Please create an account to participate in the Slashdot moderation system

 



Forgot your password?
typodupeerror
×
Power Earth The Almighty Buck Science Technology

MIT Plans To Build Nuclear Fusion Plant By 2033 170

Mallory Locklear reports via Engadget: MIT announced yesterday that it and Commonwealth Fusion Systems -- an MIT spinoff -- are working on a project that aims to make harvesting energy from nuclear fusion a reality within the next 15 years. The ultimate goal is to develop a 200-megawatt power plant. MIT also announced that Italian energy firm ENI has invested $50 million towards the project, $30 million of which will be applied to research and development at MIT over the next three years. MIT and CFS plan to use newly available superconducting materials to develop large electromagnets that can produce fields four-times stronger than any being used now. The stronger magnetic fields will allow for more power to be generated resulting in, importantly, positive net energy. The method will hopefully allow for cheaper and smaller reactors. The research team aims to develop a prototype reactor within the next 10 years, followed by a 200-megawatt pilot power plant.
This discussion has been archived. No new comments can be posted.

MIT Plans To Build Nuclear Fusion Plant By 2033

Comments Filter:
  • by Ritz_Just_Ritz ( 883997 ) on Friday March 09, 2018 @09:15PM (#56236927)

    A functional fusion reactor that's commercially viable seems to be perpetually 15-20 years in the future.

    Next......

    • Re:LOL (Score:4, Insightful)

      by sheramil ( 921315 ) on Friday March 09, 2018 @10:00PM (#56237103)

      The big money isn't in power from nuclear fusion, it's in research towards nuclear fusion.

    • A functional fusion reactor that's commercially viable seems to be perpetually 15-20 years in the future.

      Next......

      Kinda like real AI (vs. AI lite)... which one will come first, true AI, Nuclear fusion, flying cars, or hover-boards....

      • Kinda like real AI (vs. AI lite)... which one will come first, true AI, Nuclear fusion, flying cars, or hover-boards....

        1. Flying cars already exist. You can't buy one because they don't make economic sense.
        2. Nuclear fusion will come next. It is just an engineering problem.
        3. Hard AI is not yet on the horizon. We don't know how to achieve it.
        4. A hoverboard, like in "Back to the future 2" violates the known laws of physics. That will have to wait until we discover that dark matter emits an anti-gravity force. This discovery will also explain why dark matter was so hard to find. Physicists were looking for somethin

        • Dark matter was theorized to account for missing mass. If it pushes rather than pulls then it is not the missing mass. You may be thinking of dark energy...

      • Where there is lots of data, and clearly defined objectives, AI systems based on deep neural networks and reinforcement learning can usually at least equal human performance (often surpass it) today. However, this is only narrow AI. Allowing AI systems to deal with less well defined objectives, and being able to use experience in other domains to guide behavior in novel domains where limited data is available, will need further breakthroughs. It is difficult to say how long this will take, but recent progre

        • Any one claiming to have AI should be sued for false and deceptive advertising.

        • Present day AI is based on technology and concepts from the 1960s and 1970s. The only thing that changed which make it now more viable are: Computing power, advances in parallelism, faster algorithms and the availability of digital data. Beside that it is 50 year old tech.

          • With respect, I do not think you have any conception of the developments of the last few years. For example, Generative Adversarial Networks, and especially the Deep Convolutional Generative Adversarial Networks of the last 18 months or so, have really revolutionized the field. If you believe this just to be a tweak of the algorithms of 20, 30 or 40 years ago, you are simply wrong.

            • by Bengie ( 1121981 )

              If you believe this just to be a tweak of the algorithms of 20, 30 or 40 years ago, you are simply wrong.

              That's the thing about problem solving. Implemented solutions may look completely different, even if they're fundamentally the same. There's a near limitless number of ways to implement a solution, but still be the same solution. Do not conflate new implementations for new solutions.

              I don't know enough on the topic, but my brother is going into AI and he has taken a similar position that all of these "new" AIs is just 60s/70s computer science used in ways that only recently becoming feasible. They could

        • "Where there is lots of data, and clearly defined objectives, AI systems based on deep neural networks and reinforcement learning can usually at least equal human performance (often surpass it) today."

