Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Whats the holdup with Fusion Power?

  1. Nov 1, 2007 #1
    You got temperature and/or pressure, and one simple atom to play with. What's the big deal?
  2. jcsd
  3. Nov 1, 2007 #2
    Unfortunately, it takes many many atoms to produce useful amounts of energy. Now you have a volume of superheated plasma. Now the problems begin.
  4. Nov 2, 2007 #3


    User Avatar

    Staff: Mentor

    As MaWM indicated it's many ionized atoms (free nuclei and electrons) magnetically confined in a plasma. The plasma is loosing energy very rapidly due to phenomena like brehmsstrahlung and cyclotron radiation, while nuclei scatter more often than they fuse.

    The plasma densities are on the order of 1014 (nuclei and electrons)/cm3. The densities are limited by pressure which is constrained by the achievable magnetic fields and strength of the structure supporting the magnets.
  5. Nov 2, 2007 #4


    User Avatar
    Science Advisor


    "You got temperature and/or pressure". Unfortunately you need BOTH simultaneously, AND for
    a long enough time.

    Additionally, as was pointed out; you have to do this for a LOT of atoms. When you try to do that -
    you get some very wicked instabilities that wreak havoc with your attempts at fusion.

    As Astronuc pointed out, the densities that have been realized so far within the constraints of present
    magnetic confinement technology are on the order of 1014 atoms per cc.

    In any other field - something with that density would be considered a very good VACUUM!!!

    Controlled thermonuclear fusion is NOT an easy problem to solve AT ALL

    Dr. Gregory Greenman
  6. Nov 2, 2007 #5

    D H

    Staff: Mentor

    You can say just about the same for any complex and hard-to-solve technical or social challenge we as a society confront. Example: What's the holdup with solar energy? You got photons streaming in for free and lots of ways of harnessing them. What's the big deal?

    What you did in your initial post was to trivialize the problem by whitewashing over all of the huge challenges noted by Astronuc and Morbius.
  7. Nov 2, 2007 #6


    User Avatar
    Gold Member

    Confinement time ala Larsen would apply to confinement approaches, inertial or magnetic. Confinement time does not seem to apply to any of the several beam - beam approaches (e.g. IEC). That is, there's no intention to do ignition; they are purely 'driven' schemes. - Not that IEC has shown any possibility of power production
  8. Nov 2, 2007 #7


    User Avatar
    Science Advisor


    The designs for the NIF - the National Ignition Facility are intended to "do ignition".

    LLNL has developed the "Fast Ignitor" concept, and it is also being explored by
    the University of Rochester on the Omega laser:






    Dr. Gregory Greenman
  9. Nov 3, 2007 #8


    User Avatar
    Science Advisor
    Homework Helper


    As they have pointed out here, we have big difficulties with making enoguh nucleis fusion in order to gain energy.

    It is a HUGE difference between fusing 2 protons in a collider, and having 10^30 confined in a quite small region and trying to both make them fusion, continue to fusion and also to extract the energy from there.

    I hope you understand this.
  10. Nov 3, 2007 #9


    User Avatar
    Gold Member

    Yes, certainly. I meant IEC does not require ignition for net power, though it has many other problems.
  11. Nov 3, 2007 #10
    Yes, but it's hard to imagine making a practical fusion reactor with ICF, in large part because of the required repetition rate. Ignition may be achieved at NIF (that alone will be extremely challenging), but those target shots will be at best once every couple weeks - that's a very long way from firing it at 5-10 Hz, which is what you want for a practical reactor. If you scale up the energy you can get away with less-frequent bursts of energy, but that's going in the direction of a nuclear bomb, not a reactor.

    Maybe these problems will eventually be solved, but probably not in our lifetimes. Doesn't mean we shouldn't work on them, but anyone looking for a quick solution to fusion energy production will be disappointed.
  12. Nov 3, 2007 #11


    User Avatar
    Science Advisor


    We're ALREADY working on the reactor designs!!

    There are MANY designs ALREADY on the drawing boards.

    Nuclear designers can do an awful lot of work while they are waiting for the laser to be built.

    You don't have to do the designs in the order you envision. You can do A LOT of the design
    work even before you have the proof of principle. We know what the output product of a successful
    fusion reaction is - so you can take the design from there. You don't have to wait for the proof
    of principle shot.

    Dr. Gregory Greenman
  13. Nov 3, 2007 #12
    Yes, I know people are thinking about reactors - there are some interesting concepts, and some interesting work going on for example at General Atomics. But it's not just the laser (firing multimegajoule shots at 10 Hz from hundreds of separate beams, each with very fine tolerances on timing, pulse shaping, pointing, etc.), it's the stream of targets - all currently conceived (even with fast ignition) as having more-or-less beautifully smooth DT ice layers that take a day to craft, and are extremely difficult to adequately characterize. They have to be fired somehow into the reactor at 10 Hz, aimed with micron precision over meter distances, in a manner that doesn't ruin the ice.

    And at this point we don't really even understand the requirements on an ignitable target, all we have are simulation predictions that (based on long history) will almost certainly turn out to be wrong in significant ways, and we'll have to sort out how to do it properly by performing many experiments. We may learn, for example, that we really need 5-10 times more laser energy, reducing the rep-rate requirements but making much bigger individual bangs that will have a significant impact on reactor designs.

    I'm not pessimistic in the long term, clever people can figure these things out given enough time and money, but it certainly won't happen soon.
  14. Nov 4, 2007 #13


    User Avatar
    Science Advisor


    ALL of your concerns HAVE been addressed. Those involved have been designing and testing such devices. After all, you don't have to
    wait for a shot to reach ignition before you design, test, and certify equipment that can position tartgets with the required precision at 10 Hz.
    [ I believe current designs may go to 20 Hz ].

