New 'Stellarator' Design for Fusion Reactors

eldavojohn writes "The holy grail of fusion reactors has always seemed 'just a few years off' for many decades. But a recent design enhancement termed a 'Stellarator' may change all that. The point at which a fusion reactor crashes is when particles begin escaping due to disruptions in the plasma. A NYU team has discovered that coiling specific wires to form a magnetic field may contain the plasma. This may be a a viable way to create a plasma body with axial symmetry, and a far better chance of remaining stable. Like other forms of containment this does require energy itself, but could bring us closer to a stable fusion reactor. It may not be cold fusion or 'table top' fusion but it certainly is a step forward. The paper is up for peer review in the Proceedings of the National Academy of Sciences."

Comments

Princeton 1951 called...

[Anonymous Coward]

...they want credit.

If they used...

[jd]

...Axl symmetry, they could produce something that was violently unstable but produced vast amounts of marketable energy and money.

Re:If they used...

[heinousjay]

I hope you're serious, I take great solace in knowing the world is full of crazy people.

Re:If they used...

[jd]

I don't - they keep getting elected. What we don't need right now is more crazies in power. What we need is someone so totally and utterly insane, they'll spend two or three trillion dollars a year on getting a full-scale fusion reactor built and operational before they get kicked out of office or shot. Yes, there are many unsolved problems, but we're running a little low on time and researchers are too busy on corner cases that might never happen in a real reactor under normal conditions. Building a live system and requiring the scientists to live within blast radius would likely get faster results and just as much reliability.

Re:If they used...

[The_Wilschon]

Blast radius my foot. A fusion reactor is immensely safer than a fission reactor. Furthermore, fission reactors are really very safe (far safer than, say, oil refineries). Even Chernobyl was primarily a /chemical/ explosion (although caused by problems with the reactor), which happened to scatter radioactive debris over half the globe. A chemical explosion at a fusion plant would scatter hydrogen. Oh boy. Even the unstable isotopes of hydrogen are still light enough that they would float to the top of the atmosphere and escape into space in very little time. A fusion reactor is not a controlled H-bomb. Unlike a fission reactor, which requires a carefully tuned reaction to walk the knife's edge between dying out and going critical, the hard part with fusion is keeping it going. Fusion is very fussy. If the density, and the temperature, and the composition of the plasma are not just exactly right, then reaction dies out in a fraction of a second, the time it takes to exhaust the really tiny amount of fuel that is available to it at any given time. To keep it going, you have to keep feeding it more fuel, as well as carefully tuning things. If there were even a very very tiny explosion, the worst it would do is damage the devices tuning the plasma's parameters, and then the reaction would die out. Even if the fuel feeders went crazy and started flooding hydrogen in as fast as they could, it would still just die out. There is no way that the reactor, even in an undamaged state, could bring enough hydrogen to the needed density and temperature quickly enough to cause a thermonuclear explosion even on the scale of a pipe bomb. So, I say, blast radius my foot, unless you want to compress the researchers down very very small and put them inside the plasma itself.

Researchers are not involved in corner cases that might never happen. Nor are they worried about reliability yet (in the sense of preventing another Chernobyl, as opposed to the sense of very little downtime). They are just trying to get the blamed thing to produce enough energy to sustain itself, with some left over. (Although, if you're feeling pessimistic enough, you might call that a corner case that might never happen!)

I agree that we need to get a lot of funding to fusion research, but throwing money at the problem won't necessarily solve it. It is a very hard problem. Furthermore, we'd need not just one crazy (I presume you refer to the office of the President), but a whole bunch of crazies (half of Congress), because Congress makes the budget.

Re:If they used...

[joto]

We already can get ALL the energy we need from renewable sources

That might be so, but it certainly isn't economical, otherwise we would already have all our energy from renewable sources. Furthermore, very few renewable sources of energy shows any sign of promise in the short term, although solar certainly seems interesting once someone comes up with a "breakthrough". Oh, and we could build more dams, they can certainly be profitable, but often takes a huge toll on mother nature.

a bunch of overblown assholes who aren't remotely interested in finding a CHEAP energy source - they want a CENTRALISED energy source that a government can completely control.

