4 ms·

Is the Age of Fusion Upon Us?

12 points by nsoonhui 4y ago

burlesona 4y ago
Great article. It’s both a long write up and also an effective short summary of the state of fusion systems and how far we have left to go.
jjk166 4y ago
I think a big issue with any discussion of Fusion is that the popsci narrative and the actual research are at a huge disconnect about what the fusion age means. As the article touches upon, fusion is often touted as the future of global energy production which is somehow immune to all engineering and economic realities beyond the physics of achieving fusion reactions, and it will give us limitless free energy in perpetuity. This narrative being obviously false, there are plenty who throw the baby out with the bathwater - if it can't produce free unlimited clean energy then what's the point? The fact is that fusion is not going to be powering your home any time soon. It's a big jump from scientific breakeven to engineering breakeven, a larger jump to commercial breakeven, and a still larger jump to "too cheap to meter." Realistically, scientific breakeven may happen in this decade, engineering breakeven won't be till the second half of this century, commercial breakeven might not be possible with DT fusion and free fusion power is just a pipe dream. Certainly on the timescales most of us care about we're not going to see fusion power plants in every city replacing all other options. But no technology needs to be the one true power source. It's fine if fusion just occupies a niche. And that niche may not even be within the realm of electricity production. The major advantage of nuclear power and especially fusion is the extreme energy density of its fuel. This is irrelevant for a commercial power plant, but of great importance where weight is an issue. There are lots of cases like spacecraft, ships, remote locations, etc where the advantages of a nuclear reactor would outweigh the increased cost if you just didn't need to worry about something going wrong with a lot of uranium, which is exactly what fusion offers. Beyond even power production, fusion's main engineering concern can actually be turned into a useful product as an extremely intense neutron source (indeed fusion neutron sources are already a commercially available product, though they currently use much less efficient fusion methods). Process heat, isotope production, hybrid fission-fusion reactors, x-ray sources, there are many ways for fusion to prove useful before commercial or even engineering breakeven. But even this is missing the point. The fact is fusion is something we know we are unable to do. Like an undeciphered ancient language or an unclimbed mountain, there may be nothing of practical utility at the end of the journey, but what kind of society would we be if we didn't try to tackle the problem? By setting our sights on fusion we have committed ourselves to building better super conducting magnets, better vacuum chambers, better lasers, better micromanufacturing methods, better alloys, better simulations, and many more. We are viewing a realm of physical reality that previously has only ever been observed in nuclear tests, and even then only from a distance. Where we are at right now is still very much in the science experiment realm. It's difficult to put a dollar value on the benefits of scientific research, but for things of similar expense to fusion experiments like large telescopes, particle accelerators, mars rovers, etc we don't typically have to. No one expects any return on their investment in the James Webb Space Telescope, no one expects it to solve any worldwide problems, we're not waiting for the telescope age to be upon us, its only mission is to satisfy our curiosity. Once you realize that the goal of these experiments is to learn, then the fact that every time we build a new fusion reactor we find some new quirk of plasma physics that pushes economic power production further out of reach stops being a failure and becomes a ringing endorsement for further research.
ohiovr 4y ago
I was going to post this as an ask hn question but since this article got a lot of eyeballs I thought I'd just post it here in the comments: Ask HN: Is ITER built to last? ITER is a demonstration model of a fusion power plant that is expected to yield enough energy to be comparable to utility scale power. That it may do, or may get close, or lead to a discovery that could make a derivative design mostly practical for these ends. ITER is an experiment so they don’t need to worry about utility hookups quite yet. Why fusion energy? Fuel abundance is the selling point more than any other because water contains a great deal of fusion fuel in the small amount of deuterium scattered in it. Extracting deuterium is safer, cleaner, and cheaper than handling uranium or thorium ores and refining them into purified metals. Enrichment is very high technology and an industrial accident is extremely deadly. Deuterium as a fuel is inert and safe. Tritium is like deuterium but unstable because a lone proton argues with two wives. But we need it’s instability and reactiveness for ITER. Deuterium and Deuterium fusion is so hard to pull off we can’t imagine how to at this time (besides bombs) in amounts that are interesting. Unfortunately at this time Tritium can only be acquired in large quantities from fission but we had enough for experiments. ITER address this with tritium breeding blankets made of lithium. A 17 mev neutron slams through a very hot bottle of liquid lithium and hopefully bounces around long enough to thermalize to where it can be absorbed by the lithium contained in the bottle. Upon merger of lithium and the neutron, lithium will sometimes break down into Tritium which we need for fuel. So far so good. Big deal, it converts neutrons to high test nuclear fuel. Whats the problem? Fusion experiments (where fusion is being made in any significant amount) have usually been very short lived. So they haven’t really had to worry very much about what happens to those neutrons slamming everywhere at velocities impossible to imagine. Thinking here, what kind of neutron flux are we dealing with? Beyond the napkin calculations I reason a few points to consider. First what kind of neutron flux is expected with current generations of fission plants? Well fission of uranium is a well known science and it turns out that most of the energy carried away from the process is in the kinetic energy found in the fragments. Only 30 Mev on average from a 200 Mev fission event is something in the form other than fast moving fragments. Per event it maybe similar to the energy fusion neutrons, I’ve not really dug deep into that. But overall, neutron kinetic energy is less than 15% of the total. But unfortunately for us, fusion of Tritium and Deuterium's energy is basically all in the kinetic energy of the neutron. So comparatively speaking, fusion plants will have to deal with a neutron flux about 6.7 times higher per unit energy delivered. Looking at fission reactors, ideally, neutrons don’t really get much farther than the fuel and the pressureized coolant and they are all submerged together. For safety reasons, concrete and lead can line the pressure vessel from the outside to prevent the from baddies getting out. And they are very bad, transmuting anything unlucky to brush up against a slow one. And that brings me back to thinking about a practical fusion reactor. Sure it can stop most of those neutrons in a very very thick blanket of lithium but anything found before contents of that bottle is going to be blasted with a very very high amount of fast transmuting particles every minute, and every hour of the day. Eventually the structure and the magnets are going to be touched by some of these and cause problems. How long would it take? I can’t estimate that really. And maybe this is is all wrong and I’m not seeing this clearly. But I’ve not seen this discussion about the sheer amount and power of the neutron flux. Its hard to imagine anything in there lasting very long. Transmutations have effects on grain boundries and can lead to lattice dislocations. I am guessing that a significant mass fraction of the reactor may have to be scrapped and condemned within the first year of its tests. Is there an answer to this? I don’t think handling neutrons this way is going to work out. Hotter reactors can burn pure deuterium but that also leads to significant neutron flux. Finding a material that is ideal as a shield that has the right magnetic permeability and other properties would be hard. Not to mention in general it is kind of out of this world trying pure isotope fusion. It just gets so much harder beyond that with exotics like boron proton fusion that really is kind of an odd way of fissioning light nucli with protons. ITER will probably do what the scientists expect and they might be happy at first. I hope they thought of the long term. Could be I’m behind the times and someone already leap frogged ITER or solved it within the original plan. But I’ve not been up on the latest.

