24 ms·

2040s? 20+ years from now!?! Too fuckin slow!! For all Musk's faults, he recognizes such timelines are untenable, and pushes people to do the 'impossible'.

howardr 6y ago
I read a letter in the Financial Times over the weekend... > Forty years ago, when studying for my engineering degree, I learnt a rule of thumb that said that nuclear fusion is always 20 years away. I was therefore reassured by the date given in the report (“Sites sought for Step change in energy supply”, December 3) for the Step nuclear fusion plant — 2040. (Edit: Unfortunately it’s behind a paywall but will include the link anyway https://www.ft.com/content/e5af6548-a3a9-451f-8f92-268679d5d7b1 https://www.ft.com/content/e5af6548-a3a9-451f-8f92-268679d5d...)
- garmaine 6y ago
  For most of my life fusion has been perpetually 30 years away. Now it is only perpetually 20 years away. That's progress!
  - gbh444g 6y ago
    So the horizon is approaching us at a what? logarithmic scale? What can we infer from this about the topology of the planning surface?
  - mchusma 6y ago
    Both ITER is scheduled to be done and SPARC are trying to make net energy in 5 years. They probably won't hit those dates, but we aren't really talking about 20 years anymore. (I agree with several commenters on this thread that any project with a 20 years to completion timeline should be shelved for things with <10 year timelines).
    - garmaine 6y ago
      Those are research plants, not regulatory approval for commercial designs. We're still at least 20+ years away.
    - pfdietz 6y ago
      ITER won't even start burning DT until 2035 or so (assuming they control disruptions enough to be allowed to introduce tritium.)
  - wiml 6y ago
    This graph often gets mentioned in the context of "always 30 years away": https://imgur.com/3vYLQmm https://imgur.com/3vYLQmm It's not on "schedule" because we aren't putting the resources towards it.
    - credit_guy 6y ago
      Whenever I see this graph, I wonder what’s supposed to represent. Total world investment in fusion, total world government spending, only US spending, or maybe only the part of the US Department of Energy earmarked for fusion research? It probably is the last one, considering that the enacted 2012 budget of the US DOE for Fusion Energy Sciences was $401 MM ([1], p 16). But then, why is only the US DOE supposed to invest in fusion? In any case, as of 2020, this budget was increased to $671 MM [2]. [1] https://www.energy.gov/sites/prod/files/FY13Highlights.pdf https://www.energy.gov/sites/prod/files/FY13Highlights.pdf [2] https://www.aip.org/fyi/2020/final-fy20-appropriations-doe-office-science https://www.aip.org/fyi/2020/final-fy20-appropriations-doe-o...
    - pstuart 6y ago
      But we can spend ~$40B per year on our misadventures in Afghanistan. It's maddening.
      - frockington1 6y ago
        Not just Afghanistan, Biden we'll get us back to bombing Libya and other nations soon
    - pfdietz 6y ago
      That graph was for a crash program under the assumption that tokamaks work better than it turns out they do. So, if that money had been allocated, it would have been a failure. There was also not the appreciation then of the grave nature of the engineering challenges facing fusion, even if the plasma physics worked wonderfully. The implication that we'd have had fusion if that money had been spent is not supported by the evidence.
    - cbozeman 6y ago
trhway 6y ago
And fortunately soon Musk will need fusion for Mars travel and colonies. He just personally got extra 100B to burn - i think SpaceX will start to hire nuclear scientists and engineers soon :)
- gpm 6y ago
  Fission seems like the better option for mars. The technology is ready now, and no one cares about a bit of nuclear waste on mars anyways.
  - Mediterraneo10 6y ago
    Space probes require a minuscule amount of radioactive fuel, but launching that into space has still occasionally caused concern among environmentalists, as something might go wrong during launch. Can you imagine the resistance if someone wanted to launch enough processed uranium to support a Mars colony?
    - nine_k 6y ago
      If it's a pebble bed design, easily. Pebbles are uranium coated with multiple protective layers, extremely strong. They could be collected and reused if the launch vehicle plainly explodes on the launch pad.
      - WJW 6y ago
        I think the problem is more when it explodes when the rocket is already going mach 10 and the resulting explosion scatters thousands of radioactive pebbles across several thousand square kilometres of Atlantic ocean, possibly hundreds of metres deep.
        Vecr 6y ago
        Why's that a problem? The uranium won't leak out, the ocean will provide enough cooling, and even if it eventually does, it would be slowly and near the bottom, causing no problems at all.
