki/Polywell https://en.wikipedia.org/wiki/Polywell Cool thing with the Polywell - if the physics actually pan out, the full size one should be able to fuse hydrogen and boron, which won't produce neutron radiation, just alpha particles that can be directly converted to electricity.

physicsguy 9y ago

I went to a talk by one of the materials scientists at JET a few years ago, and he stressed that in terms of that part, the biggest challenge is the neutron bombardment of the steel reactor vessel. He was saying that the biggest problem is that current calculations mean you'd need to replace the entire highly irradiated reactor vessel after 3/4/5 years or so because of structural defects and helium bubbles forming in the cracks.

epistasis 9y ago

When you start comparing the $20B cost for ITER versus "big" projects like the human genome, which took $1B over a similar time scale, ITER really does seem like a boondoggle. It's sucking up massive amounts of research dollars and attention on scientific/engineering goals that don't necessarily even move in the right direction! Suppose they succeed; then what? What has been learned that can spur the next $20B? Will the next tokamok take an order of magnitude less money to produce, like the next genome did the day after the first was released? Big projects always use ridiculous and deceptive hype, but "unlimited energy" takes that to a new level, IMHO.

Jedd 9y ago
It's possible that your assessment is predicated on the idea that all advancement can be reduced to $-values, normalised over large time spans. My feeling -- undeniably hard to substantiate -- is that this is a difficult position to defend or attack.
- epistasis 9y ago
  I'm certainly not reducing the advancement to dollar values. In contrast, the expenditure necessarily must be evaluated in money and time, and it's an age old problem: who gets the research allocations? The advancements are often unknown, but even if the genome project hadn't had the side effect of spinning off a massive amount of technology, the end product would be a significant advancement that would have accelerated all manner of biological research. This was biology's biggest project ever, and it is massively dwarfed by the boondoggles of physics. In contrast, what can we learn form ITER? The opportunity for learning seems awfully minimal, and if it does complete its goal, as the article points out, it's not terribly desirable. And if the end goal isn't desirable, where's the side-effects of tech development? There appears to be none. However many contractors get rich. Potential for advancements can be difficult to judge, and the biggest advancements are almost never expected. But when judging must be done about potential, it's typically best done by those close to the field rather than by political means. ITER seems to be more of a political beast [1] than a scientific one. A political beast that has now robbed the field of fusion of $20B of dollars that could have gone to other research projects. (There's significant intra-science politics of course, that greatly steer the course of research as much as scientific goals. However I'd much rather go with those political winds than the politics of nations when deciding research.) [1] https://news.ycombinator.com/item?id=16380390 https://news.ycombinator.com/item?id=16380390

anovikov 9y ago

Who cares? How much were the solar cells on the market recently? 22 cents a watt? Any chance fusion could beat that price, even taken into account solar's dismal load factors?

okket 9y ago
But you need access to sunlight for solar panels to work. You can't have that constantly due to the rotation of the earth. Or if you move further away from the sun. Or into the constant shadow of a planet or moon, to shield your scientific equipment from sun radiation. Or even underwater/underground on earth. Etc. So there will always be a use case for fusion energy, if it works.
- anovikov 9y ago
  True, there may be uses cases for it. But with renewables price collapse in the last 10 years, and very likely further at least 2x price reduction of solar and slight on wind, this is no longer a make or break thing for humanity. More research and commercialization of advanced batteries would further reduce importance of fusion. I am getting to think that fusion has nearly ran out of time. 10 years down the road, potential market for it may become too small to justify the immense expenses - and by that point, 'mainstream' projects are not expected to bear fruit yet.
  - wiz21c 9y ago
    Problem is : even they're cheap, you have to build massive amounts of them... So fusion energy (or anything nuclear) is still required I think. But well, I'm no expert, just repeating what I've heard (check Jean-Marc Jancovici, french guy)
  - acidburnNSA 9y ago
    It's easy to make the leap you're making but while intermittent renewable costs have indeed been falling, they're doing it in an environment with cheap dispatchable backup natural gas. Once you get to around 50% intermittent renewable penetration the storage costs get crazy. In a four-state area in the Pacific Northwest from December 5 to December 15th, all 4 billion watts of wind generation sat dormant because of a wind lull. Batteries for that would cost $90 billion and take up a football field 100 stories tall.