          Lots of data and clearly defined objectives? Yeah, those are called computer programs.
    • Was there even a single fusion plant project announced before this? Is there one which was announced and should have been active by this year (i.e., which failed to deliver)?

    • In the 70s it was always 40 or 50 years in the future In the 90s and 00s this dropped to about 30 years in the future
      Seems like people may be shifting their goals and expectations to about 20 years in the future now.
      I would not be surprised that this decrease is due to people expecting things to happen more quickly now---not because of the technological advancements occurring in this field.
      Also, let's face it, the only fusion-derived electricity sold to consumers currently living will be from solar panels

    • People like you were saying the same thing about airplanes and rockets. Thank goodness the inventors kept trying.

    • by Maritz ( 1829006 )

      A functional fusion reactor that's commercially viable seems to be perpetually 15-20 years in the future.

      Next......

      I was wondering which slashdot genius would be making this incredibly facile point, and how long it would take.

      Pretty much the first real comment. Thanks for your profound insights. Everybody has heard that line a hundred times. Now, fuck off.

  • by Anonymous Coward

    Call it " hydrogen fusion" or people might confuse it with "nuclear" and that negative associated sentiment that goes with it

  • It's been "in 20 years" for 40 years so the slope is 1 year every 8 years. That means we should take 120 years for the remaining "in 15 years".

    Now note that world total liquids will peak in the next decade and so will coal probably. In 15 years you won't even recognize the world economy.

  • by Sgs-Cruz ( 526085 ) on Friday March 09, 2018 @09:38PM (#56237025) Homepage Journal
    FYI, the CEO of this new company (Bob Mumgaard) and CTO (Dan Brunner) helped answer the questions asked in the Ask MIT Fusion Researchers About Fusion Power in April 2012: https://hardware.slashdot.org/... [slashdot.org] Prof. Dennis Whyte and Dr. Martin Greenwald were also on that thread and are now core members of the founding team of the new startup (although they remain employed by MIT).
  • by elrous0 ( 869638 ) on Friday March 09, 2018 @09:42PM (#56237039)

    The secret to getting away with a bullshit promise is to set the promised payoff so far in the future that you'll be long gone by the time people realize you were full of shit. See every American President for the last 50 years who's promised we'll put a man on Mars just 30 years after he leaves office.

    • by Tablizer ( 95088 )

      The secret to getting away with a bullshit promise is to set the promised payoff so far in the future that you'll be long gone by the time people realize you were full of shit.

      This is often why budget deficits and pension problems happen: those who negotiate the deals will only be around to get the upsides of their deals. They sell off the future to get benefits during their reign. Those deal-makers are usually long-gone when the monster awakes.

    • by Jeremi ( 14640 )

      OTOH, the secret to never accomplishing anything is to stop trying.

      • by Kjella ( 173770 ) on Saturday March 10, 2018 @11:45AM (#56239165) Homepage

        OTOH, the secret to never accomplishing anything is to stop trying.

        Yeah, even though it's ridiculously hard I think research into high energy power is essential, you can always say we should become greener and smarter but in the end physics dictate that it takes a certain amount of power to drive all the household appliances. Sure for a CPU/GPU you can improve calculations/watt but for a water boiler it takes 1 calorie of energy to heat 1g of water 1C and if you can change that you'll have a closet full of Nobel prizes. If we want to give ~10 billion people a modern standard of living we need energy. If we want to start a Mars colony we need energy. If we want to explore the universe we need energy. I don't know how feasible it is to make a miniature sun here on earth, but it's one helluva power source. It's the kind of thing it's probably worth mastering even if it takes us 100 years or 1000 years. I'll admit I'd like to see results a little sooner, but it's like the people researching longevity and immortality. For humanity it looks like a smart topic of research even if it won't arrive in time to save my ass.

        Of course you will always have speculative and sham research looking for grants. You will always have dead ends and people beating a dead horse. But I feel pretty confident that these researchers believe in what they're doing and is making an honest attempt. There's a helluva lot of medical researchers trying to find the cure for cancer, many of them won't achieve much at all. But I think the vast majority is genuinely trying. Comparing them to a politician posturing for his reputation while not realistically even beginning to fund the necessary programs is grossly unfair.