hn_throwaway_99 6y ago
I'm not a fan of this sentiment. Sometimes things are just really f'ing hard, and no amount of "pushing people" is going to change the reality of how long those things take. I mean, if you look at both Tesla and SpaceX, nothing they've done was thought to be "impossible" from the outset from a scientific perspective. Electric cars already existed when Tesla was started, and we've been shooting rockets into space for decades. This isn't at all meant to minimize the huge achievements of those companies, but the science was never really in question. Viable fusion is simply much harder.
- BorisTheBrave 6y ago
  While you are right that Musk favors engineering/cost improvements on existing ideas. I don't think there's a clear dividing line between science and engineering. SpaceX has numberous "firsts", including reusable rockets and some engine designs. So it's not right to say they just do things we've been doing for decades. And on the other end, the physics behind fusion is pretty well understood. The difficulties/expense are in building the thing, and dealing with issues like plasma instability. We call that "science" mostly because it is a state funded project. If a private company were doing the same thing, it would be called R&D.
  - lambdatronics 6y ago
    >We call that "science" mostly because it is a state funded project. Interesting perspective. I agree that arguably most of the difficulties are more in the applied side of things, although calling it engineering might be going too far. The big looming physics uncertainty is crossing the 'burning plasma' threshold, where the plasma becomes dominantly self-heating. There are two aspects to this: (1) will the plasma settle into a nice self-consistent steady state? (2) will the large quantity of fast fusion-born helium nuclei destabilize the plasma in an unexpected way? Theory says it should work, but the proof is in the experiment (which is why SPARC & ITER are being built).
- global-citizen 6y ago
  I totally disagree, but accept that maybe Musk might have not been the best example. Better ones, driven by the government (albeit during a hot and cold war): the Manhattan Project and landing humans on the Moon, respectively. Both had insane timelines, no clear or certainty of success but the pressure, backing and motivation delivered. We need more of this, and yesterday.
dmos62 6y ago
I'd argue that one of Musk's strengths is _not_ doing the very hard things, but finding projects that are relatively low-hanging fruit. Electric cars, multi-use rockets, driving AI. Those aren't the harder problems. Fusion, for example, is on a different level.
Razengan 6y ago
You can’t just get 9 women and make a baby in 1 month.
gwbas1c 6y ago
(I'm not a fusion expert) I get the impression that there's still plenty of details that we don't understand how to do. It's not like solar + batteries where we have plenty of working solar + battery setups, and we just need to figure out how to make more of them more cheaply. It looks like there's still some fundamental research needed before fusion can generate more power than it consumes. Heck, 10-ish years ago, I met someone who told me his cabin in the woods was off-grid solar because an off-grid system was cheaper than running electricity to the cabin. We're not even close to that state with Fusion.
benhurmarcel 6y ago
And yet quite a few Musk promises perpetually planned in the coming years will probably end up with that kind of timeframe too.
- jsight 6y ago
  That's a fair point. I can easily think of several examples too... the boring machine (still slow), self-driving (perpetually delayed), and the hypeloop. I'm sure there are plenty of others.
lambdatronics 6y ago
>2040s? 20+ years from now!?! Too fuckin slow!! I'm impatient too, but perspective is important here. Even the leading private efforts don't have plans to put power on the grid for at least 10 years. Also, before this report, there wasn't an official plan of any sort for the US federal fusion program to get to a pilot plant -- so this is progress. (FWIW, I participated in the early community input stages of this report.) >For all Musk's faults, he recognizes such timelines are untenable, and pushes people to do the 'impossible'. Musk has a good nose for what's possible to commercialize in a medium-term (5-10 year) time horizon on a few $bn budget. I have reason to believe he's considered fusion (he has a background in physics after all). Instead, he's made a play in batteries & solar. Now, if someone had ~$1-5bn to gamble, it might be possible to leapfrog the existing crop of private efforts by ~5 years. Just pick one concept and build the engineering-breakeven experiment (gain ~ 20) without the intermediate scientific-breakeven experiment (gain ~ 1). It would be significantly riskier, but it would save time if it worked.