Gravityloss 9y ago

ITER is a very large and tightly packed network of piping, wiring and components. A change to one part will cascade through the system. Having spent some time with computer assisted design systems, I suspect making these changes is a very laborious manual process. Are we software limited here too?

indescions_2018 9y ago

The contrarian opinion is to increase budget 100X, construct Helium-3 mining colonies on the lunar surface, and deploy rocket foundries for missions to Titan and beyond ;) http://www.moonexpress.com/ http://www.moonexpress.com/ Direct Fusion Drive For A Human Mars Orbital Mission https://bp.pppl.gov/pub_report/2014/PPPL-5064.pdf https://bp.pppl.gov/pub_report/2014/PPPL-5064.pdf Future Chinese Lunar Missions https://nssdc.gsfc.nasa.gov/planetary/lunar/cnsa_moon_future.html https://nssdc.gsfc.nasa.gov/planetary/lunar/cnsa_moon_future...

jlebrech 9y ago

Do we really need to build something anymore if we can simulate it in a computer first? How did the simulation go and is it a net gain? I guess if the data varies that it will also help us learn more about physics.

misnome 9y ago
Question: How do you know if your simulation is correct?
xyzzyz 9y ago
Fluid dynamics is notoriously hard to simulate.

gwbas1c 9y ago

With all the waste heat, you'd think some kind of district heating would be attempted?

pfdietz 9y ago
ITER will only be able to run at full power for a few weeks before its materials reach their radiation damage limits, so trying to use its heat output would not make any sense. And unless they can get disruptions well under control on non-tritium shots, it will never be allowed to do any DT runs.

pfdietz 9y ago

ITER is a dead end. It would require multiple miracles to get anything remotely competitive out of it, and that's just not going to happen. First, realize what's holding back nuclear fission. It's not nuclear waste, or safety, or Greenpeace, or the cost of uranium. It's the capital cost of the powerplants, and to a lesser extent the non-fuel operating cost of the plants. Fusion would make these two main problems worse, not better. The fusion power density of ITER's plasma (note: just the plasma, not including the total volume of the reactor) is 0.6 MW/m^3. The power density of the core of a pressurized water reactor, on the other hand, is 100 MW/m^3. And the machinery around a PWR core is much simpler than all the magnets and such around a fusion core. (The ARC reactor has somewhat higher power density, but still sucks compared to a PWR.) That fusion power density sucks has been known for decades. It follows from basic physical considerations (square-cube law) and is largely independent of the details of the reactor design. See Lawrence Lidsky's famous 1983 article in MIT's Technology Review, "The Trouble With Fusion". The inevitable result of this is that fusion power will be more expensive than fission power. The capital cost will be much higher. ITER's high cost has been explained as due to international coordination, but actually the cost was lowballed from the start. If they had gone with cost estimates of previous reactors, and scaled them on the cost/size plot, ITER should have been estimated at 3-4x what they said it would cost. Now, operating cost... Fusion reactors are extremely complex, much more so than fission reactors. This means they will break more often. And all the parts inboard of a DT (or DD, or D3He) reactor's biological shield (in ARC, this includes the magnets) will be too hot for hands-on maintenance. So all the repairs will have to be done by robots. Good luck with that! In a reactor the size of ITER, all the armor on the first wall will have to be water cooled. Each component slab of this armor cannot fail. A single leak in any of them will render the reactor incapable of sustaining a plasma. I understand that when conventional reliability estimation techniques are applied to these things, one concludes the reactor will be able to operate only a few percent of the time. This isn't even good enough for a research reactor, never mind a commercial reactor. And what will fusion face if, by some miracle, a "working" reactor is finally made? In the last 40 years photovoltaic modules have declined in cost by about a factor of 200. What will happen in the next 40 years? ITER, if all its fusion output were converted to electrical power at 40% efficiency (and none needed to be fed back) would cost > $100/W. PV modules are now about a factor o 300x less than that. Even if we divide the PV number by four, ITER is so far out of the running it's absurd. To finish, I'll add that to truly replace fossil fuels, electrical power has to come in cheap enough that resistive heat displaces natural gas. This requires electricity around $0.01/kWh. There appears to be no chance ITER, ARC, or any of the other fusion schemes could approach this, even with hugely optimistic assumptions.