  • Not completely silly (Score:4, Interesting)

    by joe_frisch ( 1366229 ) on Friday March 09, 2018 @10:13PM (#56237143)

    If they can use higher magnetic fields, that increases the pressure and decreases required volume of the reactor to get to breakeven.

    That said, the picture the show looks really small even with high field magnets .

    We'll see. There have been a lot of claims of practical fusion in the next few years. So far non have worked, but its not fundamentally impossible.

    • by quanminoan ( 812306 ) on Friday March 09, 2018 @11:46PM (#56237461)

      A lot of great fusion designs rely on very high fields ~20 T or larger, this has been known for a while. The superconducting technology is now just getting there so some exciting possibilities are becoming realities. Still, not a walk in the park designing large magnets with high temperature superconductor (HTS). HTS joining of cables (splicing) is very tricky as many are powder-in-tube, so for various reasons an internal splice in a solenoid is a trick (that I have not seen demonstrated). It can be figured out though. Right now the biggest hindrance is cost, most HTS requires silver in the powder tube for chemistry reasons, making the cost very high - thousands of USD per meter.

      As other posters have mentioned high magnetic field allows reducing the volume, higher densities, maybe even newer modes. As an engineer however one thing I always see in many of these new designs is a lack of respect for radiation damage on the superconductors; they can't handle high radiation so while it's tempting to put them as close to the plasma region for increased densities etc., you need a reasonable balance. The lockheed design was very guilty of this.

    • by HiThere ( 15173 )

      It's not completely silly, but I'm not going to believe any promises until they have a pilot plant working, and then I'll be a bit dubious. I understand that it's a hard problem, and they currently have a few new tools, and slightly altered approaches, but this has happened before.

      That said, we won't be able to determine the downsides until they have it working. I hope they do, because that's a crucial piece of kit for building long-term space habitats...especially ones that move out past Jupiter's orbit.

  • I'm not sure fusion will ever be economical even if we get it working. Fusion has to compete against direct conversion technologies, where energy is directly converted from its original form into electricity.

    Solar voltaic converts light energy directly into electricity. Wind turbines turn energy from moving air directly into electricity. Gas turbines burn natural gas directly in turbines that generate electricity.

    Most fusion reactions create a lot of their energy in the form of fast neutrons, whose energy can't be converted to electricity directly, but must instead be used to heat up steam, and the hot steam then is used to turn turbines and generate electricity. This is indirect conversion, and the argument I've heard is that steam conversion plants cost more all by themselves than many direct conversion technologies do--therefore fusion reactions that generate the bulk of energy in fast neutrons will be uneconomical by comparison.

    Coal plants too, incidentally--there's a reason no new coal plants are being built in the USA--they're uneconomical compared to natural gas turbine generation. And fusion plants will be extremely capital intensive.

    Furthermore, plasmas in thermal equilibrium that produce energy in charged particles instead of neutrons (which would allow for direct conversion), cool off faster via Bremsstrahlung radiation than they self-heat from their own fusion reactions. So direct conversion from fusion would have to come from nonequilibrium plasmas. And nonequilibrium plasmas are really, really unstable--they tend to thermalize very, very fast.

    Bottom line, I'm not optimistic about terrestrial fusion in any form being economical when it has to compete with solar, wind, and natural gas. Leave planet Earth and go past the orbit of say, Jupiter, and I could see it being a good solution way out there.

    • I agree fully with your remarks. There seems no prospect that fusion is going produce cost-competitive energy.

      However I would like to see them build a practical (though very expensive) power plant. That will be an important step for developing fusion for where will really need it some day in deep space. Not until sometime next century at the earliest - but we can make progress in that direction now.

      • by AJWM ( 19027 )

        Hopefully before next century, but yes. Past a certain distance from the sun (somewhere around the asteroid belt, or a bit further), solar cells don't do much for you, and uranium or thorium are hard to come by. Lots of deuterium-laden hydrogen, however.

        If (when?) we get to the point of manufacturing things as complex as fusion reactors in space, we can go interstellar -- one Oort cloud object at a time, like the polynesian islanders in their canoes.

        (Yeah, I'd rather have warp drives, but working fusion

        • Um, how does "fusion power" translate into interstellar travel? We can't travel between stars, no matter what fuel source you use.
    • I agree, the MILLIONS they are wasting on this could be put better spent on the California high speed train which is now pushing a 70 BILLION price tag /s

    • by Anonymous Coward

      You're only considering the efficiency factor, not the amount of energy produced. If I had a 100% efficient solar panel (impossible) and another source that was 1% efficient but producing 101 times the raw energy, it would still be more net energy.

