Skip to main content

Fusion Finally on the Horizon?

If you're betting that we will one day rely on clean nuclear fusion for our growing energy needs, your odds are getting a lot better thanks to a laser-based technique called Inertial Confinement Fusion (ICF).

For many people, nuclear fusion is the ultimate alternative to dirty, planet-destroying fossil fuels and tsunami-vulnerable fission plants. The fact that we can look to the sun to see fusion in action makes it seem so easy, safe, and wholesome. There's just one small problem - stars can only shine if they're really, really big because you need lots of gravity to hold the churning mass of hydrogen and helium together tightly enough for fusion to take place.

Because we can't rely on gravity to squeeze fusion fuel together down here on Earth, physicists have been working on a number of alternatives. For the past half century or so most of the focus has been on using magnetic fields to confine gas-like plasmas of super hot hydrogen. Unfortunately, plasmas are slippery and holding onto them for long is very, very difficult. So difficult, in fact, that the approach has been the butt of a longstanding joke that goes something like this, "Practical plasma fusion is only twenty years away, and always will be."

It's funny, 'cause it's true. When I was a physics student in the 1980's, the plasma physicists I was working for told me we'd have magnetic fusion reactors in twenty years, and they are saying the same thing today. (Some are even saying it may take thirty years, essentially indicating that the field is making anti-progress towards its lofty goal.)

In the meantime, an entirely different approach appears to be moving ahead at leaps and bounds. The National Ignition Facility (NIF) seems to be on track to burn a bit of fusion fuel within the year, and produce megawatts of power for the electrical grid in about ten years, using Inertial Confinement Fusion.

Most ICF designs rely on heating a capsule of fuel with a whole mess of high power lasers. That causes  the outer shell of the capsule to explode and compress a tiny bit of fuel inside. Once compressed, the hot fuel tends to rapidly expand, but before it can do that it has to overcome inertia, as described by Newton's first law.  That doesn't give us a lot of time, but it's enough to create a minuscule version of the sun, even without massive amounts of gravity.

All you need to do at that point is harness the energy from the tiny, fleeting sun to heat water and use the resulting steam to turn a few turbines. Yeah, yeah, it's clearly not so easy. Still 60 years of work has proven magnetically confined fusion to be durn near impossible. That's why I'd consider doubling down on the ICF folks, while taking my chips off the magnetic fusion hand.

Of course, the cynic in me says that in ten years, we could still be saying that practical ICF is ten, or even twenty, years away. That would result in another sad, and unfunny, joke.

Personally, I think we're on the right track this time. We may be laughing about ICF in a decade, but I'm pretty sure we'll be laughing all the way to the bank.


  1. The odds get even better with particle accelerator technology instead of lasers:

  2. 60 years of poorly-funded, disorganized work into magnetically confined fusion has in fact proven it to be quite feasible, despite the limited resources allocated to the research and constant threat of projects being canceled. The JT-60 research reactor achieved conditions that would have produced net gain in D-T plasma back in 1998 (as a plasma confinement research reactor, tritium is not used in JT-60), and ITER is expected to produce 500 MW of output power from 50 MW input, with first plasma starting in 2019 and fusion experiments starting in 2026.

    The NIF has to cool its lasers and optical systems for hours between is hoped to eventually achieve 700 shots (with ~10 mg of fuel each) a year. This is nowhere near what's needed for power production, even if each shot somehow manages to produce more power than it consumes. Working against that is the fact that the losses in the laser system are enormous, and the needed lasers will never be highly efficient. ICF is nowhere close to the current state of magneticallly confined fusion and nowhere close to catching up, it certainly isn't going to be producing grid power in 10 years.

  3. Cjames, you may be right, but thirty years of waiting for magnetic confinement to work has made me weary (at least the thirty years since I started paying attention, it was already 30-40 years old by then). The NIF is ripping right along in comparison.

