Hacker News Reader

Comment by sam | Hacker News Reader

It is an important milestone. But to have a commercially viable fusion reactor, you'll need a factor of 50-100 more energy out than in to make up for inefficiencies in electricity generation using this kind of scheme.

The real story here is that this facility allows the US to do nuclear weapons research without violating the nuclear test ban treaty. If the goal was to develop a commercially viable fusion reactor, the $3,500,000,000 spent so far could have been put into projects geared towards small scale fusion experiments investigating novel confinement schemes.

Like the one I was working on, http://meetings.aps.org/Meeting/DPP07/Event/71002 whose funding has since been cut and has been mothballed.

View on HN

Finally someone that's actually in the fusion field. Could you say whether or not the NIF research represents a practical path towards a fusion based power source? My understanding is that it would at least be capable of capturing some much needed high pressure empirical data necessary for validating theory.

On a side note concerning the NIF funding, the US gov will spend huge sums of money studying Nuclear weapons (including their disposal) regardless of the NIF project. If it is at all possible to combine that research with other fields then really it's a perk. I'm not justifying the expenditure, simply attempting to frame the situation in a different light.

Finally, as a fellow scientist with funding woes, I feel your pain.

I've worked in ICF for a nearly a decade. I think NIF research is a necessary step to a ICF based power source. If NIF demonstrates ignition, it can be used to validate our physics models of ICF (turbulence, fusion product transport, thermal conduction, etc). These physics questions are shared by all ICF approaches. However, a fusion power plant will use entirely different laser technology [1], and most likely a different target design [2]. After showing that the physics works, there is still a lot of engineering work needed before building a reactor.

[1] An ICF power plant will be pulsed at 1-10 Hz. NIF is a flash-lamp pumped glass laser, which takes ~12 hours to cool between shots. A power plant would likely be diode-pumped sold state laser since these can meet the required repitition-rate.

[2] The indirect drive target that LLNL is pursuing on NIF is not very efficient. You spend a lot of energy heating the hohlraum. Directly driven targets (blast the capsule directly rather than heating a gold can to make x-rays) should be much more efficient. There are also several ideas for ways to ignite a target more efficiently (shock ignition, fast ignition), but these need additional laser hardware.

@sam Though I work in ICF, I was sad to see the innovative confinement concepts (magnetic confinement) cut a couple years ago. I think it is short sighted. Fusion need to work and get smaller and we should keep our options open. Hope you managed to get a thesis out before the walls fell.

I thought the goal was to have a self sustaining reaction that would only require laser pulses to hold it in place. Is a continuously pulsed laser necessary to provide thermal energy to drive the reaction?

Also I've read (sorry no citation) that the indirect drive was necessary to ensure even dispersal of the heat generated by the initiating laser blast.

Any clarification would be much appreciated, this isn't my field.

ICF is intrinsically pulsed (magnetic confinement like ITER is completely different). You use a laser to spherically compress a 1 mm diameter spherical capsule. The capsule implodes, stagnates, and blows (releasing energy in the explosion). Then you do do it again 0.1 seconds later.

NIF uses an "indirect drive" design. Instead of directly illuminating the spherical target with a bunch of lasers, you blast the inner surface of a gold cylinder with the shell at the center of the cylinder. The cylinder gets hot, emits x-rays, which are absorbed by the capsule. The x-ray drive tends to be "smoother" than direct illumination.

The big problem with ICF is hydrodynamic stability. It is like trying to squeeze a water balloon with your fingers. If you don't squeeze it perfectly symmetrically, it will squirt through your fingers and pop rather than getting compressed by a factor of 20.

>The big problem with ICF is hydrodynamic stability. It is like trying to squeeze a water balloon with your fingers. If you don't squeeze it perfectly symmetrically, it will squirt through your fingers and pop rather than getting compressed by a factor of 20.

Well said.

I agree with your general philosophy towards ICF.

You get the same problem with laser ignition, funnily enough.

My layman's understanding of the subject is that you want (and get) anything but a self-sustaining reaction with fusion, and that's actually one of its safety features compared to fission (where a chain reaction is allowed to happen in a controlled way -- if something goes wrong the reaction keeps going, out of control).

