Nuke Fusion power--is this NY Times article realistic?

Okay, we all know that electric power from nuclear fusion is “only 20 years away” —and always will be.

But Thomas Friedman at the NewYork Timessounds hopeful. (Though, yes, he also admits that “some scientists are skeptical”, and that it is “really hard” even in the lab, and will cost $10 billion).

So what’s the latest Dope? Have there been new developments to hint we might actually see this work within the next couple decades?

Laser-induced fusion has been in the works since the 1960’s. I would bet against it being commercially viable in the next decade.

First of all, the article cited in the o.p. is an op-ed, not even a pop-tech article. The National Ignition Facility, while designed to investigate phenomena pertaining to inertial confinement fusion, is not in any way intended to be a practical research tool for applied power generating nuclear fusion; it is designed to examine the conditions in which self-sustaining fusion (i.e. fusion powered by exothermal energy flux) occurs, if but briefly. In fact, the NIF was (according to some sources) primarily intended to provide correlation data for nuclear weapon simulation, and a large portion of DOE funding was predicated on this utility.

One particular statement from the article highlights the author’s misconceptions about the purpose of the facility and the potential for practical energy generating fusion: “The next step — which the N.I.F. expects to achieve some time in the next two to three years — is to prove that it can, under lab conditions, repeatedly fire its 192 lasers at multiple hydrogen pellets and produce more energy from the pellets than the laser energy that is injected. That’s called “energy gain.”” What Freidman doesn’ t seem to understand (or perhaps willfully misstates) is that it isn’t enough to just have overunity power production; you actually need to generate enough power to overcome all systemic losses in keeping the reaction going indefinitely. This is called the fusion energy gain factor or Q (quality) factor, and is generally agreed that power generating fusion probably needs to have Q>10 to be practical.

[post=9759343]Here[/post] is an old thread that discusses the potential for practical fusion in some detail.


I like Thomas Friedman’s columns, but he’s been criticized for excessive optimism. There is even a Wikipedia article for a six-month unit of time, because his columns frequently predict that events are six months out.

I note that the OP of that thread says it’s always 30 years away, while this one says it’s always 20. So clearly some progress has been made.:smiley:

Or at the very least, greater than 1, which nobody has been able to achieve yet either.

Seriously, fusion is easy. There are undergraduate physics labs where you can do it, with an apparatus that fits on a tabletop. It’s practical fusion that’s hard.

It’s even easier than that; all you need is to put together a few octillion tonnes of hydrogen, and Newton will do the rest. The FOB bill is horrendous, however.

It’s very easy to talk about how very easy nuclear fusion is to accomplish. It is very difficult, however, to talk about how very difficult it is to overcome the actual details of making confinement fusion work. Barring some unlikely major revelation in plasma physics, it’ll be thirty or forty years before we can even begin to talk realistically about power generation from nuclear fusion.


Could you elaborate on this please? I’m a science teacher and am greatly interested in this subject.

Fusion is easy!

ITER is an international project to build a nuclear fusion reactor. It is scheduled to be finished in 2018 and some of their goals are to pass the break-even point by up to ten times (Q=10) and be able to sustain the pulse for up to 8 minutes. However, it will be an experimental reactor and won’t convert the generated heat into electricity. It will be a Tokamak design which uses magnetic confinement and not inertial as the laser example given.

The planned successor of ITER is called DEMO, with the objective of reaching a continuous 2000 MW output. You’ll probably have to wait until at least the fourties or fifties before that happens though.

They’re both wrong.

It’s always been a quarter century away.


The biggest problem with sustaining fusion is that the plasma leaks too much energy. X-rays and high-energy neutrons carry energy out of the plasma. Yet plasma is very strongly coupled to electromagnetic energy; it’s only because the plasma in tokamaks is so small and thin that it’s optically transparent.

I wonder: how big would a fusion plasma have to be before it was effectively opaque to x-rays, and neutrons would reach thermal equilibrium? Could the answer to fusion simply be “make the whole thing gigantic”? Maybe we’re in the position of intelligent fleas trying to build a tiny flea-sized campfire- the physics just doesn’t scale down that small.

I have this vision of a ginormous (like a mile across) plasma containment vessel, with fusion taking place in the center and the rest of the plasma serving to keep the core hot by scattering energy back towards the center.

