Nuclear fusion is here!

[Oleron]

Well? Is it?

Is it ever likely to be?

Sorry about the blatantly misleading thread title but it's a competitive market out there.

My point is this - What is the current state of play with genuine fusion research. Is there even such a thing?

Will we ever see fusion power stations and, if so, when?

 

[Matabiri]

Read all about it

The current timing is something like:

2015 - 2020: Next generation research reactor (ITER).
At the same time the fusion materials research facility (IFMIF) should come on line in Japan.
2035 or thereabouts: Commercial fusion reactors.

The physicists keep assuring us that they can solve the problems associated with controlling the plasma. The real problems are materials ones - the first wall of the reactor is bombarded with high-energy (14MeV) neutrons, which cause massive microstructural damage, as well as activation (the materials have to be disposed of as radioactive waste). The activation isn't that much of an issue. The microstructural damage is. Additionally, almost all the data we have is from fission reactors (much lower energy) and accelerators (different energy spectrum), and so although we can try and make models and predictions (what I'm working on), we won't really know how these materials will react under a genuine fusion neutron spectrum until IFMIF is completed.

IFMIF has an irradiation volume of about two litres.

Currently we're trying to use steels, because they're nice. The alternative is to use ceramics, which are much more resistant to radiation damage, but are brittle. So long as the physicists do their job, this is no problem - the reactor's all in vacuum and so the torus is in compression - but if there's an error it'll crack and implode. Also not a real safety risk - the amount of tritium in a fusion reactor is on the order of 2 - 3 grams - but a big PR disaster.

There's a lot of international collaboration going on - the US is welcome back into the fold. They went off and did their own thing for a decade or two, but are back involved again now. Fusion research is currently very sexy, and there's a lot of money going into it. This is, however, going into a small number of big projects (there was planned competition to ITER, but then everyone decided to hitch their wagons to the same project) rather than competetive research.

Hope that helps. I can try and answer any other questions, but can't guarantee to be absolutely up to date.

 

[DaveW]

Well? Is it?

Is it ever likely to be?

Sorry about the blatantly misleading thread title but it's a competitive market out there.

My point is this - What is the current state of play with genuine fusion research. Is there even such a thing?

Will we ever see fusion power stations and, if so, when?

Tough questions. Last I paid much attention to it was about 2.5 years ago. Princeton Plasma Physics Laboratory was still performaing experiments, and was even in the process of designing/building/participating in a new type of containment. The big problem was that noone has been able to sustain the reaction for more than a few seconds (because of the rapid fuel expansion and state of superconductor/magnet technology and containment field design). They could get quite a bit of power out in that few seconds, but not enough to make up for the massive power requirements of a startup (they spin up a large flywheel to drive a generator for startup, if I remember correctly, which takes quite a bit of power). They were getting somewhere around 0.6 Breakeven (I couldn't find the number online, sorry).

The ITER program is aiming for a reactor that can suatain a reaction for 10 minutes, according to a PPPL presentation I found on their website (www.pppl.gov), which would be the first fusion reactor to actually have a shot at commercial viability.

 

[Matabiri]

Re: Re: Nuclear fusion is here!

(they spin up a large flywheel to drive a generator for startup, if I remember correctly, which takes quite a bit of power)

They still do that. And there're no brakes, and they can't leave it spinning, so if the shot doesn't go ahead some poor guy's got to stay behind for many hours while it spins down.

 

[tracer]

Nuclear fusion is here! Well? Is it? Is it ever likely to be?

But of course! It's been here for decades, in the form of the Farnsworth-Hirsch machine.

You didn't say it had to be efficient nulcear fusion.

 

[BPSCG]

Well? Is it?

Is it ever likely to be?

Well, sure it is:

1952

 

[Oleron]

Thanks for the replies guys. I guess I'll still need to order heating oil for this winter.

Let me see if I understand this tokamak stuff correctly. Super heated fuel, in the form of plasma, is introduced into a torus shaped magnetic field containment machine. The fusion reaction is initiated by adding extra energy to the system. In theory the reaction could continue as long as fuel exists and containment is maintained. The magnetic field is for stopping the plasma escaping and making sure the experimenters don't lose their eyebrows. Normal reaction vessels would vapourise at fusion temperatures.
Am I close?

Am I right in saying that Lithium is used only as a heat conductor to convert the energy from heat to mechanical?

As such it doesn't actually deplete?

 

[Matabiri]

Thanks for the replies guys. I guess I'll still need to order heating oil for this winter.

Let me see if I understand this tokamak stuff correctly. Super heated fuel, in the form of plasma, is introduced into a torus shaped magnetic field containment machine. The fusion reaction is initiated by adding extra energy to the system.

Pretty much. The energy is pumped in through lasers or ion beams.

One pretty consequence of the physics of plasmas is that the flow lines don't cross - the plasma is, to all intents and purposes, non-turbulent. So getting fuel into the hottest, most pressurised part of the plasma (the centre) is tricky. Current technology involves firing pellets of frozen tritium into the reactor, in the hope that they'll reach the centre before they ionise.

In theory the reaction could continue as long as fuel exists and containment is maintained. The magnetic field is for stopping the plasma escaping and making sure the experimenters don't lose their eyebrows. Normal reaction vessels would vapourise at fusion temperatures.
Am I close?