    • by WaffleMonster ( 969671 ) on Saturday March 10, 2018 @12:29AM (#56237561)

      Fusion has to compete against direct conversion technologies, where energy is directly converted from its original form into electricity.

      Solar voltaic converts light energy directly into electricity. Wind turbines turn energy from moving air directly into electricity. Gas turbines burn natural gas directly in turbines that generate electricity.

      Of these only natural gas is base load and cheap gas can't last forever. It would be necessary to factor in necessary investments in storage/conversion and transmission to compare the true overall cost of each option.

      • You can create your base load with renewables most of the time and only crank up your gas turbines when you have a shortage. That's somewhat expensive but uses little nat gas and creates little CO2 on average.

      • by Uberbah ( 647458 )

        Of these only natural gas is base load

        Zombie talking point. Wind and solar power generation would be spaced across a grid - same as coal and nuclear are. Excess power may be saved via a pumped storage [wikipedia.org] facility like the Ludington plant in Michigan - which is used to back up a nuclear power plant - to be used when needed.

        • by Agripa ( 139780 )

          Zombie talking point. Wind and solar power generation would be spaced across a grid - same as coal and nuclear are. Excess power may be saved via a pumped storage [wikipedia.org] facility like the Ludington plant in Michigan - which is used to back up a nuclear power plant - to be used when needed.

          The amount of available and potential pumped storage capacity is woefully insufficient for this. That leaves much more expensive secondary and flow batteries.

          • by Uberbah ( 647458 )

            The amount of available and potential pumped storage capacity is woefully insufficient for this.

            As opposed to new nuclear power plants that just magically sprout from the ground in the morning???

            • by Agripa ( 139780 )

              The amount of available and potential pumped storage capacity is woefully insufficient for this.

              As opposed to new nuclear power plants that just magically sprout from the ground in the morning???

              You can build a nuclear power plant anywhere although they are better where cooling water is available. Pumped storage is limited to a fraction of hydroelectric sites.

              • by Uberbah ( 647458 )

                Pumped storage is limited to a fraction of hydroelectric sites.

                Ludington is an artificial reservoir. And even in the most arid of climates, pumped storage power could be built into existing water towers, or new ones constructed for that purpose. We have water towers and hydroelectric dams in use today that are over a century old, so for an up-front investment you can have infrastructure that will last a very long time. An investment nuke fans can't really complain about when a new nuke plant will cost t

      • Chances are that by the 2030s, renewable electricity will be so cheap that electrolytic hydrogen to be burned in CCGT plants could be a competitive means of large-scale storage.
      • Of these only natural gas is base load

        You are stuck in the past [skepticalscience.com], please join us in the present, as we work towards the future.

    • In most fussion reactor designs, the neutrons are soaked up by lithium which then undergoes fission to produce heat and fuel. So they are really hybrid fussion/fission reactors.

      • Worth elaborating that these are just designs, no one has built a lithium blanket reactor since all fusion experiments are not yet made for power generation. DEMO will very likely have a lithium blanket, slated for design/construction after ITER. Also, not a hybrid reactor - lithium doesn't split and produce energy (elements up to iron only give energy by fusion, after by fission), and it's not a chain reaction. The reason for using lithium is mainly to capture fast neutrons, and get decay products that pro

    • Direct or indirect conversion is irrelevant. What matters (from an energy standpoint) is the efficiency.
      • Steam turbines have a maximum efficiency of about 40%-43%. In real-world use, they are typically about 33% efficient (2/3rds of the energy becomes waste heat).
      • Photovoltaic conversion is been pushed up to about 44% in the lab. But PV cells commonly in commercial production are about 14%-19% efficient [wikipedia.org]. The rest of the sunlight is either reflected or heats up the PV cell.
      • Wind turbines have a theoreti [wikipedia.org]
      • I was making an argument on based on economics. For that, efficiency is irrelevant. What matters (from an economics standpoint) is how much profit on energy sold we deliver per dollar invested. Even a 95% efficient energy conversion process, that doesn't require ANY inputs whatsoever, is going to be uneconomical if the capital cost is so high that you can't pay interest on the capital invested with the power you sell.

