4 ms·

For reference the temperature of fusion plasmas are in the 100 million degrees C range. So very high ISPs are not that crazy. There are many other problems to s

For reference the temperature of fusion plasmas are in the 100 million degrees C range. So very high ISPs are not that crazy. There are many other problems to solve in building a usable fusion rocket, but high ISP comes for free. One thing you should realize about a fusion reactor is that the easiest (which is still really difficult to do) reactions for fusion release a lot of their power in neutrons. These can not be harnessed directly to produce electricity, unlike moving electrons or ions theoretically could be. They can be used to boil water, just like a coal plant, and use steam to run a generator to make electricity. This part of the power plant is expensive. And the fusion reactor is expensive. You don't just get a Mr. Fusion on your car roof outputting Gigawatts of power. As to it being a great investment, an electric power generation method is only going to be a great one if it is much cheaper than current methods. This is why ITER is such a boondoggle with its $10-$20 billion dollar price tag. It is very unlikely to be the path to putting fusion power onto the grid. Also, who knows if people will be just as afraid of nuclear fusion plants and nuclear fission. If that is the case they will not be economical to build. Maybe a name change is in order? Relativistic magnetohydrodynamics is a hard subject matter. Investors with $30 million dollars to speculate on such a project are not going to know enough to distinguish a possible project from unlikely ones. Physicists who can know enough to see that some plan might work are unlikely to have the funds to invest. I saw Sam Altman had a fusion book on one of his lists at one point and Elon Musk might know a physicist he would trust, but getting those people to take a look is not easy. I'd be happy to make introductions though. :)

philwelch 8y ago
> One thing you should realize about a fusion reactor is that the easiest (which is still really difficult to do) reactions for fusion release a lot of their power in neutrons. These can not be harnessed directly to produce electricity, unlike moving electrons or ions theoretically could be. They can be used to boil water, just like a coal plant, and use steam to run a generator to make electricity. This part of the power plant is expensive. And the fusion reactor is expensive. You don't just get a Mr. Fusion on your car roof outputting Gigawatts of power. I would kind of expect a completely new technology to be expensive to begin with. > As to it being a great investment, an electric power generation method is only going to be a great one if it is much cheaper than current methods. This is why ITER is such a boondoggle with its $10-$20 billion dollar price tag. It is very unlikely to be the path to putting fusion power onto the grid. I'm not sure how that follows--this might not be a great comparison, but the cost of building the first two atomic bombs was orders of magnitude higher than the cost of building the most recent two atomic bombs. Even if it costs $20-50 billion dollars per fusion power plant, it still might work out if each plant can produce enough economy of scale. > Relativistic magnetohydrodynamics is a hard subject matter. Investors with $30 million dollars to speculate on such a project are not going to know enough to distinguish a possible project from unlikely ones. Physicists who can know enough to see that some plan might work are unlikely to have the funds to invest. I saw Sam Altman had a fusion book on one of his lists at one point and Elon Musk might know a physicist he would trust, but getting those people to take a look is not easy. I'd be happy to make introductions though. :) Yeah, that makes sense. There has to be some reason that a $30 million plan to build a proof-of-concept viable fusion reactor hasn't been funded yet.
- njarboe 8y ago
  Tokamak fusion plants 10 times the output of the biggest power plants today are likely to work, but no one would want such huge plants on the grid. When it needs to shut down for maintenance, that amount of power cannot be easily replaced. If it was used only for hydrocarbon synthesis, hydrogen generation, or seawater desalination, this could work, but not for electricity generation.
  - philwelch 8y ago
    > Tokamak fusion plants 10 times the output of the biggest power plants today are likely to work, but no one would want such huge plants on the grid. When it needs to shut down for maintenance, that amount of power cannot be easily replaced. If it was used only for hydrocarbon synthesis, hydrogen generation, or seawater desalination, this could work, but not for electricity generation. The United States has, according to a quick Google, over 8,000 power plants. If you replaced them with 800 fusion plants, wouldn't you still have enough redundancy to take them offline for maintenance? Sure, having a smaller number of power plants increases line loss, but there has to be some economy of scale where tokamak fusion starts winning. Thinking even bigger, if you have a 100x or 1000x scale fusion plant and used it for hydrocarbon synthesis, as you suggest, couldn't you use that plant to supply several downstream power plants that burn hydrocarbons and still, potentially, end up net-ahead? I dunno--my impression is that the theoretical yield of fusion is so high that you can very, very quickly start reaping benefits even if you're doing so in extremely inefficient ways. Once you get above break-even, even if you're only a couple percentage points above break-even, if can build a huge enough reactor, it adds up really, really, really fast. Fast enough that some rich technocratic dictatorship could, say, build a bunch of tokamaks and effectively corner the world oil market without drilling for a single barrel of natural oil. You'd have to think that fusion starts doing things to the cost of energy that Moore's Law did to the cost of computing, to the point where people actually wrap user-facing applications in gratuitous web browsers.
    - baking 8y ago
      You still have the cost of steam turbines and generators just like a coal/oil/gas power plant. The main cost savings is the fuel cost, which is a small but significant fraction of today's electric power cost. There are theoretical proposals for direct conversion of electricity using the charged particles of certain fusion reactions (usually using Helium-3 mined from the moon) in different reactor designs but those are even further out than the toroidal magnetic confinement reactors (tokamaks, spherical tokamaks, and stellarators.)