12 points by nsoonhui 4y ago

Great article. It’s both a long write up and also an effective short summary of the state of fusion systems and how far we have left to go.

jjk166 4y ago

I think a big issue with any discussion of Fusion is that the popsci narrative and the actual research are at a huge disconnect about what the fusion age means. As the article touches upon, fusion is often touted as the future of global energy production which is somehow immune to all engineering and economic realities beyond the physics of achieving fusion reactions, and it will give us limitless free energy in perpetuity. This narrative being obviously false, there are plenty who throw the baby out with the bathwater - if it can't produce free unlimited clean energy then what's the point? The fact is that fusion is not going to be powering your home any time soon. It's a big jump from scientific breakeven to engineering breakeven, a larger jump to commercial breakeven, and a still larger jump to "too cheap to meter." Realistically, scientific breakeven may happen in this decade, engineering breakeven won't be till the second half of this century, commercial breakeven might not be possible with DT fusion and free fusion power is just a pipe dream. Certainly on the timescales most of us care about we're not going to see fusion power plants in every city replacing all other options. But no technology needs to be the one true power source. It's fine if fusion just occupies a niche. And that niche may not even be within the realm of electricity production. The major advantage of nuclear power and especially fusion is the extreme energy density of its fuel. This is irrelevant for a commercial power plant, but of great importance where weight is an issue. There are lots of cases like spacecraft, ships, remote locations, etc where the advantages of a nuclear reactor would outweigh the increased cost if you just didn't need to worry about something going wrong with a lot of uranium, which is exactly what fusion offers. Beyond even power production, fusion's main engineering concern can actually be turned into a useful product as an extremely intense neutron source (indeed fusion neutron sources are already a commercially available product, though they currently use much less efficient fusion methods). Process heat, isotope production, hybrid fission-fusion reactors, x-ray sources, there are many ways for fusion to prove useful before commercial or even engineering breakeven. But even this is missing the point. The fact is fusion is something we know we are unable to do. Like an undeciphered ancient language or an unclimbed mountain, there may be nothing of practical utility at the end of the journey, but what kind of society would we be if we didn't try to tackle the problem? By setting our sights on fusion we have committed ourselves to building better super conducting magnets, better vacuum chambers, better lasers, better micromanufacturing methods, better alloys, better simulations, and many more. We are viewing a realm of physical reality that previously has only ever been observed in nuclear tests, and even then only from a distance. Where we are at right now is still very much in the science experiment realm. It's difficult to put a dollar value on the benefits of scientific research, but for things of similar expense to fusion experiments like large telescopes, particle accelerators, mars rovers, etc we don't typically have to. No one expects any return on their investment in the James Webb Space Telescope, no one expects it to solve any worldwide problems, we're not waiting for the telescope age to be upon us, its only mission is to satisfy our curiosity. Once you realize that the goal of these experiments is to learn, then the fact that every time we build a new fusion reactor we find some new quirk of plasma physics that pushes economic power production further out of reach stops being a failure and becomes a ringing endorsement for further research.