    In general, holding things in magnetic bottles is a hard thing to do, but inertia is comparatively simple, both in concept and in practice. Given a difficult solution and an easy one, I'll take the easy one every time.

  4. Actually Buzz thats not correct. TFTR produced 10 MW of fusion energy in 1994. Inertial confinement fusion has yet to produce ANY fusion energy. So your logic that NIF is making progress in leaps and bounds is not correct. Yea they can build a giant 2 MJ laser for billions of dollars. Big deal! If I had a billion dollars I could get a Q>1 on JET just by buying a shit load of neutral beams. Magnetic confinement is a proven technilogy, ICF is not. In fact NIF is just a nuclear weapons program posing as a fusion experiment.

  5. Galan, 10 MW of power in 1994, and still no power plant 18 years later? ICF has really only begun in earnest since the development of high power lasers in the 1970's, while magnetic confinement started in the 1940's. So ICF seems to be progressing steadily to me, while magnetic confinement is still crawling along.

    Magnetic confinement requires dealing with lots of complex problems, while ICF is more brute force. One approach is a balancing act, while the other relies on a great big hammer. We may be getting better at the balancing act, but it's so much easier to buy a bigger hammer. ICF benefits from the fact that laser technology, which is really all there is to ICF (to oversimplify things a tad), keeps improving. Magnetic confinement has no similar enabling technology, as far as I know, so we have to endlessly struggle with the same issues that have been there for decades.

  6. ICF has been sputtering along since the 1970s, it's not the new, rapidly advancing approach you paint it as. Nova was built in 1984 specifically with the purpose of achieving ignition, and failed to do so. Even looking just at the NIF, construction started in 1997 and experienced delay after delay.

    And inertia is simple, but nothing else about laser based ICF is. Lasers and the frequency conversion/pulse forming optics required are inefficient and produce tremendous amounts of waste heat, and the shots need to be extremely precisely timed and aligned. Suggestions for ICF fusion plants assume shot rates of 1-10 Hz...the NIF at its eventual best will take *hours* to cool down and ready for the next shot. Lasers have indeed advanced greatly, but there's no reason to think this will continue to happen or that they'll ever be efficient enough for ICF to be's betting on further revolutionary improvements in lasers. In comparison, only a relative few magnetic confinement experiments have even used superconducting magnets. No further "enabling technology" is needed (if there was one, it was the development of computers and models sophisticated enough for the problem), it's largely a matter of building a few large research reactors that nobody has been willing to fund.

  7. There is good reason to think that lasers will continue to advance rapidly - that reason is that the technology has advanced rapidly since lasers were invented. There is no reason to expect that to come to an end soon.

    It is in part because there is no enabling technology for magnetic confinement that I have lost my taste for it. If we could say "the only thing standing between us and magnetic fusion is X, and a reasonable projection is that X will be sufficiently advanced by Y date," then I might get back on board.

    The thing is, magnetic confinement is basically an effort to duplicate the sun, except that we're can't use one of the main factors that lets the sun burn - extraordinarily high gravitation. While for inertial confinement, we're duplicating hydrogen bombs on small scale, which we have known how to build for as long as we've been struggling with magnetic confinement.

    1. > There is good reason to think that lasers will continue to advance rapidly

      No, there isn't. We may very well have already achieved a large fraction of the maximum performance that's physically possible. There's room for improvements, certainly, but very possibly not the orders of magnitude of improvements needed. I don't think you appreciate how far NIF has to go...we're talking something like a 20000-200000 times (that's 20-200 thousand times) increase in firing rate, and equipment that can sustain firing for years at a time without downtime.

      > If we could say "the only thing standing between us and magnetic fusion is X, and a reasonable projection is that X will be sufficiently advanced by Y date," then I might get back on board.

      We *can* say that. The major remaining roadblocks research into materials that can withstand the neutron exposure and funds and the will to build large reactors. ITER finally has the latter and will be doing the former. Once again, we are currently constructing a tokamak system that will achieve first plasma 7 years from now and start fusion experiments 7 years later, designed to produce 500 MW of fusion power. ICF is nowhere close to attempting such a feat.

