Cracking the Fusion Nut

[moffy]

I watched this video:
about a group that are shooting copper discs at a plastic cube about 1cm3, with a bubble of hydrogen/helium in it. They manage to create a fusion reaction without the need for lasers or magnetic containment toroids. Of course the energy out is a lot less than the energy in, but it looks promising.
Excuse me if you have already seen the video.

[T3sl4co1l]

What's old is new again?  Shiva Star I believe is the project that comes to mind.

Tim

[I wanted a rude username]

Is it just me or do all the inertial methods seem impractical for energy production?

[moffy]

I don't know about how impractical the kinetic method is. Maybe in my naivety I am just seeing it as being much simpler than toroidal magnetic containment of a 100 Million C plasma. They only recover 80 percent of the energy, that leaves 20 percent losses. Guess like most things, it's the detail that will make or break it.

[Amper]

Pretty much all pulsed fusion reactors like z-pinch are at best useful for research and understanding but most of them are actually built for weapons research. They all have some sort of "we will revolutionize the world" story to get funds but then later on after doing a few shots are modified for mil applications, laser fused and scalable fusion bombs. Continuous is the only way that actual useful power production may be done.

If you are just after fusion, farnsworth fusor or beam on target is the way to go, you will have a nice neutron source but no energy production.

[donotdespisethesnake]

It might surprise people, but creating a fusion reaction is quite easy. You just need temps of 150 million degrees and a little bit of hydrogen and there you go.

Sustaining that reaction and turning it into a practical source of energy production is a completely different thing. This is a stark contrast to fission, which sustains by itself, the tricky part of fission is preventing runaway.

Although it looks quite steampunk, I think these guys may have a chance.
 

[Amper]

There is no way it will work continuously for years at a time...

[ejeffrey]

It's not that hard to get a small amount of fusion.  The problem is that one way or another you have to add a ton of energy to the fuel.  The problem is that you need to have a moderately high probability for any given particle fusing, otherwise you don't get enough energy back to pay for the initial investment.

The problem with magnetic confinement is that you have to hold onto that hot plasma for a long time and it is continually losing heat through non-fusion mechanisms like bremsstrahlung radiation.  That also makes anything but D-T fusion almost certainly impractical.

The good thing about intertial confinement is that once ignited the fusion reaction itself provides heat to increase the temperature further and proceeds as a supersonic shock wave so that you don't have to confine the plasma at the burning temperature.  The downside is mostly the rep rate.  Old systems like NIF using flashlamp pumped lasers are really slow and and really inefficient.  A more modern system using DPSS lasers could reach much higher wall plug efficiency and also much higher rep rate.  However, it is still going to be hard to get the rep rate up.  In addition, intertial confinement requires precision manufactured fuel pellets -- they need to be nearly perfectly spherical to implode properly.  It's not clear that cost can ever be acceptable. 

[Alex Eisenhut]

The problem is how you get to boil water to spin a turbine. For all the gee-whiz physics at the end of the day this isn't Star Trek, we need heat to boil water. The energy from fusion flies out as neutrons and you need to capture them and slow them to heat something up.

That's actually more complex than the fusion part and I have yet to see anything seriously address that, at the "There is no way it will work continuously for years at a time..." level Amper said. 

General Fusion seems to combine the fusion with the capturing part so I have some hope for them. Plus they're Canadian. 

Nobody took my idea seriously, to blow up hydrogen bombs to spin wind turbines. 

[duak]

I used to work with some of the people that are now at General Fusion.  In particular, one of founders, who is a very smart and pragmatic physicist.  I'm not a nukular fizisist so I can't make a judgement on the technical merits.  My personal opinion is that they won't succeed with repeatable break even fusion, but they have made significant numbers of neutrons so they might be on to something. 

[donotdespisethesnake]

I have a hunch that fusion will never be a practical commercial proposition. The reality may never match the dream. We may be able to achieve something a little better than existing sources but at a greater capital cost. Investors don't like putting all their eggs into one basket, getting funding is already an issue for nuclear fission.

I am quite certain that the path being followed by ITER will not lead to a practical commercial reactor. The field is open for a different approach, I have trouble seeing what that will be. Wrangling a hot plasma seems like something that is never going to be easy.

I am also pretty sure that I won't live long enough to see the result.

[ejeffrey]

Getting the energy out isn't really that hard.  It's obviously not entirely trivial but if you can get the fusion part working with enough heat output building a power plant around it is not going to be the obstacle.

[Nerull]

There's an entire group of hobbyists who build fusion reactors in their garages. Fusion is easy, useful fusion is really, really hard.

[donotdespisethesnake]

Getting the energy out isn't really that hard.  It's obviously not entirely trivial but if you can get the fusion part working with enough heat output building a power plant around it is not going to be the obstacle.

Most of the energy output of the D-T reaction is in the form of fast neutrons? I think capturing those and turning them to heat on a GW scale is fairly hard.
A lot of the steps look doable on paper or in a lab, but putting it all together in a working power plant is something we just don't know until we've tried.

By comparison, nuclear fission is so easy it only took a few years to develop a power plant. And despite being easy, fission plants are expensive and complex. A fusion plant is only going to be more expensive and complex.