Actually, centralized makes a lot of economic sense. There is only one way of generating power that is at least potentially profitable at a small scale, and that is solar cells. But even here, you get economic benefits by simply aggregating them in a small area, which simplifies maintenance and infrastructure. Nuclear just doesn't scale down at all, and everything else, is simply more efficient at a larger scale, whether it's wind-turbines, dams, wave-turbines, geothermal, waste energy from industry, burning of garbage or methane from a landfill, or even gas, diesel, or coal. If you don't believe me, try putting up a small propeller in your backyard, and compare the $/watt of this to a typical 100 meter wind turbine, or to a huge turbine farm with 100 meter turbines. Or compare the $/watt of a small waterwheel to the Hoover Dam. Or compare the efficiency of a diesel-generator you can afford, to one that you can only afford to rent, or to one that is able to power a whole city. These things might scale down, but certainly not in an economical way.

It's always "It'll be ready in 50 years' time". i.e. in 50 years' time, they'll STILL be saying "It'll be ready in 50 years' time".

This is certainly a valid criticism for fusion power research. But I hope you don't seriously believe that just because you can't have it today, we should stop researching it. The benefits of such a technology would be incredible, and the money we spend on it is not that much. Still, one can debate whether pouring money into tokamak research is justified given its track record so far, This, however, is not a tokamak, and as such is a potential theoretical breakthrough, even if it might be an unlikely one (I'm not qualified to judge that).

Stellarators aren't new

[Anonymous Coward]

The summary makes it sound like stellarators are something novel, which they are not. Research has been going on for decades, most notably with the German Wendelstein experimental reactors.

Re:Stellarators aren't new

[iamlucky13]

In fact, the stellarator design is almost as old as the Tokamak design. The first one was built in 1951.

Somebody over at physorg got a little too excited about a fairly low-impact paper from NYU. If you read the abstract, you'll see that the paper just deals with the design of the coils for a stellarator.

Most likely, this is for the National Compact Stellarator Experiment (NCSX) being built at nearby Princeton, which will be the first stellarator designed with a computer optimized plasma geometry. I think it will also be the largest stellarator to date, with 12 MW of heating capacity. In contrast, the JET Tokamak has 37 MW and the ITER Tokamak will have 110 MW of heating. Unlike ITER, NCSX will not be capable of break-even operation.

Stellarators often get mentioned in fusion power discussions because they provide a more stable containment design, whereas a Tokamak needs one extra set of electromagnets to deal with the fact that the magnetic field is weaker at the outside of a torus of magnets than at the inside. Although a stellarator is therefore a little simpler in that regard, the geometry and plasma modelling is much more complex, and this in turn creates problems for designing the coils and the exhaust diverter. Because of this, most of the funding and research effort has gone to the Tokamaks.

A little more info here: http://en.wikipedia.org/wiki/Stellarator

Anybody care to bet on whether this shows up on CNN's tech page in a day or two as some major "recent design enhancement?"

input-output

[polar red]

Like other forms of containment this does require energy itself I find it weird that the amount of energy needed to contain, is less than the energy contained in the plasma. Can anyone explain this ?

Re:input-output

[Paul Pierce]

Like other forms of containment this does require energy itself I find it weird that the amount of energy needed to contain, is less than the energy contained in the plasma. Can anyone explain this ?
Picture Chinese handcuffs

Re:input-output

[feepness]

Picture Chinese handcuffs Great now I can't let go of that image.

Re:input-output

[thanatos_x]

It's intrinsically harder to do useful work that raw work, just like it's easier to destroy than create....

It could also be that it's a brute force attempt to force cohesion, and since force must be met with equal force it's very difficult. That also assumes it could concentrate the exact amount of energy at exactly the right point. Just imagine trying to not only stop a terrorist attack, but subdue them without lethal force. They need one leak to win. You need a perfect record.

Re:input-output

[iamlucky13]

Well, I'm not a plasma physicist, so I'm not intimately familiar with all the details, but one thing that jumps out at me right away is the distinction between energy and power.

Energy is the ability to do work. Power is the rate at which work is done or energy is extracted.