      > While for inertial confinement, we're duplicating hydrogen bombs on small scale, which we have known how to build for as long as we've been struggling with magnetic confinement.

      This comparison is completely absurd. The huge optical apparatus and fuel pellets used for NIF-style ICF are nothing remotely like anything in a hydrogen bomb, and magnetically confined fusion doesn't even approximate the conditions inside stars, or make any attempt to do so.

  8. This comment has been removed by the author.

  9. So, the limiting technology in developing tokamaks is . . . wait for it . . . tokamaks. Well, we've plodded along for 60 years, so maybe we'll be there in another 60. Except the funding isn't available because I'm not the only person who's tired of it.

    Magnetic confinement is a steady state approach (like stars) and inertial confinement is not (that is, it's like bombs). Magnetic fields simply suck at replacing gravity. But the basic physics of making small explosions and big explosions with fusion are the same, even if the technologies that create fusion conditions are different.

    1. No, the limiting technology is materials, which a tokamak is being built in large part to research. The need for materials that can handle the neutron exposure is not unique to magnetic confinement, they are a requirement for every other D-T fusion system, including ICF.

      And you are once again ignoring the fact that a 500 MW tokamak reactor is under construction...not being talked about being built in 60 years, they're building it RIGHT NOW, with first plasma expected in 7 years and fusion 7 years after that. We're nowhere close to even beginning to design a ICF system that can do what we're building ITER to do.

      The superficial resemblance of ICF to fusion bombs is completely irrelevant. The facts are that ICF is vastly more complicated than you claim, and magnetically confined fusion vastly more successful and further along than you seem willing to admit.

    2. I'm saying that magnetic fields are a lousy way to hold hot, slippery plasma for extended times. On the other hand we've been using a variety of ICF to create fusion (i.e. fusion bombs) for as long as magnetic confinement research has been crawling along.

      You say MCF will give us fusion in 14 years. According to the article linked above "LIFE director Mike Dunne says that the capital costs for the pilot plant would be about $4 billion, and it could be putting hundreds of megawatts into the grid by the early 2020s." That would put ICF well ahead of MCF, even though it got underway decades after the first magnetic confinement machines were built.

      You may may not believe Dunne, but that means your argument is with him, not with me.

      In any case, I'm on the ICF bandwagon because I feel the technological challenges are more tractable than with MCF.

    3. Once again, the tokamak approach has *already* given us fusion. Why do you keep pretending that tokamaks have failed when they have not only demonstrated the ability to confine plasma under conditions that will produce fusion, but have produced megawatts of fusion power?

    4. Sure, tokomaks have produced fusion, but on the road to a power plant you say "we are currently constructing a tokamak system that will achieve first plasma 7 years from now and start fusion experiments 7 years later." That still puts you as much as a decade behind ICF, which Dunne says will be putting power into the grid by the early 2020's.

  10. @cjameshuff wrote multiple times that the NIF concept will never work for fusion power generation because NIF lasers can't fire quickly enough.

    This is true, insofar as the NIF lasers were never designed to fire more than a few times a day. They use passive cooling and flashlamps to pump the lasers, both of which means you must wait for hours between shots (at least if you fire all of the beams.

    However, NIF was never designed for power generation. It was designed to STUDY the physics involved and for that it only needs to fire a few times a day, while being able to collect a ton of data from each shot, so that they can analyze the results and experiment with different configurations.

    The project to commercialize the NIF technique is not NIF itself, but rather the LIFE project.

    Unlike NIF it uses a laser design designed to operate at 10hz using diodes instead of flash lamps, with active cooling across the beamline, as demonstrated with the Mercury laser, which was developed to validate the basic technologies needed for LIFE.

    In principle, there aren't any basic technological barriers to deploying the LIFE concept on an industrial scale, other than waiting on NIF to demonstrate the basic concept by achieving energy gain using their ICF concept.