howardr 6y ago

I read a letter in the Financial Times over the weekend... > Forty years ago, when studying for my engineering degree, I learnt a rule of thumb that said that nuclear fusion is always 20 years away. I was therefore reassured by the date given in the report (“Sites sought for Step change in energy supply”, December 3) for the Step nuclear fusion plant — 2040. (Edit: Unfortunately it’s behind a paywall but will include the link anyway https://www.ft.com/content/e5af6548-a3a9-451f-8f92-268679d5d7b1 https://www.ft.com/content/e5af6548-a3a9-451f-8f92-268679d5d...)

garmaine 6y ago
For most of my life fusion has been perpetually 30 years away. Now it is only perpetually 20 years away. That's progress!
- gbh444g 6y ago
  So the horizon is approaching us at a what? logarithmic scale? What can we infer from this about the topology of the planning surface?
- mchusma 6y ago
  Both ITER is scheduled to be done and SPARC are trying to make net energy in 5 years. They probably won't hit those dates, but we aren't really talking about 20 years anymore. (I agree with several commenters on this thread that any project with a 20 years to completion timeline should be shelved for things with <10 year timelines).
  - garmaine 6y ago
    Those are research plants, not regulatory approval for commercial designs. We're still at least 20+ years away.
  - pfdietz 6y ago
    ITER won't even start burning DT until 2035 or so (assuming they control disruptions enough to be allowed to introduce tritium.)
- wiml 6y ago
  This graph often gets mentioned in the context of "always 30 years away": https://imgur.com/3vYLQmm https://imgur.com/3vYLQmm It's not on "schedule" because we aren't putting the resources towards it.
  - credit_guy 6y ago
    Whenever I see this graph, I wonder what’s supposed to represent. Total world investment in fusion, total world government spending, only US spending, or maybe only the part of the US Department of Energy earmarked for fusion research? It probably is the last one, considering that the enacted 2012 budget of the US DOE for Fusion Energy Sciences was $401 MM ([1], p 16). But then, why is only the US DOE supposed to invest in fusion? In any case, as of 2020, this budget was increased to $671 MM [2]. [1] https://www.energy.gov/sites/prod/files/FY13Highlights.pdf https://www.energy.gov/sites/prod/files/FY13Highlights.pdf [2] https://www.aip.org/fyi/2020/final-fy20-appropriations-doe-office-science https://www.aip.org/fyi/2020/final-fy20-appropriations-doe-o...
  - pstuart 6y ago
    But we can spend ~$40B per year on our misadventures in Afghanistan. It's maddening.
    - frockington1 6y ago
      Not just Afghanistan, Biden we'll get us back to bombing Libya and other nations soon
  - pfdietz 6y ago
    That graph was for a crash program under the assumption that tokamaks work better than it turns out they do. So, if that money had been allocated, it would have been a failure. There was also not the appreciation then of the grave nature of the engineering challenges facing fusion, even if the plasma physics worked wonderfully. The implication that we'd have had fusion if that money had been spent is not supported by the evidence.
  - cbozeman 6y ago

trhway 6y ago

And fortunately soon Musk will need fusion for Mars travel and colonies. He just personally got extra 100B to burn - i think SpaceX will start to hire nuclear scientists and engineers soon :)

gpm 6y ago
Fission seems like the better option for mars. The technology is ready now, and no one cares about a bit of nuclear waste on mars anyways.
- Mediterraneo10 6y ago
  Space probes require a minuscule amount of radioactive fuel, but launching that into space has still occasionally caused concern among environmentalists, as something might go wrong during launch. Can you imagine the resistance if someone wanted to launch enough processed uranium to support a Mars colony?
  - nine_k 6y ago
    If it's a pebble bed design, easily. Pebbles are uranium coated with multiple protective layers, extremely strong. They could be collected and reused if the launch vehicle plainly explodes on the launch pad.
    - WJW 6y ago
      I think the problem is more when it explodes when the rocket is already going mach 10 and the resulting explosion scatters thousands of radioactive pebbles across several thousand square kilometres of Atlantic ocean, possibly hundreds of metres deep.
      - Vecr 6y ago
        Why's that a problem? The uranium won't leak out, the ocean will provide enough cooling, and even if it eventually does, it would be slowly and near the bottom, causing no problems at all.