The plasma contains a great amount of thermal energy with a tendency to do work (by difussing to the reactor walls), so you have to set up a barrier to accomplishing that work. This is analogous to a dam holding back water. The water, due to it's elevation, has a lot of potential energy, but no power is required to hold it back. Power is extracted as it's let through the turbines.

It's a little more complicated for a plasma. A charged particle moving through a varying magnetic field (like that surrounding the reactor) does work and thereby loses energy. As a result, there is a tendency, although less definite than with a dam and water, for the hydrogen ions to only move around in the reactor along lines of constant magnetic field strength.

Once a magnetic field is established, it ideally takes no energy to maintain, except as charged particles move through it. So power only has to be supplied to the electromagnets to account for their inefficiency (0 under ideal conditions in a superconducting Tokamak) or as work is done on the field by charged particles escaping. Since most of the energy from the reactions is carried away by neutrons, which have no electric charge and therefore don't affect the field, the containment power is sufficiently smaller than the reaction power that this is theoretically feasible as a power plant.

Actually, the biggest power demand in a Tokamak as I understand is for heating the plasma to a temperature where fusion will take place. The hotter it gets, the faster fusion occurs, eventually reaching a breakeven point energy is released by fusion faster than it is carried away by escaping neutrons and gamma rays. Then the plasma can sustain itself. We haven't gotten there yet.

Sorry, the dam analogy isn't great and talking about charged particles in a magnetic field is a little abstract. Hope this helps.

Re:input-output

[counterfriction]

"Energy" in the context of containing a plasma is actually work. They have the same units, so they're like exchangeable currencies (i.e. some energy will buy you work, and some negative work will buy you energy)
The energy that a plasma intrinsically has (like kinetic energy) is just that; energy.

Here's a related (but certainly not airtight) analogy: A brick can have some gravitational potential energy relative to the earth's surface. If you're standing on the ground, that brick will have some nominal gravitational potential energy. If you lift that brick 1 meter, you'll do some amount of work. If you're hanging over the edge of a helicopter at a couple hundred meters, that brick has substantially higher gravitational potential energy. However, if you lift the brick a distance of 1 meter, you'll still do the same amount of work.

So, what's going on here is that a plasma can indeed have a lot of energy (relative to the earth's environment). However, the "energy" we're putting in is actually work to contain that plasma.

Huh...

[Greyfox]

SCO Loses and they figure out fusion* on the same day. Coincidence? I think NOT!

* Sure it doesn't say they figured it out in TFA but humanity will point to this day and say 'That is the day SCO lost and they figured out fusion.'

Thorium reactors

[Bombula]

I was reading about thorium reactors recently. Seems like that's much closer to being rolled out, and its developers are claiming it solves a lot of the problems with existing reactors: it's more stable because thorium reactions don't chain the same way, it doesn't produce waste or plutonium, it can actually burn up other waste - including plutonium, and it can be used in some types of existing reactors (there are trials in Russian reactors right now).

Does anyone know any more about this?

Re:Thorium reactors

[Bombula]

Thorium good, but if possible, fusion even better.

It's important to define 'better' here. Cost would seem to be an important consideration, for example. I don't know what the price tag of fusion is so far, but it's awfully, awfully high already and without a great deal to show for it. If we've already got a pretty good thing in thorium, and we already have the reactors, and there's enough thorium and uranium to keep us in electricity at present consumption rates for thousands of years, and it's non-polluting and all the rest, then how is fusion - a hugely expensive, so far unproductive technology - 'even better'. I'm not quibbling or trying to be antagonistic here - it's a serious question, and it needs a serious answer considering what's at stake: we need clean, non-polluting power that doesn't ultimately come from politically volatile parts of the world.

Re:Thorium reactors

[mdsolar]

Solar and wind power fit the bill of being clean and local. A lot of our nuclear fuel these days comes from Russian weapons stockpiles. But the process of diluting it back down from weapons grade to fuel grade is not going all that well. In an accident in Tennessee last year that was covered up until congress stepped in, the plant managers thought that a big spill of highly enriched uranium soulution, enough to cause the kind of accident that killed 2 people in Japan 1999, was natural uranium. There were two places where the spill might have accumulated and cause criticality. That is pretty poor materials control if you don't know what it is that you are working with.