You do want to use the energy from one reaction to power the next, but not directly. No one's actually trying to create a star on the surface of the earth.

As with you, though, not my field so someone who knows better can feel free to correct me.

"one of its safety features compared to fission"

Not exactly. Off the top of my head I can't think of a reactor thats unstable in operation. Hands off (like that TV series life after humans where the people all disappear) almost all stabilize and eventually shutdown on their own. Maybe those crazy graphite reactors are not inherently stable.

There's two linked issues.

Fission reactors get about 10% of their heat from decay products (the "waste") decaying away. That means there is no instant off switch. The "off" position is still 10% out for hours/days/weeks (well it decays away eventually...). As the Japanese found out the hard way, 10% of a huge amount of power is enough residual heat to cause an awful disaster.

The other is surface area / volume ratio. Couple gigawatts thermal like the Japanese and there's not enough surface area to cool it without giant pumps when shutdown. Couple MW like a nuclear sub and the surface area ratio is better, theoretically you could probably walk away from one without anything awful happening. The people who know aren't talking. Fusion you're talking about something a millimeter across not multiple meters. Walk away or have a computer crash or whatever and it seems impossible for something that small to cause much damage.

There are some practical physics reasons why making a fusion reactor the size of the sun is really easy and the smaller you go the harder it gets, but there are practical engineering reasons why making one bigger than a millimeter would be a huge PITA. On the other hand if you could make a fission reactor the size of a millimeter that would certainly solve a lot of painful thermal engineering problems, but the physics just doesn't work (long story...)

There's also the small matter of the hohlraum costing $100,000 a shot.

Good to see a commenter with an ICF background (most of what I know of ICF was only from water cooler chit-chat).

I didn't manage to get a thesis out. I left the PhD program before the funding was cut, but the writing was on the wall.

Yes, it is short sighted to cut funding of small scale plasma confinement concepts. They are high risk / high reward projects which are not expensive ~$1M. And even if they don't pan out as viable confinement schemes, they make for great training platforms for graduate students.

This is definitely true, and the BBC totally omitted these facts. The official NIF announcement[1] tried a little bit to cover their tracks. They posted some flashy numbers about the total energy output from neutrons, but nestled in the bottom of the page is the total energy dumped into the entire system:

1.7 megajoules.

In comparison, the fusion power produced was a paltry 8 kilojoules, which they compare to the "energy required to form the plasma".

For this technology to be actually effective, there needs to be a multiplier of several thousand. As other commenters have mentioned, this is mostly just nuclear weapons development.

[1] http://www.physicstoday.org/daily_edition/politics_and_polic...

The BBC did say:

> the amount of energy released through the fusion reaction exceeded the amount of energy being absorbed by the fuel

As soon as I saw that, I immediately understood that they're nowhere near breakeven -- much less power generation.

It's an extremely misleading half-truth. Worse, Q must be an order of magnitude greater than 1 IIRC to be economically viable.

They did not mention Q at all in the article and their claims cannot be compared to any intellectually honest definition of "break even" in good faith without more information

IIRC from a recent HN discussion, ITER has attained Q of 0.65 and NIF is nowhere close to that.

Are you sure you're not thinking of JET[1]? ITER's not scheduled to be operational until 2020 AFAICR.

Incidentally when I went to see JET (it's up the road from me, not that far from Oxford) the scale of the thing was enormous. They mentioned that they'd had been able to sustain plasma for 600 seconds. I've got some pictures up[2] on flickr.

[1] - http://www.i15.p.lodz.pl/strony/EIC/ne/future_2_1.html

[2] - http://www.flickr.com/photos/z3r0kl3w/sets/72157623979315835...

I'm jealous you got to see JET.

One of my favorite stories is the disruption. This happens when things don't go just right, and all the plasma with its energy squirts onto one spot. This typically puts the equipment in jeopardy. People try to prevent it, but as they push the boundaries of what's been done, it's going to happen.

Keep in mind that there's only about as much mass in the entire chamber as in a few cubic centimeters of air; it's basically a vacuum. But what little there is is so hot, and there's so much energy wound up in the fields. When JET had a disruption, the entire machine, the largest tokamak in the world, LIFTED ITSELF OFF THE GROUND AND JUMPED A FEW FEET OVER.