  11. Bubble-confined Sonoluminescent-laser Fusion (BSF) combines ideas from laser Inertial Confinement Fusion (ICF), sonofusion, and piezoelectric energy harvesting. The resulting power plant has: extremely high power generating capacity, exceptional fuel containment, inexpensive and fabrication-free targets that self-heat to low ignition temperatures that make higher gains with larger burn-up fractions possible, very efficient energy conversion, a higher tritium breeding ratio, and no activation issues. More info is available at

  12. (part 2)

    7.Attaining tritium self-sufficiency might be fusion’s most difficult challenge. There are no practical, external sources of tritium, so fusion plants must breed their own; current inventories are extracted from heavy-water reactors, which produce 1.7 kg/year, and this supply will peak around 2025 at a mere 27 kg, enough to run a 1 GW fusion plant for six months. The main source of tritium is expected to come from breeding by capture of fusion neutrons in lithium contained in a blanket surrounding the fusion core. Because any lost neutrons would result in a tritium-breeding ratio less than 1.0, a neutron multiplier is usually employed with the hope of overcoming the negative effects of having a front wall ahead of the breeding blanket. Ironically, for an unlimited fuel source, fuel supplies (short-term) will determine how quickly fusion plants can be brought online and how effective fusion can be toward addressing our current worldwide energy crisis. Two key parameters, that influence how long it takes to produce enough tritium for a subsequent plant’s start-up are the fractional burn-up rate (low burn-up fractions require extra fuel-cycles, leading to higher retention-times and greater fuel losses through beta decay) and the trapped (inside the blanket) inventory size.
    8.Because laser-compressed ICF capsules obtain temperatures & pressures existing in the cores of stars, they also obtain stellar pressures. Without the weight of an entire star to confine it, ICF plasma disperses rapidly (~0.1 ns) into the vacuum, so that only a small fraction of the fuel gets burnt.
    9.A large portion of the laser energy is wasted, backscatter and bremsstrahlung.
    10.The ICF rocket compression scheme is not very inefficient (5% - 15%); most of the energy is carried away from the target area by high-energy outgoing ablation material. The peak efficiency of an ablation-driven rocket is typically a factor of 4 or more smaller than that of an ideal rocket because the exhaust is continually heated by the incident flux driving the implosion.

  13. (part 3)
    11.High-energy x-ray measurements indicate that up to 50% of the absorbed laser light ends up in hot electrons. The presence of these high-energy electrons, which generally have a temperature of 50-60 keV, make it difficult to achieve the high density compression that is required for a successful burn.
    12.Most mainline systems (except for liquid-metal-wall ICF reactors, such as HYLIFE) have steel first walls, which are necessary to maintain a good quality vacuum and to endure the intense x-ray and neutron radiation. The first walls of all such reactors will be highly radioactive (2 to 5 billion curies). In addition, these first walls will require replacement every few years because of neutron-induced damage, either from helium embrittlement or from atomic displacements. Because both neutron energy and neutron population are reduced in the steel first walls of these reactors, neutron multipliers (such as lead or beryllium) or isotopic enrichment of Li-6 are usually required to achieve acceptable tritium breeding ratios. The same applies to magnetic fusion reactor chamber walls. For example, the STARFIRE tokamak walls will have a radioactivity of more than 5 billion curies and must be replaced every four or five years. The significance of this should not be ignored, chamber walls exposed to damage rates of 35 dpa/yr (displacements per atom per year) will require replacement every 5-7 years. Assuming that only the inner structural walls need to be replaced at 30% of the original reactor vessel cost, then about 5% of the plants lifetime must be devoted to replacement activities.
    13.ICF laser firing times need to be increased. NIF requires a timeout for cool-down and recovery after each firing. The high precision laser optics that ICF uses must cool for several hours between firings to recover from thermal expansion.
    14.Cooling the laser medium is very inefficient. NIF uses external flash tubes that create significant amounts of non-recoverable low level waste heat which must actively be removed (requiring extra energy) from the gain medium.
    15.The Halite-Centurian tests in Nevada apparently showed the DT targets might require up to 20 MJ to ignite. Current ICF designs produce less than 3 MJ.