Uranium reserves are estimated to be about 85 years at present use. Plans to extend the life of nuclear power all pretty much include breeder reactors (such as thorium reactors) and have unresolved fuel cycle problems. Fast breeder reactors are also illegal in the US owing to proliferation concerns. Their prototypes have also tended to melt down.

The new reactor being planned for Calvert Cliffs has an estimated price tag of $2.50/Watt for construction alone, though with federal loan guarranties included in the Senate Energy Bill, this price will likely rise substantially. The price compares poorly with wind and solar, both at about $1.30/watt to build, but with much less in the way of operating costs, and obviously no fuel or long term waste disposal costs.

The level of effort put into fusion has not really been that large. You hear about it, but compared to the Manhatten Project, out of which nuclear power came, it gets much less in the way of GDP. Renewables get even less than that. This was deliberate. The idea was to give it enough effort so that it would be ready when oil and coal ran out. The problem is that at the time, the growth in the use of coal and oil was not foreseen. So, fusion is actually right about on schedule. When it is here, there may be some trouble siting it since nuclear power plants squat on some of the better cooling resources and our storage in place policy for nuclear waste may keep these prime resources tied up for hundreds of years. But, wind was 20% of new generation in 2006 and is growing at 50% per year, while solar is growing at 30% per year and this should accelerate as the silicon purification bottleneck clears. So, fusion may enter a market that is already dominated by clean inexpensive power and thus find only niche applications in any case.
--
Go solar the easy way: http://mdsolar.blogspot.com/2007/01/slashdot-users -selling-solar.html

Re:Thorium reactors

[barawn]

That's not exactly true. The ideal output of most fusion cycles is stable. You get side production of tritium and a few other radioactive isotopes, though.

But fusion does, however, produce a large amount of radioactive waste. Not through the products of the reaction. Through the byproducts of the high level of irradiation.

The difference is that fission radioactive byproducts are long lived. Fusion radioactive byproducts are extremely radioactive, but very short lived, and therefore easier to deal with

Re:Thorium reactors

[Your.Master]

Advantages of Fusion:

* Fuel for fusion can be extracted from water, including non-potable water. Fission requires Uranium & Thorium to be mined and transported, and your country might not have it. By the time the fuel runs out, our sun will be a red giant, so we should worry about escaping the solar system before doing any better than that.
* No weapons material generation (Thorium is in some respects similar here).
* Radioactive waste: there's a lot less of the stuff sticking around, and really no high-level waste at all. This can save a lot of money with disposal and some with safety equipment, not to mention avoiding the headaches of dealing with people who believe that radioactive waste should not be produced at any cost. Note that this is actually partly a problem, since many nuclear "waste" materials have important industrial and medical uses, so we will likely continue to run some of these fission reactors anyway (or perhaps figure out ways to produce these radioactive isotopes more directly).
* Although fission plants are perfectly safe, there are a lot of people who still fear them for a variety of reasons, mentioning any of which is likely to lead to a fruitless and flame filled side-discussion if anybody reads my post at all. Fusion power is inherently even safer, which might satisfy a few of these people.

Re:Thorium reactors

[rossifer]

Almost all of the waste from a molten salt Thorium fuel cycle reactor has a half life of 30 years or less (total storage of 300 years and it's as clean as the thorium ore it came from). Also, the mass/volume of the waste to be stored is substantially lower than a light water reactor because you can continuously mechanically and chemically extract the waste from the liquid fuel. With a solid fuel reactor, the waste is physically tied to 90% of the still-good U235 and the now damaged ceramic that makes up the rest of the pellet. You have to sequester all of that unfissioned U235 along with the pellet body which amounts to about 200x more waste to deal with.

Because the waste from a liquid fueled reactor can be continuously extracted in very small quantities, it's fairly easy to make uber-safe small containment vessels and constantly courier it to your long-term storage site. (a 1GW reactor would produce less than a half liter of waste product per day so you could make each container hold 100cc of the stabilized waste products and only need a modified armored vehicle to safely transport the five uber-bottles each day) With solid fuel reactors, you have large refueling events that generate multiple tons of waste per event. The quantity that must be managed at once makes it that much more dangerous to handle and transport safely.