philwelch 8y ago

> One thing you should realize about a fusion reactor is that the easiest (which is still really difficult to do) reactions for fusion release a lot of their power in neutrons. These can not be harnessed directly to produce electricity, unlike moving electrons or ions theoretically could be. They can be used to boil water, just like a coal plant, and use steam to run a generator to make electricity. This part of the power plant is expensive. And the fusion reactor is expensive. You don't just get a Mr. Fusion on your car roof outputting Gigawatts of power. I would kind of expect a completely new technology to be expensive to begin with. > As to it being a great investment, an electric power generation method is only going to be a great one if it is much cheaper than current methods. This is why ITER is such a boondoggle with its $10-$20 billion dollar price tag. It is very unlikely to be the path to putting fusion power onto the grid. I'm not sure how that follows--this might not be a great comparison, but the cost of building the first two atomic bombs was orders of magnitude higher than the cost of building the most recent two atomic bombs. Even if it costs $20-50 billion dollars per fusion power plant, it still might work out if each plant can produce enough economy of scale. > Relativistic magnetohydrodynamics is a hard subject matter. Investors with $30 million dollars to speculate on such a project are not going to know enough to distinguish a possible project from unlikely ones. Physicists who can know enough to see that some plan might work are unlikely to have the funds to invest. I saw Sam Altman had a fusion book on one of his lists at one point and Elon Musk might know a physicist he would trust, but getting those people to take a look is not easy. I'd be happy to make introductions though. :) Yeah, that makes sense. There has to be some reason that a $30 million plan to build a proof-of-concept viable fusion reactor hasn't been funded yet.

njarboe 8y ago
Tokamak fusion plants 10 times the output of the biggest power plants today are likely to work, but no one would want such huge plants on the grid. When it needs to shut down for maintenance, that amount of power cannot be easily replaced. If it was used only for hydrocarbon synthesis, hydrogen generation, or seawater desalination, this could work, but not for electricity generation.
- philwelch 8y ago
  > Tokamak fusion plants 10 times the output of the biggest power plants today are likely to work, but no one would want such huge plants on the grid. When it needs to shut down for maintenance, that amount of power cannot be easily replaced. If it was used only for hydrocarbon synthesis, hydrogen generation, or seawater desalination, this could work, but not for electricity generation. The United States has, according to a quick Google, over 8,000 power plants. If you replaced them with 800 fusion plants, wouldn't you still have enough redundancy to take them offline for maintenance? Sure, having a smaller number of power plants increases line loss, but there has to be some economy of scale where tokamak fusion starts winning. Thinking even bigger, if you have a 100x or 1000x scale fusion plant and used it for hydrocarbon synthesis, as you suggest, couldn't you use that plant to supply several downstream power plants that burn hydrocarbons and still, potentially, end up net-ahead? I dunno--my impression is that the theoretical yield of fusion is so high that you can very, very quickly start reaping benefits even if you're doing so in extremely inefficient ways. Once you get above break-even, even if you're only a couple percentage points above break-even, if can build a huge enough reactor, it adds up really, really, really fast. Fast enough that some rich technocratic dictatorship could, say, build a bunch of tokamaks and effectively corner the world oil market without drilling for a single barrel of natural oil. You'd have to think that fusion starts doing things to the cost of energy that Moore's Law did to the cost of computing, to the point where people actually wrap user-facing applications in gratuitous web browsers.
  - baking 8y ago
    You still have the cost of steam turbines and generators just like a coal/oil/gas power plant. The main cost savings is the fuel cost, which is a small but significant fraction of today's electric power cost. There are theoretical proposals for direct conversion of electricity using the charged particles of certain fusion reactions (usually using Helium-3 mined from the moon) in different reactor designs but those are even further out than the toroidal magnetic confinement reactors (tokamaks, spherical tokamaks, and stellarators.)