hn_throwaway_99 6y ago

I'm not a fan of this sentiment. Sometimes things are just really f'ing hard, and no amount of "pushing people" is going to change the reality of how long those things take. I mean, if you look at both Tesla and SpaceX, nothing they've done was thought to be "impossible" from the outset from a scientific perspective. Electric cars already existed when Tesla was started, and we've been shooting rockets into space for decades. This isn't at all meant to minimize the huge achievements of those companies, but the science was never really in question. Viable fusion is simply much harder.

BorisTheBrave 6y ago
While you are right that Musk favors engineering/cost improvements on existing ideas. I don't think there's a clear dividing line between science and engineering. SpaceX has numberous "firsts", including reusable rockets and some engine designs. So it's not right to say they just do things we've been doing for decades. And on the other end, the physics behind fusion is pretty well understood. The difficulties/expense are in building the thing, and dealing with issues like plasma instability. We call that "science" mostly because it is a state funded project. If a private company were doing the same thing, it would be called R&D.
- lambdatronics 6y ago
  >We call that "science" mostly because it is a state funded project. Interesting perspective. I agree that arguably most of the difficulties are more in the applied side of things, although calling it engineering might be going too far. The big looming physics uncertainty is crossing the 'burning plasma' threshold, where the plasma becomes dominantly self-heating. There are two aspects to this: (1) will the plasma settle into a nice self-consistent steady state? (2) will the large quantity of fast fusion-born helium nuclei destabilize the plasma in an unexpected way? Theory says it should work, but the proof is in the experiment (which is why SPARC & ITER are being built).
global-citizen 6y ago
I totally disagree, but accept that maybe Musk might have not been the best example. Better ones, driven by the government (albeit during a hot and cold war): the Manhattan Project and landing humans on the Moon, respectively. Both had insane timelines, no clear or certainty of success but the pressure, backing and motivation delivered. We need more of this, and yesterday.

dmos62 6y ago

I'd argue that one of Musk's strengths is _not_ doing the very hard things, but finding projects that are relatively low-hanging fruit. Electric cars, multi-use rockets, driving AI. Those aren't the harder problems. Fusion, for example, is on a different level.

Razengan 6y ago

You can’t just get 9 women and make a baby in 1 month.

gwbas1c 6y ago

(I'm not a fusion expert) I get the impression that there's still plenty of details that we don't understand how to do. It's not like solar + batteries where we have plenty of working solar + battery setups, and we just need to figure out how to make more of them more cheaply. It looks like there's still some fundamental research needed before fusion can generate more power than it consumes. Heck, 10-ish years ago, I met someone who told me his cabin in the woods was off-grid solar because an off-grid system was cheaper than running electricity to the cabin. We're not even close to that state with Fusion.

benhurmarcel 6y ago

And yet quite a few Musk promises perpetually planned in the coming years will probably end up with that kind of timeframe too.

jsight 6y ago
That's a fair point. I can easily think of several examples too... the boring machine (still slow), self-driving (perpetually delayed), and the hypeloop. I'm sure there are plenty of others.

lambdatronics 6y ago

>2040s? 20+ years from now!?! Too fuckin slow!! I'm impatient too, but perspective is important here. Even the leading private efforts don't have plans to put power on the grid for at least 10 years. Also, before this report, there wasn't an official plan of any sort for the US federal fusion program to get to a pilot plant -- so this is progress. (FWIW, I participated in the early community input stages of this report.) >For all Musk's faults, he recognizes such timelines are untenable, and pushes people to do the 'impossible'. Musk has a good nose for what's possible to commercialize in a medium-term (5-10 year) time horizon on a few $bn budget. I have reason to believe he's considered fusion (he has a background in physics after all). Instead, he's made a play in batteries & solar. Now, if someone had ~$1-5bn to gamble, it might be possible to leapfrog the existing crop of private efforts by ~5 years. Just pick one concept and build the engineering-breakeven experiment (gain ~ 20) without the intermediate scientific-breakeven experiment (gain ~ 1). It would be significantly riskier, but it would save time if it worked.