Another nice thing about molten salt reactors is that they can be 97% fuel efficient. Unlike our current light water reactors which only burn about 10% of the fuel before the solid fuel is too contaminated with reaction poisoning fission products to keep an efficient reaction going any more. This efficiency is mostly due to on-site constant fuel reprocessing. There's an alternative molten salt reactor approach that doesn't involve reprocessing that is about 50% fuel efficient but gives you a lot of crappy waste every 20 years. I prefer frequent small quantities myself...

Possibly the nicest thing is that if you use a dual fuel configuration (a LiF/BiF2/UF4 kernel and a LiF/BiF2/ThF4 shielding/breeding layer) the core is thermally self-limiting. As the reactor heats up, the salt mixture expands and reduces the reaction rate until the whole thing stabilizes around 1500-1900C (the final temp depends on the exact fraction of UF4 in the mixture). You don't need control rods or any of the additional equipment to maintain moving parts in the reactor. All you need are pumps to cycle the kernel fluid and the primary cooling fluid and if those shut down, there's nothing to go wrong. The whole thing heats up and sits there radiating heat. If you want to put an automatic stop to the heat, put a thermal plug in the bottom that will melt if primary cooling ever stops and drain the whole core into subcritical storage containers underneath the core.

One big problem with the molten salt reactor is that the existing nuclear equipment industry makes most of it's money from fuel manufacturing. Molten salt reactors are constantly reforming the liquid fuel on-site, which means that the existing nuclear infrastructure has to change their business model (or be supplanted) before it can possibly work.

The other huge problem is that most of the advantages of the Thorium fuel cycle come from the fact that it's a breeding/reprocessing cycle. Both of which (fuel breeding and fuel reprocessing) are currently illegal in most first world countries due to nuclear proliferation concerns. Laws can be changed, but governments would have to be reassured about the risks. The thorium fuel cycle can be spiked with U232 which will prevent the creation of nuclear bombs (because U232 decays into a hard gamma emitter that destroys nearby electronics), but wouldn't prevent the construction of radiological "dirty" bombs (basically what Korea can build right now).

Ultimately, the big advantage of newer fission reactors over fusion reactors is that we can build the fission reactors today. Further, if we decide to breed fuel, there's enough known U238 and Th232 in the upper crust to prov

Re:Thorium reactors

[adrianmonk]

we need clean, non-polluting power that doesn't ultimately come from politically volatile parts of the world.

I don't know if that will ever happen. With most sources of energy, the fuel is unevenly spread around the globe. And being some small country that has a huge reserve of some kind of fuel will tend to mess your country up in the same way that people who inherit a lot of money (and never have to work a day in their lives) get messed up.

Trade doesn't always have to create political instability, but it certainly doesn't help when you have a country that controls some resource that society must have in order to keep functioning. If someone must have something, they're willing to take it. And conversely if you have something that someone else needs, you tend to behave however you damn well please.

Oh, great!

[Sloppy]

I just bought a fusion reactor that uses the old design!

Stellarators have been around as an idea for years

[Zarhan]

...and as prototypes too.

http://en.wikipedia.org/wiki/Stellarator

Anyway, basically what I know about this is that stellarator designs avoids lots of the problems that are present in Tokamak - namely, degrading of the reaction chamber due to escaped neutrons. A fusion reactor using stellarator instead of Tokamak would, in effect, last forever since the material does not become radioactive.

Especially the Germans have been researching this stuff a lot, however, most of the big money is currently in Tokamak designs, including ITER. Which is kinda a shame - since we're not in the Manhattan Project-type "if you have 3 designs and think one of them might work, build all three, here's the money"-situation..so these nice ideas may only be developed further if Tokamak fails to become viable..

Doh!

[the eric conspiracy]

A stellarator is not a new design. The first examples were built here in 1951.

Why reinvent the wheel?

[viking80]

Design parameters for fusion reactor:
1. Contain a plasma ball with high density for fusion reaction. Ball is much better than doughnut if you just can figure out a way to keep the plasma together.
2. Make a wall that is far enough away to not melt from this plasma ball to absorb heat/radiation to make power, and keep it close enough to get high enough energy density on its face.
3. Make the wall 1 ton/m^2 to protect the people outside
4. Use magnetic field outside plasma ball to contain radiation.