The ITER reactor will not generate electricity, they will dump the heat to cooling towers. The one after that, DEMO will have generation capability. Then the one after that, PROTO, will be a template power generation plant. So sometime near 2070, maybe...

I went to visit JET at Culham in the 1980s, and they of course repeated the adage "fusion is always 30 years away". That was 40 years ago. Now it's 50 years away. It seems the longer they spend on it, the further away it becomes. 

[RoGeorge]

@21:55 Wow, just one shot and the vacuum chamber looks kind of ruined 

[moffy]

yeah, like a bomb went off
I guess that is some of the waste energy they have yet to deal with.

I must say that I have found the comments and discussion illuminating.

[excitedbox]

I like the focus fusion method. They don´t need to boil water or anything to spin a turbine either. They create an alternating current directly. In a sort of foil ball. It is also by far the cheapest method. I think the main researcher comes off as kind of a nut from his appearance and that could be holding them back from getting real funding. I wish there was less politics etc. in research funding.

[moffy]

I have to agree, I hope they get funding also. They have pioneered much of the theory, it is a shame that they can't progress to the next level.

[pwlps]

Most of the energy output of the D-T reaction is in the form of fast neutrons? I think capturing those and turning them to heat on a GW scale is fairly hard.

This is why the Z machine approach might be interesting, it can reach high enough energies for aneutronic processes like the B-p fusion.

[ejeffrey]

Getting the energy out isn't really that hard.  It's obviously not entirely trivial but if you can get the fusion part working with enough heat output building a power plant around it is not going to be the obstacle.

Most of the energy output of the D-T reaction is in the form of fast neutrons? I think capturing those and turning them to heat on a GW scale is fairly hard.
A lot of the steps look doable on paper or in a lab, but putting it all together in a working power plant is something we just don't know until we've tried.

Depends on what you mean.  Slowing down neutrons and capturing the heat isn't really that hard.  After all, we do that in fission too.  The fusion neutrons are higher energy (about 7x higher KE), but slowing them down still isn't a problem.

The issues with DT fusion in particular are that the neutrons tend to destroy the reactor and create radio isotopes in the process.  While DT fusion doesn't have any risk of meltdown like a fission reactor, it still creates a large quantity of radioactive waste.  The other problem is that a DT fusion reactor needs to produce tritium to complete its fuel cycle.  Some of the fast neutrons have to be captured to produce tritium -- and it is a lot.  You need a net tritium yield of 100% -- every fusion event produces 1 neutron which must produce on average 1 tritium nucleus to replace the one used.  Lithium 7 can produce multiple tritium from a single high energy neutron, but you still must capture a significant fraction of the the neutrons in your lithium blanket, and that has to fit along with your magnets, lasers, power supplies, or whatever else you need to confine and generate the fusion.

P-B fusion gets rid of most of these disadvantages--it produces minimal neutrons, its inputs are naturally occurring, and poses no proliferation risk.  However, the ignition temperature is 10x higher and the electron number (and therefore bremsstrahlung losses) are much higher as well.

[CatalinaWOW]

Pretty much all pulsed fusion reactors like z-pinch are at best useful for research and understanding but most of them are actually built for weapons research. They all have some sort of "we will revolutionize the world" story to get funds but then later on after doing a few shots are modified for mil applications, laser fused and scalable fusion bombs. Continuous is the only way that actual useful power production may be done.

If you are just after fusion, farnsworth fusor or beam on target is the way to go, you will have a nice neutron source but no energy production.

I am not sure of the basis for your assertion.  Discontinuous energy bursts are widely used for power production (although rarely, if ever, at Gigawatt levels).  They are called internal combustion engines.

Fundamentally all that is needed is for the pulses to repeat on a time scale that is short with respect to the period of the next item in the conversion chain.  If you are boiling water for steam I would assume that that means a large fraction of a minute.

[NivagSwerdna]

I went to visit JET at Culham in the 1980s, and they of course repeated the adage "fusion is always 30 years away". That was 40 years ago. Now it's 50 years away. It seems the longer they spend on it, the further away it becomes. 

I remember a talk when I was at school in the sixth form about JET... sounded so convincing... must have been circa 1982... as you say... it's 2020 now... obviously a tricky problem.

[jmelson]

Getting the energy out isn't really that hard.  It's obviously not entirely trivial but if you can get the fusion part working with enough heat output building a power plant around it is not going to be the obstacle.

Yes, actually getting the energy out is THE PROBLEM.  Unless the plasma is insanely compressed, it will radiate all the heat away in microseconds.  That's the issue, you have to prevent all the heat leaking away from the plasma.  So, you have this plasma at tens of million K, and the heat flows away from it, and the walls of the machine have some system to remove the heat.  That can be used to run a steam turbine.

And, I think that's the whole debacle of Tokamak-style fusion, drawing the plasma out into a many meters long torus, no matter how thin, leads to huge surface area from which the heat is lost.
That's why inertial confinement has an edge, the plasma is compressed into a tiny volume, therefore reducing the heat loss.

Jon

[Amper]

Because internal COMBUSTION engines are not called internal detonation engines. Do you have a feeling for milli micro and nanoseconds and how matter starts becoming liquid at such pressures and times? To a piston the 20ms it takes for every revolution is a loooooong time. Also consider that Its at temperatures that solid matter can handle, not insane ones like in fusion that will blast layer by layer away from a surface in every shot.