Slashdot Mirror
Wendelstein 7-X Fusion Reactor Produces Its First Flash of Hydrogen Plasma (gizmag.com)
Zothecula writes: Experimentation with Germany's newest fusion reactor is beginning to heat up, to temperatures of around 80 million degrees Celsius, to be precise. Having fired up the Wendelstein 7-X to produce helium plasma late last year, researchers have built on their early success to generate its first hydrogen plasma, an event they say begins the true scientific operation of the world's largest fusion stellarator.
You should feel whatever you feel, unless you're a robot, in which case: /apply feeling hopeful.
Part of the fusion problem is keeping the hydrogen confined in the plasma. A stellator does this by shaping the magnetic field in such a way that the plasma twists and constricts itself. So instead of constraining a moving plasma, the moving plasma constrains itself.
This requires a precise shaping of the magnetic field via superconducting magnets, and the design of these has only recently become possible with advanced calculations on supercomputers.
So this is a test run of a new kind of fusion reactor. If it works, it will change the world. And so far so good, but we won't know until it works until all following tests succeed.
As a former program officer for the Office of Fusion Energy, US Department of Energy I can assure you even if the Stellarator "works", it will not be a practical source of power. The complex engineering and cost make harvesting energy from fusion impractical.
I could fill a page on enumerating them. For one -- fast neutrons can destroy any material known. No one has come up with a design for the the first wall that captures the neutrons and energy.
The old quip is "Fusion has been 25 years in the future for the last 50 years.
No, "more accurate" would be correct ... an exact temperature would be "precise".
That's kinda-precise-ish in a vague hand-wavy kind of way. Kind of the opposite of "precise".
"around 80 million degrees Celsius, to be precise" is sure as hell NOT precise.
That could be +- 5 million degrees and still be "around".
Lost at C:>. Found at C.
They aren't intending to generate energy with this reactor; the goal is to sustain plasma at temperatures high enough to eventually get to fusion. The article says they are at 80 million deg C, which is about 7 keV. They need to get to 14 keV for a D-T reaction (look at the minimum for the Lawson Criterion) . That's excellent work, and if they can sustain it for thirty minutes, even better. When they are done, the design will be proven and then they can do the harder problem of building a reactor that can withstand the neutrons and recover the heat for a secondary cycle.
"Who are you?" "No one of consequence." "I must know." "Get used to disappointment."
Interestingly enough, for d-t fusion, the neutrons are not an unwanted waste product, but actually essential. Tritium doesn't grow on trees, you have to make it. And more importantly, d-t fusion only gives off one neutron, and it takes one neutron captured by 6Li to breed 1 tritium (you can also make tritium from 7Li bombardment and not consume the neutron, but due to the cross sections and energies involved its usually not as interesting). So if you use one neutron to make the fuel that produces one neutron, and you can't capture 100% of the neutrons, you're in trouble! You get around this by using a lithium-beryllium blanket, as beryllium is a good neutron "multiplier" (capturing one high energy neutron and yielding two lower energy neutrons). It's also rare, expensive as heck and its dusts are highly toxic, but it's consumed at a tiny rate, so it's mainly just an initial cost (heavy elements like lead can also be used as multipliers but they're not very effective in this context, their cross sections don't extend down as far as beryllium and their (n, Xn) reactions where X>2 don't make up for it). So basically, while you lose some neutrons to unwanted reactions, you overall end up producing enough to produce enough tritium for your reactor to consume. The key point is, you want the neutrons to be hitting your reactor, they're doing you a service ;)
There will of course be unwanted neutron captures, but when you engineer it you're choosing specifically what materials are going to be bombarded, so you can pick materials with low neutron capture cross sections and which capture to isotopes that are either stable or have short half lives. Concrete is great for how cheap it is (light elements in general are, and concrete is mostly made of light stuff). As far as metals go, aluminum is great where heat loads or mechanical stresses aren't excessive. Beryllium is even better, as well as stronger and lighter... but see the aforementioned issues with it. Steel is "okay", usually fine if you're careful about what you alloy it with. You generally want to avoid titanium. Graphite is superb if you run it hot enough (otherwise you risk Wigner energy problems). Composites likewise, although they're more temperature limited. Most common ceramics are made of light elements, which makes them very good to use, although those with heavy elements (like tungsten carbide) should be avoided. Tungsten in general should be avoided unless necessary. Some ceramics like boron carbide/nitride are highly heat and corrosion tolerant, high compressive strength, huge neutron absorbers and don't yield dangerous byproducts, which lets them fit multiple roles at once - so long as there's little tensile or shear stresses. In some cases you may want more of a neutron "window", wherein things like zirconium or lead would be good - particularly specific isotopes of them if you're willing to pay for enrichment. It all depends on the operating environment and geometry.
It's times like this I wish I had a friend named 'The Professor'.
There is also the element of funding for R&D. In the late 70's the DoE produced a fusion roadmap based on different funding levels. There was a crash program forcast which would have led to commercial fusion in 10-15 years, a robust development program that would led to fusion in 15-20 years, and a point where if funding remained below a certain level, would never lead to commercial fusion. Guess what funding level was chosen (well below the "fusion never" level). So the joke of "fusion is the technology of the future and always will be", is a result of no real investment being made. Sure ITER may be a $15billion project, but its also a 50 year long project. First announced in 1985, first plasma wont occur till 2025, that's 40 fricken years later, not exactly demonstrative of an intensive focus on developing fusion energy. Compared to what we invest in developing other sources of energy, its chump change
As an engineer working in the fusion field, I would not agree it's quite so rosy a picture. *Lots* of issues need to be solved technologically, although I agree with you the physics side of Tokamaks is relatively understood. I.e. it would be a huge shocker if ITER didn't produce the power expected. Tokamaks are however very unreliable with stability, and whether or not these can be controlled and mitigated enough for reliable power production remains to be seen. Further, going from ITER to DEMO is like launching a rocket to space vs. going to the moon; the high energy neutron flux from a fusion reactor will centimeters of the first wall to powder. Getting enough lithium around the wall for tritium breeding and heat removal for a steam cycle is very difficult.
In the end, it's all economics as you say. I can't imagine with the present state of technology a viable commercial fusion reactor online until past 2100. ITER will be ~2030, DEMO ~2070 if ITER cost/is any clue. Say you're making a decision for a company - would you rather spend $20 billion dollars on a very finicky tokamak fusion reactor with tremendous maintenance costs (tritium recycling, lithium management, disruption and instability mitigation systems, etc.), or a gen III or IV nuclear reactor - perhaps a thorium molten salt reactor - that produces the same power reliably for a small fraction of the cost?
Commercial fusion will happen eventually, but in my opinion not without tremendous advances in materials science and superconducting magnets. One can imagine with clever first wall materials and >20 T fields using advanced BSCCO superconducting materials (or other) a reactor might become as affordable as a fission reactor of the same power output. Contrary to what fusion researchers will have you believe, fusion will always be in economic competition with fission.
This PDF sums it up pretty nicely: http://www.askmar.com/Robert%20Bussard/The%20Trouble%20With%20Fusion.pdf