The LHC beam-dumps[1] at CERN are also at that holy-shit-physics scale. A more explanatory article[2] also covers the superconducting magnet quenches, in which up to 10GJ of stored energy suddenly decides it really ought to be following Ohm's law again.

Any good refs for this disruption situation?

[1] http://lhc-machine-outreach.web.cern.ch/lhc-machine-outreach...

[2] http://www.symmetrymagazine.org/article/december-2007/protec...

> total energy dumped into the entire system: 1.7 megajoules. In comparison, the fusion power produced was a paltry 8 kilojoules

In other words, the National Ignition Facility hasn't achieved ignition yet.

If you can actually achieve ignition, then nobody will care about the megajoules needed to get it ignited. Because once you've got ignition, you can get an arbitrary amount of energy out.

But if you can't get ignition, well, then, that pesky 0.3% heating efficiency will still leave you a couple of orders of magnitude away from engineering breakeven.

> get an arbitrary amount of energy out.

Not really. You have to build larger, or more, power plants. So you have to make enough extra energy to sell to pay for plants.

While they have had problems in the recent past: https://www.sciencenews.org/article/ignition-failed

~~ They've since had ignition: http://www.independentnews.com/news/article_792110c0-2c5b-11... ~~

[I'm positive that the ICF's predecessors also achieved ~~ignition~~ fusion on smaller scales.]

The real trick is that [0] you need to get out more than you put in (exceed break-even) [1] you need to be able to harness the energy produced by the ignition and [2] be able to do it over and over again, producing reliable power.

> They've since had ignition

No. Notice how it is worded.

> "More importantly, Moses said, the mechanism that generated the powerful energy burst appeared to be the self-sustaining thermonuclear fusion process that defines ignition."

In other words, they did not achieve ignition. They achieved something that is necessary to get ignition. Specifically, alpha heating.

This is another important step on the road to ignition. It is not, however, ignition itself.

Why do you think the title of the article says "Marks Progress Toward Ignition," rather than "Finally Achieves Ignition?"

You're right. In my head "ignition"=fusion, so I was auto-translating as I was skimming. They're looking for something slightly different, and probably using energy output (among other things) as indications of self-sustained fusion, which we've certainly obtained in bombs (fission explosion sets off fusion reaction ☞ big bada boom).

And this seems to hand out a wealth of information as to why they haven't gotten there yet: http://www.reddit.com/r/technology/comments/1nxtyb/nuclear_f...

"50-100 more energy out than in to make up for inefficiencies in electricity generation using this kind of scheme."

Completely true. Theoretically, the NIF target has enough fuel to produce 10-20x the amount of laser energy driving it. Once things work, it isn't a huge step to get to an energy gain (energy_out/energy_in) of ~50. That said, the real world is always more complicated and NIF has yet to ignite.

Speculating about their "real" motives is just that; speculation.

I'm sorry your project was defunded, but that in itself is not evidence that the facility is actually a sham.

NIF is explicitly and publicly a weapons research facility; it's right there on their website[0]. If they happen to do some useful scientific or commercial work then yay but it's a nice byproduct.

[0] https://lasers.llnl.gov/about/missions/

Actually, the site you linked lists 3 goals. Weapons research is only one of them. The other two are commercial and scientific research. Those are not mere byproducts, but explicit goals of the facility.

They're fig leaves.

He's not saying the facility is a sham, he is simply saying it's goals are not to generate electricity, it's simply a research tool to provide data to improve empirical models. Those models of course have all sorts of uses, and certainly a big one is nuclear weapon modelling.

You say "speculation" as if it were a bad thing.

I could speculate that space aliens secretly landed in Roswell and took over the government 60 years ago, but without evidence, it's just a story.

If we could get a fusion design that works a bit over unity, couldn't you, well, chain them?

E.g. 1MW basic plant that drives a 1.5MW plant that drives a 2.25MW -> 4 -> 6 -> 9 etc, however many stages you need.

Yes, but you have to pay for the power plant, and being just a bit over unity won't cover the expense of the next stage in the chain.