Plasma temperatures are about 3 million degrees, although this isn't really a meaningful number. Really, the ions have a kinetic energy equivalent to a temperature of 3 million degrees. But yes, you're close.

Am I right in saying that Lithium is used only as a heat conductor to convert the energy from heat to mechanical?

As such it doesn't actually deplete?

Also true. If lithium gets into the plasma, it has a massive cooling effect (lithium ions are large and radiate, depleting the energy in the plasma) and this can kill the reaction. Modern tokamaks have divertors, which are cooled plates sitting under a crossing point in the magnetic field (sorry; can't find a diagram offhand), so that the heavier ions get pulled out. Helium build-up is also a problem, but the divertors should handle this.

One of the possible designs involves running liquid helium down the inside of the reactor to cool it and soak up neutrons...

 

[Oleron]

Thanks Matabiri.

 

[Hamish]

Am I right in saying that Lithium is used only as a heat conductor to convert the energy from heat to mechanical?

As such it doesn't actually deplete?

Ok, I'm not a tokamak guy, I'm with the other lot doing inertial confinement fusion. IIRC the lithium is not just used as a heat conductor. By surrounding a reactor with lithium you can breed more tritium as the litium absorbs neutrons. The problem is that the neutron stopping range in lithium is huge. I heard someone suggest that you'd need a metre thick blanket of lithium around the reactor to stop neutrons.

Just to get the other side of the story straight, there are other approaches to fusion apart from ITER/JET style tokamaks (tokamaks get more publicity and, it must be said, are substantially closer at the moment).

Inertial Confinement Fusion (ICF) relies on extremely symmetrical compression and heating of a millimetre sized frozen sphere of deuterium-tritium fuel in order to initiate fusion. There are several potential ways to do this, the most studied being laser driven compression (which can be direct or indirect). Two huge lasers are getting ready to try to achieve ignition, the National Ignition Facility (NIF) in the US and the Laser Megajoule in France. Ignition basically means getting more energy from the fusion reaction than goes in. NIF was built primarily with military money as a safe way to test H-bomb technology so the chances of it beating ITER to any sort of commercial energy production are small.

Several novel schemes to improve efficiency of laser driven ICF are being investigated but are still young and suffer from stumbling blocks in the engineering technology required for commercial viability. Schemes like Fast Ignition or unorthodox compression geometries show promise but need a lot more work.

There's also the Z-pinch which never gets any attention but is potentially up there with the others. The idea is to put so much current trhough a vertical array of thin wires that they vapourise and form a plasma almost immediately. The current flowing through them causes a magnetic field which "pinches" the plasma onto the central axis. As it collapses itgives out a phenomenal burst of x-rays. These could, in theory, be used to compress a fusion fuel capsule as in ICF schemes. The problem is that you put so much energy into the machine that to be viable as a reactor you need to put in a rather big fuel pellet (the x-ray energy is enough to handle quite big volumes of fuel). The problem then becomes one of containing the energy from what is effectively a small H-bomb. Whacky schemes proposed have involved letting the whole surrounding assembly melt, dribble through a hole in the floor where the energy can then be extracted before a replacement wire-fuel assembly is slotted into place. They're some way off yet but they do have one of the coolest pictures of any of the projects(electrical breakdown of water in the Z tank)

There are a few more that just didn't make it. The magnetic bottles that were too leaky. The theta-pinch. The fascinating (but just not good enough) muon catalysed fusion idea.

Of course, we do rely on fusion power every day. Gravitational confinement fusion has been shown to work all over the universe. We're very lucky to have a huge reactor only 150 milliopn kilometres away with enough fuel to last a few billion years.

 

[Matabiri]

Just to get the other side of the story straight, there are other approaches to fusion apart from ITER/JET style tokamaks (tokamaks get more publicity and, it must be said, are substantially closer at the moment).

Inertial Confinement Fusion (ICF) relies on extremely symmetrical compression and heating of a millimetre sized frozen sphere of deuterium-tritium fuel in order to initiate fusion.

The trouble with ICF and Z-pinch are that they're not continuous, isn't it?

Sticking with tokamaks for a second, though, small aspect ratio tokamaks such as MAST are looking more efficient than the large ones like JET, although they're cruder at the moment.

 

[Hamish]

The trouble with ICF and Z-pinch are that they're not continuous, isn't it?

One problem, yes. It's more of a problem with lasers than with Z-pinch. With laser driven ICF you'd probably need 10 shots per second to get decent power output. The main problems are that if you use just one reaction chamber, you fill it up with debris quite quickly. The lasing media required would have to cool down incredibly quickly between shots as well - Nd Glass lasers just can't do it.

The idea originally was to use lasers to investigate symmetry requirements and capsule implosion and then change to ion beams for actual reactors. Where that plan stands I have no idea. People seem to dodge the question and point to all the interesting physics they're doing in the mean time.

With the Z-pinch it's not so bad. The fuel capsules would be big enough so that you only needed one shot per minute. There's still all the problems with energy extraction though.

 

[Matabiri]

With the Z-pinch it's not so bad. The fuel capsules would be big enough so that you only needed one shot per minute. There's still all the problems with energy extraction though.

I still love the fact that, when it comes to power generation, regardless of how advanced the energy production method is, the most efficient method still comes down to boiling water and pumping steam through a turbine...