        That's EXACTLY what I'm saying is going to be the case with fusion. The alternatives wi

    • by swell ( 195815 )

      uneconomical for who?

      The existing economy favors monopolies. All the talk lately about the imbalance between rich & poor relates to power, control and dependence. Ownership of the electric grid or the oil supply or the communication network is the source of power and wealth. Dependence on them is the source of weakness and poverty.

      A successful fusion design will help consolidate energy distribution in a few very wealthy hands. It may be able to fight off wind & solar competitors with the 'right' leg

    • For a scathing commentary on fusion and why it may never be commercially viable, read "The Trouble with Fusion" by Lidsky (1985 MIT Tech Review Article).

      http://orcutt.net/weblog/wp-co... [orcutt.net]

      IIRC he was in charge of MIT's fusion program and got booted for this article and his views. He's absolutely correct though. We need to rethink our approach and tokamaks may never compete with fission.

  • Fusion is the power source of the future, and always will be.

    Joking aside, I don't really care if they actually get a net positive reactor going, we need to keep pursuing these reactors. Eventually we will get one to work, even if it's as ugly as the matchup between deep blue and Gary Kasparov.
  • Just like carbon capture, by the time they achieve useful power generation with fusion - if ever - there will be no need for it. Solar, wind and tidal generation with battery storage will be much cheaper and less polluting.
    • by Jeremi ( 14640 )

      It might be just the thing to power your interstellar spacecraft, though.

    • by HiThere ( 15173 )

      There are definitely applications for which it will be unsuited. But there are important applications for which it would be extremely desirable. We often *could* make fission work (though not in all of them), but that has it's own problems, and refueling is a major risk.

      That said, we won't know the downsides of this approach to fusion until afterwards. But at least the fuel should be readily available. (It will need processing, but that's not new. It won't require finding bodies of ore. Water and meth

      • by ceoyoyo ( 59147 )

        Generating electricity might be a niche use case for fusion. A source of controllable (and self-sustaining) plasma could be very handy for, example, refining ore.

        • by HiThere ( 15173 )

          This is just my guess, but I don't think the plasma will be controllable in environments where purity cannot be guaranteed. So you'd do better to generate the electricity, and then use the electricity to generate the processing plasma.

          And a lot is going to depend on how small (and cheap) the smallest practical fusion reactor is. Are we talking about a homestead, a village, a town, a city, a metropolitan area, or what? And it will be a long time after the first fusion reactor is built before we really hav

  • ...given that fusion is always 20 years away...

    • by eagl ( 86459 )

      In the last 50 years, it's gone from being 20 yrs away to now being only 15 years away. Assuming it's not on a log curve, we'll see practical fusion power generation in another 150 years.

  • Economically practical fusion power plants have been 15 years away...for over 50 years now...
    • by ceoyoyo ( 59147 )

      Economically practical fusion plants were designed in the fifties. Problem is, nobody wanted to actually build one. They wanted them small, with no proliferation risks, and neighbour friendly.

  • Seems that fusion reactor design is a big complex task that these 'AI' neural nets could crunch away on til they figure out a really complex optimum design. Think what kind of crazy stellarator design skynet could think up... Might be too hard to actually construct, but I bet it works on paper (RAM?), lol.

  • by jmccue ( 834797 ) on Friday March 09, 2018 @11:05PM (#56237323) Homepage

    I would like nothing better than seeing this built, if even next door to me. But in the Boston Area ? Lots of Luck.

    You cannot even build a Dog House in that area without the following:

    1. Multiple studies on how it will impact the neighborhood.
    2. Protesters showing up at the town meetings, and you have to defend your dog house hundreds of times.
    3. Fighting with various politicians.
    4. Getting all kinds of subpoenas arriving at your door step in their pretty colored envelopes.
    5. At least 1 court appearance, lawyers will be happy.
    6. If you are lucky you hit the jackpot. Your dog house will show up as a ballot question which at best will be ignored by the politicians, or more than likely the politicians will decide to do the exact opposite.
    7. I will not even mention the cost overruns

    So maybe in 200 years you will see it build :(

  • Yet another tokamak (Score:5, Interesting)

    by WaffleMonster ( 969671 ) on Saturday March 10, 2018 @12:03AM (#56237509)

    They should have built a stellarator or literally anything other than a tokamak.