You have to be enough over unity to pay for another plant plus surplus to actually use. In which case you can just build a bunch of plants - you don't need to chain them. There are plenty of other power sources you can use to bootstart the power plants.

Even regular power plants need power to start them see https://en.wikipedia.org/wiki/Black_start

I wonder, what more do they need in nuclear weapons? Smaller size? Harder detection? Bigger contamination radius?

Somehow, I am sure, if people start a nuclear war, even the WWII's initial nuclear weaponry will be enough to destroy ourselves, knowing how many countries can now chime in with their own.

Cover the bases: you can't be sure someone else won't discover a new type of nuclear weapon or something you haven't thought of. The big one would be if someone figured out how to get kiloton->megaton yields, from a weapon which didn't produce radioactive fallout.

Sure, target a city and it's functionally indistinguishable, but what about against military formations or bunkers? It would be a big game changer if someone eliminated the fallout problem, since it changes how one could plan to use nukes in a defensive posture.

So you do the research to make sure you're always within a year or two of whatever the next state-of-the-art will be. Being the dominant hegemon has a lot of cost-of-living expenses.

The US now has a notion of "tactical" nuclear weapons.

Some reading:

http://www.fas.org/_docs/Non_Strategic_Nuclear_Weapons.pdf

God help us all.

That's been around since 1953 (less than a decade after the first atomic detonation in '45). [1]

And we've maintained a limited number of low-yield warheads that would fall under that designation. [2] (as an example)

[1] http://en.wikipedia.org/wiki/Nuclear_artillery

[2] http://en.wikipedia.org/wiki/B61_nuclear_bomb

Its the ongoing expense. Look over time, what does it cost to maintain the ability to use one?

Inflation adjusted cost of the entire B-52 program including all maintenance and crew personnel and their training? I donno, like a billion?

A missile field is cheaper to maintain (I think?) and not much more to develop, so its probably a bit under a billion.

A nuke launching sub might cost a billion but it holds alot of them so divided out is probably only hundreds of millions

Grats you've now got one so small and light a plain old fighter-bomber plane only 10 million or so can drop one.

Finally small and cheap enough to do some million dollar class cruise missiles.

I'm sure we'll eventually have cheap drones as delivery vehicles.

The point is minimizing spent money during peacetime. A bunker full of cruise missiles is infinitely cheaper than multiple operating wings of B-52.

The other (obvious?) idea is burying history. Say Canada wanted to nuke us for invading and regime changing them to steal their maple syrup. It could happen. IF we used 1940s techniques which only require a billion or so and a couple months or a whole year, anything their spies steal is very dangerous. On the other hand, "Go ahead, steal these 1980s kryton switch designs, it'll take you 50 years and a trillion bucks to figure out how to use them not to mention all the extra stuff that goes with those handy dandy trigger mechanisms". A good analogy is the time machine. So you can visit Bavaria in the 1300s if you sneak a 1500s level wooden clock escapement that would probably revolutionize Bavaria in the 1300s one way or the other. On the other hand sneak in a modern AMD64 CPU, just the bare chip, not even a heatsink or a PCB or anything, just the chip, and they're going to be all "WTF" all they have to do is get to the 2000s or so, and this will chip will leapfrog them into the 2010s, but its going to take a heck of a long time to get them from 1300s to 2000s without any help. So, yeah, go steal that modern US design, a mere couple trillion bucks and 70 years experience and you'll be stamping them out like license plates just like us. I guess by the 2080s we should be worried, but right now no problemo.

"Lawrence Livermore National Laboratory has a mission of strengthening the United States’ security through development and application of world-class science and technology to:

Enhance the nation’s defense; Reduce the global threat from terrorism and weapons of mass destruction; And respond with vision, quality, integrity and technical excellence to scientific issues of national importance."

https://www.llnl.gov/about/whatwedo.html

I am waiting for the hack of cold fusion;

Now wikipedia says that imploding bubbles create some high energies;

http://en.wikipedia.org/wiki/Mechanism_of_sonoluminescence http://en.wikipedia.org/wiki/Bubble_fusion

Maybe if you shoot up Deuterium bubbles in compressed water and push in some ultrasound waves ....

Reply on news.ycombinator.com