rootone said:
We know it is possible.
If nature would stop playing hardball with plasma containment;
I probably need to consult a solicitor, nature can be a *****
As a solicitor, I can assure you that it wouldn't help.
But, the knowing it is possible part is the rub, isn't it.
The Sun has managed it every day for four billion years without so much as instructions or metal tools with temperatures far cooler than the ones we can produce in our machines. Countless billions of stars have been doing it for 13+ billion years on the same basis. We have built stockpiles of H-bombs large enough to destroy the planet (or maybe a marauding comet fragment headed our way) many times over in uncontrolled fusion reactions. The LHC collides atoms to make all kinds of neat stuff, every once in a while resulting in nuclear fusion, in a controlled fashion millions of times a year.
All the relevant laws of physics have been staring us in the face almost fifty years, and we've measured all the physical constants we need to calculate with to more than adequate precision. We can split atoms and produce controlled nuclear power just fine thank you very much. Hell, college kids can make functional (if not very safe) fission reactions in their garages with spare parts, their laptops and a few textbooks and journal articles. We can assemble the optimal kinds of fuels with extraordinary purity without any great difficulty - scarcely more difficult than ordering a best seller on Amazon. We have nice fat budgets devoted to these projects - charges so big that even Donald Trump would notice them on his monthly credit card statement. We know exactly how much energy is produced every time we fuse atoms. We can go from zero to plasma in a few months. And yet, getting that final step to a sustainable, net energy producing reaction just isn't happening, despite all of our best efforts for roughly seventy years of trial and failure. The theory is all there and the engineering isn't happening.
Even worse, everybody out their in the bleacher seats thinks that if we can produce a sustainable, net energy producing fusion reaction that it is game over and we've won. Jetsons and Iron Man, here we come. But, it isn't that easy. It isn't good enough to produce a sustainable, net energy producing fusion reaction that is safe, has virtually free fuel costs, and is almost fully automated while producing no air pollution, no water pollution except waste heat, and no long term nuclear waste that has to be stored somewhere.
That sounds great, but we have to achieve that at about 10 cents per kilowatt hour, or we've just built an expensive curiosity that doesn't make anyone better off any more than a fine piece of sculpture does. A 1000 Megawatt fusion reactor running 100% of the time produces 8,766,000,000 Kilowatt hours a year, so you need to have a total cost of $876,600,000 per year or less for your fusion reactor to be economically competitive. Even if you can borrow the entire cost of the plant at 3% interest per year, with free labor and free fuel and no new power grid improvements, you need to build your commercial sized nuclear fusion power plant for $29.22 billion. And realistically, since labor and fuel and insurance and lawyers (yeah, eventually you'd need a solicitor, just not quite yet) to get it approved by the government, and power grid improvements aren't free, and you can't run at full capacity absolutely 100% of the time, and you have to pay off the principal costs of buying it over some period of time less than or equal to its useful life, and somebody would like to make a profit and lower energy costs for the people from the greatest and most anticipated invention ever, you really need to get the cost of a commercial sized nuclear fusion power plant down to $20 billion or less all in (including its share of R&D costs over the total number of plants built). And, the better solar and wind and tidal get, the smaller your budget gets. This is about the same as the cost of building 10,000 nuclear fusion warheads (without delivery systems).
At this point, the physics and nuclear engineering starts to matter a lot again. If your 1000 megawatt fusion reactor is running at just 1% efficiency relative to the theoretical maximum that can scale (which would be better than any previous effort the human race has made to build a fusion reactor), it is going to be a big, expensive machine (although still probably physically smaller than a conventional fossil fuel plant). If you get it up to 20% efficiency you can spread your $20 billion construction budget over a lot less machine. You can't even begin to compare p-Be v. D He3 approaches until you can figure out how efficient you will be in each approach, as mfb aptly notes, and not just the theoretical maximum output of the reaction on paper. Likewise, it really, really matters if the useful life of your reactor is 10 years or 50 years. If the process degrades the reactor badly enough that it has to be overhauled regularly, you've paid a high price in minimum efficiency to be cost effective with that tough on the machine process.
And, unless you can come up with a way to squeeze the whole process under the hood of an automobile, you'll still be struggling to use your fusion reactor to power electric cars until you also come up with a better battery as well. (We'd like to hope that that problem, which certainly seems less intractable, will be solved in the meantime.)
The fact that lots of very smart and well funded people have been trying to perfect the tokamak for decades also isn't encouraging, because it suggests that most of the low hanging fruit to improve the process has already been exploited without even managing to get the approach across the finish line. Generally there are diminishing marginal returns to additional research, not increasing ones (integrated circuits being the rare exception).
My personal intuition is that it is going to take someone going a completely different direction than what we've been trying for the last 70 years, that nobody has even considered or that somebody ruled out early on, perhaps because they didn't see a loophole that could make another approach work, before we get commercially viable nuclear fusion power generation. I just don't see how a tinker here and a tinker there from the same general approach we've been trying for a long time is going to improve enough (at least in the next half century) to become commercially viable. Maybe it would all work better at a nano-scale. Maybe there is an easier way to fuse atoms that don't happen to have the optimal energy release per reacting atom which has been the basic approach so far, even though just about any fusion reaction of any combination of atoms is still wildly energy generating (which I guess is the point of the p - Be approach). Maybe there's some crazy reason that a double helix design would work better than the tried and true torus. Hell if I know. But, I've got to think at this point that we need to try something totally different and that the Siren call of the latest project proposal along the same old lines should be ignored. Maybe we need to act like Edison and just try everything, even if it costs a boatload of money, research efficiency be damned.
Of course, the good news is that it is a problem that only has to be solved once. Twenty-one years after the design is perfected, when the patent expires, a proven solution to the greatest engineering problem of all time will be in the public domain and available for everyone for ever after and we can all start living the post-scarcity Star Trek life.