    How many tens of billions have been invested in Tokamaks thus far with very little to show for it? Other approaches consistently get shafted for serious funding due to dogma/politics and risk aversion.

    Comparatively peanuts have been spent on stellarators to date and they have already demonstrated far better results than any tokamak ever has.

    • Most of the standard plasma fusion machines can work if scaled to large enough sizes. Do stellerators scale better than tokomaks? Its a detailed technical question that probably depends on a variety of design constraints.

      I assume ITER went with a Tokomak because it scaled to smaller sizes than any of the other configurations based on the technological constraints at the time. If the technology has changed, maybe there is a better optimum, but I haven't seen a good technical comparison .

    • Have to agree with this, from what little I know, it's always the tokamak design, why keep trying it without success?

      I was reading up on ITER just last week, that thing is incredible expensive, incredible late and not going to produce power.
      The even bigger even more expensive one, they hope to get some power from and the one AFTER that might be commercial.

      Bear in mind the article I read said that Ronald Reagan and bloody Gorbachev are the two who signed off on it, to give you an idea of how long ITER is tak

  • I remember as a very young kid, seeing reports that we'd have fusion reactors in 15 years.

    That was in 1960.

    I'd like to be optimistic, really, but it's a bit of a challenge.

    • The tokamak doesn't seem to work - it certainly hasn't so far. There are many other approaches that cold do with throwing a few million or billion at. We seem to have beaten tokamaks to death without success. Yes, it's theoretically possible. But many things work in theory, just not in practice.

      • by Anonymous Coward

        Well the science disagrees. Tokamaks are so far the only fusion
        devices that have come close to break-even and have produced the
        densest, hottest and most long lasting fusion plasmas. The scientists
        also have used Tokamaks extensively to understand plasma instabilities.
        The only big problem that has slowed fusion power development
        has always been the lack of serious funding. ITER is a big gamble. But
        it will tell us at least whether the Tokamaks will work as power production
        devices.

      • Fusion isn't really mysterious. You need to keep the plasma above some temperature at a sufficiently high product of pressure and lifetime. This improves as the machine gets larger.

        Different field configurations (tokomak, stellerator etc) can provide different scaling of storage time relative to size. At the time ITER was designed, Tokomak looked the best. I don't know if any new information has changed that.

        ITER has been slow / expensive because it is a giant multi-national collaboration, IMHO the le

        • Fusion may not be mysterious, the physics is not that hard in theory, but the practice seems to elude us.
          Except for big giant world destroying bombs, of course. We can do that.

          • I think its more expensive than it is hard. I expect most of the standard fusion devices would work if scaled up - the trick is finding one that can be made at a reasonable size (with ITER maybe not qualifying as "reasonable").

  • Fusion has been just 15 years away ever since man first noticed that the sun was actually on fire. Just ask any true sci-fi author or science/technology magazine editor.

  • Looks like MIT and the fusion gang have understood that the clock is ticking and therefore the 50 year promise does no longer work. Unfortunately, it will not help even if they where able to pull it off because by 2033, we must be half through with our climate change adaptation. 2045 we are either done or finished.

  • by bigtreeman ( 565428 ) <treecolin@gDALImail.com minus painter> on Saturday March 10, 2018 @06:30AM (#56238229)

    We've already got a really big nuclear reactor.
    It's called the sun.
    It distributes power with light which we can safely convert to electricity with solar panels.
    About 150M Kms is about as close as we should get to a nuclear reaction of any type.

    • by quenda ( 644621 )

      You know how many toxic chemicals and bad vibes go into making those solar panels man?
      Natures solar panels are called trees, and that's all we need.
      Wood still burns when the sun does not shine, no need for toxic batteries.
       

  • by account_deleted ( 4530225 ) on Saturday March 10, 2018 @12:20PM (#56239339)
    Comment removed based on user account deletion
    • by novakyu ( 636495 )

      I see that you have never tried to build a self-contained survival bunker that can last for at least three years.

Top Ten Things Overheard At The ANSI C Draft Committee Meetings: (10) Sorry, but that's too useful.

Working...