We have passed "break even" conditions in a reactor for D-T for a few seconds. The real test is getting useful amounts of net energy generation which requires about 5x break even for several minutes.
IMO, ITER's design should get to that point, but I don't think fusion is going to cost less than solar let alone coal for a long time.
PS: Using lasers to create fusion has always been more about simulating bombs than it was creating useful amounts of energy.
You have achieve breakeven more broadly -- over the entire pulse, or over the electricity you pull from the grid, or perhaps most properly, including all of the material and labor costs as compared to just taking heat from the fusion reactor in the sky with wind turbines and solar collectors.
As for the simulation of bombs vs power, I think it depends on who you ask; scientists or DOE project managers?
It's interesting, after having left a PhD program in plasma physics (for fusion power,) I have a very different sense of the commercially plausible. All we're really using the fusion for is to create heat, and there are many ways to get heat inexpensively (solar, geothermal, coal, natural gas) that don't involve temperatures of hundreds of millions of degrees, that don't require superconducting coils, liquid helium cooling systems, ultra-high precision machined tungsten tiles, structural materials that are vaguely defined, completely unsourced, untested at the correct neutron fluxes, and rely upon rare earth elements, liquid lithium systems from which we're supposed to do tritium recovery (lithium is really reactive, tritium bubbles up and is explosive and diffuses through walls and embrittles them), liquid lithium walls on the inside of the chamber; high vacuum vacuum pumps, giant particle beams and incredibly non-linear and poorly understood science. You're comparing that to coal and mirrors and holes drilled in the ground and wind, all of which ultimately go into some kind of turbine (which is half or most of the expense!) The benefit of fusion just doesn't seem worth it for anything other than military applications or interstellar spacecraft. And really, we've got a lot to worry about before then...
As I recall the actual limitations where two fold, it's a research device so they did not want the extra radiation produced from an actual DT reaction so they stuck with DD. And it was cheaper to build it as a pulsed device and let it cool down, vs having an active cooling system.
From what I can tell the problem over the last 25 years with fusion is people keep saying what is the least amount of money we can spend and still make something useful. But, actually building a fusion reactor costs ~20 billion so we have been stuck prototypes when we need to shoot for the moon.
Also http://en.wikipedia.org/wiki/JT-60 "During deuterium (D–D fuel) plasma experiments in 1998 plasma conditions were achieved which would, if the D–D fuel were replaced with a 1:1 mix of deuterium and tritium (D–T fuel), have exceeded break-even"
I would like to argue that prototypes have been what's called for. We don't even have a clear idea of what a commercial fusion reactor would look like. The cooling and tritium production issues, the structural materials, and the lifetimes required of the super cooling and superconducting systems, are massive problems that haven't really been solved yet.
It's possible that replacing D-D with D-T would have yielded break-even for a few seconds, (I've heard that claim before too) but I'm very skeptical; increasing the internal heat generation is sure to change the plasma conditions: for example, there would be a lot of adiabatic expansion, which, even restrained against the tokamak fields, is liable to have a variety of unstable modes.
I feel that with a modular design that includes remote handling capability you could test things out a lot faster on a full scale device. As to internal heating issues, I don't think there is much problem extrapolating from Q0.7 at JET to Q1.2 at JT-60.
Large Plasma devices are "the sexy" but IMO it's just another engineering problem at this point. And you don't design large scale power plants to sit at the outer edge of their capabilities 24x7. So yes, the highest energy pulses are unable after a few seconds, but we have built steady state fusion reactors they just operate further from their limits.
PS: Finding T is a problem, but "just build it bigger" and suddenly D-D (or 75%D - 25%T etc) becomes reasonable.
Edit: I am suggesting building building something with 4x ITER's budget which should give you ~8x the plasma volume and a lower surface area to volume ratio etc. We can also pump of the field strength etc. But you are limited to how much heat the wall can take so the plasma conditions don't need to be as efficient.
I do agree that the problem of controlling Q > 1 plasma becomes easier as you increase the volume. However, I don't think that's the biggest problem. As you yourself point out, you're limited in how much heat the wall can take. This is fundamentally the limiting factor in fusion power plant designs, unless a much more efficient direct conversion scheme is feasible. Why not instead take the tungsten panels you'd use and point a bunch of mirrors at it, achieving the same heat flux? Mirrors are just not that expensive.
PS: Do you work in fusion, by the way?
PPS: We should chat more. Send me a message at dfong at lightsailenergy.com
Out of interest Dani, is it possible to use a relatively benign lithium salt solution rather than pure lithium to do the tritium recovery?
As for the geothermal alternative, unfortunately it seems drilling holes to geothermal depths isn't easy or cheap either. Geodynamics in South Australia are arguably at the most advanced stage and they've had multiple drilling failures at their hot fractured rock geothermal pilot plant. They originally claimed that hot fractured rock geothermal drilling was in the regular 3-5km oil and gas depth "window" and thus should be straight forward. Ubiquitous geothermal from the mantle is much deeper still, so I suspect as a viable alternative to fusion you'd just be replacing one multi-billion dollar engineering challenge with another!
Btw, on the topic of plasma physics, do you think the type of life ("self-organizing structures") proposed here possibly exists at all in the universe or is even somewhat common? http://dl.dropbox.com/u/101698/PHU_50_4_R11.pdf
It's an extremely interesting idea. However I'm not sure if the numbers work out. Most of the stable 'bits' are to be stored in magnetic topologies. Unfortunately these topological structures are spaced at extremely distances from each other, further more they decay and magnetic field lines close in time spans much shorter relative to self-organizing/replicating topological interactions than, for example, DNA base pairs.
I was doing research on a kind of self-organizing plasma structure called an FRC. One of our big results was that we could stabilize the FRC to last for 100 milliseconds before it unwound itself. We have to pump a fair amount of energy in to make it in the first place!
Well, then you just have to invent a fusion reactor that:
- does not work by creating heat but by creating electricity directly (I am sure fusion creates plenty of high energy charged particles to make some electricity), and
- does not require any isotopes but fuses ordinary and plentiful atoms, such as ordinary hydrogen.
Well you are the scientist, get to it. After there are many alternatives but we all know their limitations. Regarding fusion, we know there is a potential for incredible amounts of energy that do not suffer most of the limitations of the other methods.
To your first idea, people have thought of this but unfortunately using ordinary plasma potentials isn't effective enough to drive a current, we have no proven concepts allowing for inductive generation, extraction using photovoltaics is a practical impossibility with decent lifetimes due to the temperature and neutron flux, and other methods of current extraction (e.g. alpha channeling) are very theoretical and still have huge amounts of science around required around them and are limited in efficiency due to a variety of entropy generating processes.
As to your second idea, it's already extremely hard to achieve break even with tritium, D-D fusion is even tougher and P-P fusion much much much tougher. We're now talking billion dollar projects, all to achieve an inexhaustible source of heat, which we effectively have in the form of the sun already!
IMO, ITER's design should get to that point, but I don't think fusion is going to cost less than solar let alone coal for a long time.
PS: Using lasers to create fusion has always been more about simulating bombs than it was creating useful amounts of energy.