The Current State Of US Fusion
So bottom line: we are spending 440 million on a spectrum of groups, companies and universities following an approach that we know will never work. It will not work energy-wise nor will it work commercially. That may have been fine when there were no other fusion approaches available to us – but that is no longer true.
- The Levitating Dipole Experiment at MIT from 1998 – 2011. This was run by Dr. Jay Kesner and Dr. Micheal Manual . The team was gearing up for ignition experiments when funding was cut - for ITER - in 2011.
- The Dynomak at the University Of Washington, run by Dr. Tom Jarboe. In 2011, the group found a great way to heat plasma. This opened up a whole new approach to spheromaks . They could not get the funding and so went private.
- The POPS machine proposed by Dr. Park and a team at Los Alamos in 1998 . This concept played games with ion oscillations to avoid the fusor grid.
- The penning trap concept pushed by Dr. Dan Barnes (LANL) late nineties. Dr. Barnes wanted to trap a mostly negative plasma inside a penning trap and get the ions to accelerate using the created voltage drop .
- A multiple ion beam approach pushed by Dr. Ray Sedwick at the University of Maryland in 2011. This approach has some fundamental charge limitations and instabilities – but Ray was able to get pretty far with it .
- The Plasma Liner Experiment pushed by Dr. Hsu at Los Alamos and by Dr. Doug Witherspoon at HyperV Inc. This concept used converging plasma beams to squash plasma together. This has been around since the early 2000’s.
- The Field Reversed Configuration work done at the Redmond Plasma Physics Lab in Seattle. This group existed through the nineties and was run by Dr. Alan Hoffman & Dr. John Slough . When they could not get any more DOE funding, they formed two private companies.
- The Princeton Field Reversed Configuration. Dr. Sam Cohen at PPPL set a world record for the longest stable FRC ever created by mankind  but he runs his lab on less than 200k a year. This year, the DOE ended their funding (to support ITER) and they had to go through NASA and the Army.
If we ever want to get to a commercial plant - we need a new funding system. The US should adopt a system similar to the one used in pharmaceutical companies. Drug companies have a pipeline. They give a bunch of high risk drugs a very small amount of money. These could fail. So they support many concepts – but keep things cheap. A good amount might be a million a concept. That is typically more than most teams survive on. A second tier should be supported at the 10 to 30 million dollar range. You could support maybe ten concepts there. These are more mature designs. Finally, there should be three big efforts that should be funded at the 100 to 150 million range. This way you could incorporate your existing efforts, while adding in new ideas. For example, ITER and NIF could get 100 to 150 million per year in this system. But you could add a new big machine, like a national FRC.
Critically, you must build this system around going commercial. The money must be tied to this. Right now, all we care about in fusion is something called a triple product. That is the density x temperature x trapping time of a machine. This is a terrible meter stick. By that definition NIF is a great machine – even though it costs 3.8 billion and failed to get ignition.
We should care about a machines: run time, cost, efficiency, energy in/energy out and size. Groups must to prove they are making progress along these lines - to move up or down in funding. Below is a quick graphic I threw together on this. Finally, it would be nice to try to fund with public-private partnerships. Use public money to try to lure private investors. There are some interesting co-ownership models you could pursue. For example you could have a privately run group applying for public funds through block grants.
- Fusion rate is self-explanatory. So far, much of the world has focused solely on this, by driving at triple product.
- Conduction. This is all the energy that leaves with the mass. Metal is the enemy here. When plasma touches metal - it leaks out. This robs energy from the machine, killing it’s’ efficiency. Hence you need space around your plasma. You also want a strong trap to keep the plasma away from the walls. Lockheed Martin is following a long shot approach to make a near perfect magnetic trap – which if it works – would be a big break through [29 -33]. Foolishly, Lockheed has not published so we do not know where they stand. I have even seen people build their chamber out of insulators like glass to stop conduction losses .
- Radiation. This is all the energy leaving the plasma as light. Plasma bleeds energy away, as light. People have talked about reflecting that light back into the plasma – but that idea is very, very limited. For example is almost impossible to reflect an X-ray. Another option is to drive up your density. That could slow light loss. Personally, I think your best bet here is trying for a tuned plasma. In a perfect world you would want a plasma with lots of really cold electrons and a few hot ions. That would be best. This kind of distribution may not be possible . It is not clear at this point. Energy distributions in plasma are a function of many other things: shape, structure, magnetic and electric fields, injection, runtime, ect…. Could you tune your plasma? IDK. We need to get hard data proving this one way or the other.
- Efficiency. This is how well your device spends or collects energy. There has been some exciting work done here. In 1982, a team a Livermore was able to capture 48% of the energy coming off a fusion reactor using something called direct conversion. Basically direct conversion works by putting the charge particles coming off the reactor directly into a wire.
J. Proc. of 20th International Stellarator-Heliotron Workshop (ISHW), Max Planck Institute, Greifswald,Germany. Greifswald, 2015.