ohiovr 4y ago

I was going to post this as an ask hn question but since this article got a lot of eyeballs I thought I'd just post it here in the comments: Ask HN: Is ITER built to last? ITER is a demonstration model of a fusion power plant that is expected to yield enough energy to be comparable to utility scale power. That it may do, or may get close, or lead to a discovery that could make a derivative design mostly practical for these ends. ITER is an experiment so they don’t need to worry about utility hookups quite yet. Why fusion energy? Fuel abundance is the selling point more than any other because water contains a great deal of fusion fuel in the small amount of deuterium scattered in it. Extracting deuterium is safer, cleaner, and cheaper than handling uranium or thorium ores and refining them into purified metals. Enrichment is very high technology and an industrial accident is extremely deadly. Deuterium as a fuel is inert and safe. Tritium is like deuterium but unstable because a lone proton argues with two wives. But we need it’s instability and reactiveness for ITER. Deuterium and Deuterium fusion is so hard to pull off we can’t imagine how to at this time (besides bombs) in amounts that are interesting. Unfortunately at this time Tritium can only be acquired in large quantities from fission but we had enough for experiments. ITER address this with tritium breeding blankets made of lithium. A 17 mev neutron slams through a very hot bottle of liquid lithium and hopefully bounces around long enough to thermalize to where it can be absorbed by the lithium contained in the bottle. Upon merger of lithium and the neutron, lithium will sometimes break down into Tritium which we need for fuel. So far so good. Big deal, it converts neutrons to high test nuclear fuel. Whats the problem? Fusion experiments (where fusion is being made in any significant amount) have usually been very short lived. So they haven’t really had to worry very much about what happens to those neutrons slamming everywhere at velocities impossible to imagine. Thinking here, what kind of neutron flux are we dealing with? Beyond the napkin calculations I reason a few points to consider. First what kind of neutron flux is expected with current generations of fission plants? Well fission of uranium is a well known science and it turns out that most of the energy carried away from the process is in the kinetic energy found in the fragments. Only 30 Mev on average from a 200 Mev fission event is something in the form other than fast moving fragments. Per event it maybe similar to the energy fusion neutrons, I’ve not really dug deep into that. But overall, neutron kinetic energy is less than 15% of the total. But unfortunately for us, fusion of Tritium and Deuterium's energy is basically all in the kinetic energy of the neutron. So comparatively speaking, fusion plants will have to deal with a neutron flux about 6.7 times higher per unit energy delivered. Looking at fission reactors, ideally, neutrons don’t really get much farther than the fuel and the pressureized coolant and they are all submerged together. For safety reasons, concrete and lead can line the pressure vessel from the outside to prevent the from baddies getting out. And they are very bad, transmuting anything unlucky to brush up against a slow one. And that brings me back to thinking about a practical fusion reactor. Sure it can stop most of those neutrons in a very very thick blanket of lithium but anything found before contents of that bottle is going to be blasted with a very very high amount of fast transmuting particles every minute, and every hour of the day. Eventually the structure and the magnets are going to be touched by some of these and cause problems. How long would it take? I can’t estimate that really. And maybe this is is all wrong and I’m not seeing this clearly. But I’ve not seen this discussion about the sheer amount and power of the neutron flux. Its hard to imagine anything in there lasting very long. Transmutations have effects on grain boundries and can lead to lattice dislocations. I am guessing that a significant mass fraction of the reactor may have to be scrapped and condemned within the first year of its tests. Is there an answer to this? I don’t think handling neutrons this way is going to work out. Hotter reactors can burn pure deuterium but that also leads to significant neutron flux. Finding a material that is ideal as a shield that has the right magnetic permeability and other properties would be hard. Not to mention in general it is kind of out of this world trying pure isotope fusion. It just gets so much harder beyond that with exotics like boron proton fusion that really is kind of an odd way of fissioning light nucli with protons. ITER will probably do what the scientists expect and they might be happy at first. I hope they thought of the long term. Could be I’m behind the times and someone already leap frogged ITER or solved it within the original plan. But I’ve not been up on the latest.