Based on the writer of the article in the OP, it should then be about 50 Friedman Units :smiley:

Fusion power is still decades away at best. Here’s a PDF version of the American Physical Society’s publication APS News (Warning! PDF!) Page 3 has an article about ITER. ITER is not on line yet, and it’s projected to be a factor of five less power than a proposed fusion power plant. It’s still a basic research facility, nothing like an engineering prototype. Even if ITER worked absolutely as well as expected, said power plant would be no less than 20 years away. And that’s assuming somebody magically solved the “first wall problem”. The article I cited all but said the problem of how to construct the innermost wall of a fusion power plant has not been addressed even at a basic level. I’m thinking we’re decades of material science away from solving that problem in an R&D setting, and years more in a engineering/power production setting.

You’d think a smart guy like Friedman would at least glance at the Wikipedia article on an unfamiliar and highly technical subject before wrting about it. Or talk to one of the world-class science reporters that the Grey Lady employs. But nooooooooo.

Tommy, you’re down another notch in my respect-o-meter. But you can turn it around - the next six months are key. :rolleyes: x 180

According to Wikipedia’s article on ITER, a viable commercial fusion reactor is at least 40 years away; their hope for ITER itself is to prove that we’re going about it in the right way and they expect that to take 30 years. Since politicians and enviro-protesters slow up every power plant by another decade (minimum), we’re looking at 2060 for an ideal timeline.

So I think it’s a given that we will eventually get fusion power to work, but I’d be surprised if anyone here was alive to see reach commercial production.

I think the problem would be getting a fusion reaction going at that scale. JET says it has an input of about 16 MW of power, and ITER is planned to use about 100 MW. ITER will use 0.5 g of plasma material. So I think you can see how impractical it would be to scale up. If you wanted a ton of hydrogen, I still doubt that you’d have enough to absorb x-rays and neutrons, but you’d still need 1 million times the energy to heat it up. (And since the plasma is about 1/1000 the density of air, you would be talking about something on the order of kilometers to get a ton of plasma up to temperature).

And just imagine the damage of a containment failure of 1 ton at 100 million K…

Scaling up (though not quite in this fashion) reduces a number of problems that occur because the requirements for containment are lessened. However, you also end up with a large containment volume and initiation requirements scale more-or-less with volume, so if you have a ‘core’ that is twice the radius, the energy required to initiate fusion is eight times as much. (I’m assuming that a horse is a sphere to make the math easy, but you get the idea.) Once you scale up to a point where you mitigate some of the smaller scale effects you end up with a device the size of a small moon; unmanageable in terms of both energy requirements and physical construction. All practical energy generation fusion reactors will have to be non-equilibrium, requiring some kind of externally-controlled confinement and regulation.


Is that initiation energy requirement just what it takes to get the plasma that hot? Or is it what has to be continually pumped into the plasma to keep it from cooling off? I don’t doubt that when considering a plasma the size of a nuclear fireball, we’re talking some serious energy. But the whole volume doesn’t have to be fusion hot, just the core. I envision heating the core by electrostatically accelerating minute particles of lithium to a high enough speed/energy that they plasmize about the time they reach the center of the plasma. As long it’s disassociated into nuclei and electrons, the plasma should couple strongly to electromagnetic radiation, and thus serve the purpose of minimizing radiation loss. The core would be fusion hot, while the edges of the plasma could be “only” 10,000 C. which would also allow it to be denser (and therefore a more efficient blanket), and subject the physical walls of the vessel to far milder conditions. I was hoping this would allow the whole scheme to be “domed city” size instead of “small moon” size.

P.S. and of course if the thing did work, we’re talking about a fusion plant that would supply the energy for a continent, if not a hemisphere. But hey, sometimes you just have to think big. :stuck_out_tongue:

For all intents and purposes, scaling up of initiation power throughput scales with volume. Remember, too, this has to be evacuated except for the plasma; you can’t pressurize it from within. Structurally, a containment vessel of this size impossible, or at least massively impractical. This is not a problem that can be solved by just making things bigger.


You know, as I’m just now starting to get back into college teaching, I wonder if this idea might make a great demo when I need to give the class a much-needed break from database design…

Then, I’m thinking about how this just might resonate with parents and the university administration…

…naw, maybe w/a grad class, don’t need the yellow journalism headlines, but thanks! that was interesting.