    You are wrong again here. We DO understand - not just in simulations - but from experiment what
    the requirements of an ignitable target are. We used the one driver that we KNOW works:


    Courtesy of Princeton University; read the following on page 198:

    "However, in its FY 1987 report, the House Science and Technology Committee disclosed that
    explosions at the Nevada Test Site have provided data for the laser fusion program. In a program
    called Centurion-Halite, the intense radiation from undergound nuclear explosions have apparently
    been used in attempts to ignite ICF targets, ultimately to help design pellets that could be ignited
    with currently available drivers, thus circumventing the need for the next-generation short-pulse
    laser at a price of almost one billion dollars"

    As the above indicates; there was no need to build another experimental laser bigger than Nova in
    order to find out what conditions were necessary for ignition. This allowed the program to skip
    building another experimental laser that could not reach ignition - and jump directly to the scale of
    the NIF - National Ignition Facility.

    That's why NIF is called the National Ignition Facility; because it was designed knowing
    what conditions it needed to produce.

    Dr. Gregory Greenman
    Last edited: Nov 4, 2007
  15. Nov 4, 2007 #14
    I assure you that they have not been. Particularly the target problem, which I am intimately involved with. We struggle to make one suitable target, and there is no one working in the field, knowledgeable about the process, who is seriously thinking about making them at 10 Hz - there are only wild, utterly untested concepts from outsiders. Fast ignition might relax some of the target constraints, but that's a concept in it's infancy - we don't even know yet how we'll deliver the spark energy to the implosion.

    You sound like a designer. :smile: This is very naive, because Halite/Centurion experiments used a multi-terrajoule driver (a bomb), not a megajoule laser - you can afford to be sloppy when you have so much energy available. There were other very key target differences, perhaps (?) you are aware of them. Those experiments demonstrated the basic concept, but cannot tell us whether or not NIF will succeed at it's mission - or how our concept of an ignitable target will evolve as we learn more. We know how to ignite a target with terrajoules, but not with megajoules except through simulations - and long experience shows that every time we make a leap forward, we discover how much important physics is missing from the simulations.

    One key uncertainty that was never addressed with Halite/Centurion is laser-plasma interactions, which scatter laser light out of the target, drive plasma waves that create preheating electrons, steer beams away from their aim points, and cross-couple energy, degrading symmetry. We are very unsure what will happen in a NIF ignition target, and this uncertainty drives a large part of the uncertainty in what an ignitable target looks like for NIF - if it's worse than we think, we'll have to use more laser energy to drive a larger target with less power, but that tradeoff only goes so far (unless you're driving things with a nuclear bomb). At some point you run out of available laser energy. You can win some of it back by going to longer laser wavelengths (second harmonic vs. third, you can run with more energy because final optics damage is less severe), but that will make the laser-plasma problem worse (things tend to scale as intensity*wavelength^2).

    Another uncertainty is 3D plasma hydrodynamics. 3D simulations are very slow and expensive even on the fastest modern computers, so essentially all target design requirements are based on 2D simulations. These miss inherently 3D effects, and are themselves based on physics that is known to be incomplete, for example in the treatment of radiation transport and high-mode instability growth.

    A good review of the program from two years ago, discussing these issues and many more, is here: http://www.fas.org/irp/agency/dod/jason/nif.pdf

    I wouldn't be working in the field if I didn't think it was worthwhile, but people need to appreciate that it's difficult and uncertain, and commit to it over the long haul.
  16. Nov 5, 2007 #15


    User Avatar
    Science Advisor


    For Heaven's sake - use your BRAIN!!!

    When you have more than enough energy - but you are interested in what can be done with lesser
    amounts - you can throw a bunch of the excess energy away!!!

    FYI - I'm NOT a designer - I'm a code developer.

    Dr. Gregory Greenman
  17. Nov 5, 2007 #16


    User Avatar
    Gold Member

    From the referenced Jason's report:
    Care to comment? Did NIF implement any of the Jason report's recommendations?
  18. Nov 5, 2007 #17
    Between this and your mildly amusing private messages, I think I am done discussing this with you. :smile: I will point out, however, that having worked in ICF/NIF target experiments at LLNL for the last 15 years, and having attended the Jason review as well as having made some of the material that was presented, I've never heard of you. Perhaps if we run into each other sometime, we can discuss further over coffee.
  19. Nov 5, 2007 #18
    Yes, many of them.
  20. Nov 5, 2007 #19
    I missed this, it must have been edited out. I don't care about the FAS site, the link was to the Jason report. You appear to not know who I am, either. :smile:
  21. Nov 10, 2007 #20
    Does confinement of the 'Polywell' type for fusion have a future?

    I have seen videos of attempts (from Google Tech Talk lectures). But, there are hundreds of 'videos', and I find it hard to separate the deluded and downright fraudulent from the genuinely feasible.

    The work of the late Robert Bussard comes across to me as being the real deal.

    I am less sure of the scheme promoted by Eric Lerner where he proposes a small pulsed device (340Hz or so) where the resulting intense ion beam energy is collected from a transformer device.

    Here in UK we have had about 30 years of fine university minds working on Tokamak -type fusion devices at Culham (like JET Joint European Torus) and now there is to be a bigger effort in France (ITER). There seems no great impetus for any other scheme.

    I would welcome some informed discussion on which schemes have substantial credibility in this community.
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Whats the holdup with Fusion Power?
  1. Fusion for power? (Replies: 35)

  2. Fusion power (Replies: 2)