  14. (part 1)
    Here are a few obstacles that laser ICF power plants still need to overcome:
    1.Manufacturing the cryogenic fuel capsules requires extremely high-quality surfaces, down to the atomic level. This is prohibitively expensive, especially for a power plant that fires several capsules per second.
    2.Not only do these BB size targets have to be tracked as they fly through the reactor, but all of the laser beams have to hit them midair from several meters away, a task that for direct-drive ICF requires 50 micron accuracy.
    3.The blast chamber must be evacuated several times per second, between shots, to prevent interference with subsequent laser shots.
    4.The optics (and walls) must be protected or they will vaporize and fail structurally due to the extreme thermal impulse stresses that result from the intense x-ray, 14 MeV neutrons, and 3.5 MeV alphas. The biggest problem for ICF’s final optics is that there is no scheme yet proposed for either Direct Drive or Indirect Drive that has complete credibility. Optics protection is still one of the weak areas for laser driven ICF.
    5.In order to reach the gains necessary for a commercial power plant, self-heating of the fuel by 3.5 MeV alphas is required. Unfortunately, because ICF capsules are small, the alphas escape from the burn zone before depositing their energy. This problem cannot be solved by simply making the capsules bigger, because the blast-chamber constraints would be exceeded by the larger yield.
    6.A major obstacle that ICF capsules encounter is turbulent mixing that can quench the fuel and prevent ignition. Rayleigh-Taylor Instabilities (RTI) arise in situations where low-density fluids push into high-density fluids. ICF capsules are vulnerable to this type of instability twice during their compression, 1st at the start, when low-density, high-pressure plasma pushes the higher-density tamper material inward, and 2nd when the implosion stagnates and high-pressure fuel pushes the higher-density tamper material outward. In 2009, W.J. Nellis of Harvard University wrote a critique of the National Ignition Fusion program. The article was titled “Will NIF work?” It contained the following quotes, “Despite the financial and human resources and time spent on NIF, the key condensed matter and materials physics issues of the fuel capsule remain unsolved and the R-T instability continues to be the limiting feature of NIF performance.” “While both the physics of R-T growth and the equations of state of DT (hydrogen) and shell material must be known, the R-T instability is by far the major issue…” “It is R-T spikes that grow from such R-M instabilities under high accelerations at later times that are the show-stopper of ICF.” “Computational simulations for more than 35 years have provided no insight into eliminating the R-T instability.” If large pellets could be imploded with increased laser energy, then the thickness of the mixed region and loss of the hot-spark region would become less serious, but a method to achieve such an implosion must first be found.


Post a Comment

Popular Posts

How 4,000 Physicists Gave a Vegas Casino its Worst Week Ever

What happens when several thousand distinguished physicists, researchers, and students descend on the nation’s gambling capital for a conference? The answer is "a bad week for the casino"—but you'd never guess why.

Ask a Physicist: Phone Flash Sharpie Shock!

Lexie and Xavier, from Orlando, FL want to know: "What's going on in this video ? Our science teacher claims that the pain comes from a small electrical shock, but we believe that this is due to the absorption of light. Please help us resolve this dispute!"

The Science of Ice Cream: Part One

Even though it's been a warm couple of months already, it's officially summer. A delicious, science-filled way to beat the heat? Making homemade ice cream. (We've since updated this article to include the science behind vegan ice cream. To learn more about ice cream science, check out The Science of Ice Cream, Redux ) Image Credit: St0rmz via Flickr Over at Physics@Home there's an easy recipe for homemade ice cream. But what kind of milk should you use to make ice cream? And do you really need to chill the ice cream base before making it? Why do ice cream recipes always call for salt on ice?