This seems like a tall order, and it is, but consider the sun/earth:
1. Gravity works great compared to magnetism.
2. Well, here on the earth, it is 1kW/m^2. That is much higher than the energy consumption in most cities. Should be good.
3. Our atmosphere stupid.
4. The earth again has a great magnetic field that protects us pretty well.

Bottom line: Why reinvent the wheel?

Re:Why reinvent the wheel?

[OeLeWaPpErKe]

4. Use magnetic field outside plasma ball to contain radiation

This seems like the exact reason why basic physics should be mandatory in schools. Dear God. How exactly would a magnetic field contain neutral photons ? They will generate zero flux and will not interact with the field at all.

Re:Why reinvent the wheel?

[Wonko the Sane]

3. Make the wall 1 ton/m^2 to protect the people outside

Houston, we have a unit problem

Re:Why reinvent the wheel?

[lawpoop]

This seems like the exact reason why basic physics should be mandatory in schools. Okay...

Dear God. How exactly would a magnetic field contain neutral photons ? They will generate zero flux and will not interact with the field at all. Is this the kind of basic physics that the average student would understand in their mandatory class?

Re:Why reinvent the wheel?

[rossifer]

Actually, you'd be correct if you were to say that neutrons are not affected by electric fields. But neutrons are fermions with magnetic spin and are affected by (and can be moved around with) magnetic fields, so...

Regards,
Ross

Magnetically confined plasma fusion reactors

[the_kanzure]

Related links: * LDX@MIT
* Physics of magnetically confined fusion [pdf]
* The main principles of magnetic fusion
* Magnetic fusion experiments at LANL
* High density magnetic fusion
* Has a good bit on magnetic confinement
* Can a magnetic field be used to contain plasma?
* International Thermonuclear Experimental Reactor
* What's happening in fusion?
* Design of magnetic fields for fusion experiments [pdf]
* Wikipedia article on the topic
* Magnetized target fusion bibliography
* Plasma physics bibliography
* Databases for plasma physics

* Plasma physics laboratories
* List of plasma physicists
* Plasma on the internet

"Stellarator"

[flickwipe]

Refrigerator full of Stella?

Nice News for Nerds but...

[swokm]

If society won't even accept safe fission designs, what makes you think we will ever get far more powerful fusion reactors built? I think the largest problem now is the culture of misinformation and fear, not the problem of technology.

Unless I'm wrong, the production of non-military nuclear reactor designs in the US for the last 30 years have been... zero. Unless you count the Galileo, Ulysses, and Cassini space probes. Call me when we upgrade all of our reactors from 1973 designs to a much safer and cleaner Gen IV design -- like this bad boy (now with free hydrogen!) instead of taking high-level radioactives --potential fuel-- driving them recklessly around the country in truck, and shoving it into a salt mine, or some similar brilliant idea.

Besides, though I lust for the sheer coolness of magnetically confined plasma as much as any proper geek, the the simple fact is we have had the technology to use fusion for power for quite some time now(press release from 1998, although building the X-1 was promptly cancelled without reason) with Z-pinch inertial confinement on the insanely cool Z machine at Sandia.

Yawn. Wake me went the politics of our time aren't ruled by Luddites with pitchforks and torches...

From the chasing your own tail dept.

[Lord Balto]

For the $1,000,000,000,000 Monkey Boy will spend in Iraq we could have put solar collectors on every home in America for free. So they finally figure out how to make fusion work. Energy will still be monopolized by the power companies and you'll still be paying through the nose. And if you try to do anything about it they'll call you an enemy combatant and send you to Guantanamo. There is no technological fix. There is only a political fix.

Still concave toward the plasma.

[Ungrounded Lightning]

I seem to recall one of Bussard's points in his talk Should Google Go Nuclear? was that plasma confinement by magnetic fields is inherently unstable when the confinement is concave toward the plasma, no matter how you twist them. Thus Stellarators, Tokamaks, etc. are (in his opinion) doomed. (And that's why his design is conVEX toward the plasma.)

(My take on that has been that even if passive geometries are unstable, if you can get it stable enough that instability growth occurs at no more than an HF rate you might be able to use an active system to finish the job of stabilizing the confinement. But that's a separate issue.)