The spherical approach seemed like a great idea until they actually built them. Now it's pretty clear the economics are no better than the conventional MFP approaches. See the Disadvantages of this article, especially the first two items listed:

https://en.wikipedia.org/wiki/Spherical_tokamak

> The russians have been operating sodium / lead fast reactors for decades

Sure, especially on their submarines. Excellent safety record there.

> The next step BN-1200

Riiiight. Like the Gen III reactors were going to be so much better than Gen II. Look how that turned out.

> That $100 car battery? A lithium-ion equivalent that's 1/10th the weight for the same
> capacity and probably even more cold cranking amps might be $80.

Sheesh. Why not also demand it be made out of unicorn tails and magic dust?

Li-ion is 1/3rd the weight. 1/3rd, not 1/10th. It doesn't have to be any lighter.

Li-ion also has less *power*. Be sure you understand the difference between *power* and *energy*. A li-ion battery will have *less* cranking amps, not more.

> I know it's not XKCD, but there's relevant SMBC [smbc-comics.com] and PHD [phdcomics.com] comics.

Minor complaint with the second: we know from studies that the problem is not the university PR departments, but the researchers themselves.

> Why do physicists insist on treating gravity as a force?

Because everything else works that way.

> Since Einstein, we know gravity is the curvature of space-time

No, since Einstein we know that Einstein's model is that gravity is the curvature of space-time.

Before Einstein, we thought it was a force between objects, or objects and a space-filling field.

There's no reason to suggest one model is inherently "more correct" than the other. Personally, I *like* the geometric model more, which almost certainly means it's wrong.

The good news is that such an experiment is likely far, far less expensive than the LHC. Therefore it is also more likely to happen.

Someone needs to write a paper on the inverse relationship between CAPEX and chance that the experiment is carried out. Of course, that relationship is likely identical to many, many others.

> Except those who continue working on it.

Maybe. But there's a long history of people working on projects they know are going nowhere while keeping up a brave face. I'm sure you've worked on a few yourself.

> You may call this hand waving, but the best way to establish this is an actual experiment.

Absolutely! Which is why I brought up the Teller example. The pattern is *exactly* analogous.

> This article from 2 years ago [slashdot.org] and its accompanying chart [imgur.com] make
> a good case that we'd have fusion already if we as a civilization seriously funded it.

It doesn't make a difference how fast you throw money at it. If the parts that go into the reactor cost more than the economic value of the electricity that comes out, then no one is going to ever build one. And right now, that's absolutely the case.

It's a little hard to do the math on something like ITER, which runs continually. On the other hand, it's really easy to do the math for something like NIF, where the inputs are nicely quantized. NIF burns a fuel packet that costs thousands of dollars. Under the most ridiculous future scenarios, they thing they can get that down to 50 dollars. Mind you we're not talking about the machine here, just the fuel.

When burned perfectly, which of course we can't actually do, we expect to get about 13 MJ of fusion. We might extract 25% of that energy as electricity. That sells for about 5 cents.

So $50 in, $0.05 out. And that's the best case scenario.

> Actually, it was shut down because Canada had a surplus of heavy water

Which says a lot about the industry as it currently stands. In spite of numerous technical advantages, actually selling a D2O reactor seemed beyond the capabilities of the country.

> Canada developed a new technology for enriching deuterium from water, based on catalyzed exchange

Currently small-scale system suitable for lab production and make-up supplies for the existing reactors, based on semi-enriched feedstock. That feedstock comes from LPCE.

> I thought it was kind of a general purpose device that could do other things.

Oh sure, but no one was *banking* on it. The basic long and short is that if you detect the Higgs everyone gets a Nobel and a slap on the back. If you don't detect anything, you get nothing.

So the entire project is focused on Higgs, because we already know it exists, as opposed to actually useful science like supersymetric partners which no one really has any clue if they exist or how to really look for them.

It is entirely possible LHC will return really useful new science. It is equally likely it will not. In comparison, I *guarantee* you that the EELT will generate new science, science that the standard model can't explain. So in terms of cash-for-outcomes, it's no contest.

> The old FermiLab accelerator could do all manner of experiments

Indeed, and it spent the last decade of its life spending hundreds of millions measuring the top quark mass to the 6th decimal. One can imagine less useful ways to spend money, but you'll have a hard time doing so.

> With regards to the Polywell design

No one, and I mean no one, expects the Polywell will escape the Ritter issues.

The team has done some fantastic hand-waving to pretend it doesn't effect them. Until *proven* otherwise, it does. That's the way science works.

And if you think this is unfair, you *really* need to go read the history of fusion research. For instance, back in '54 Teller gave a talk at the Gun Club where he outlined a problem and noted that any reactor design that used fields shaped like *this* probably wouldn't work. Many of the people in the audience (or all of them) had reactors with fields shaped like *that*. They all went away and came up with excuses as to why Teller's concern didn't effect them. Guess what, it did, and they all failed.

So I'll believe the polywell has escaped Ritter's bremsstrahlung concerns when I see the polywell escape Ritter's bremsstrahlung concerns.

> As to General Fusion

Mechanical MTF. Good luck with that.

> The watts per square meter are still very low, the panels very expensive, the land and installation requirements still onerous

All-in, including land, clearing it, levelling it, installing equipment, trenching lines, all CAPEX and REG, every single penny from one end to the other, costs $1.79 a Watt.

In comparison, fission plants are currently going in for at least $5 a Watt, but have overrun their budgets almost every time.

Fusion reactors would be fantastically more complex and expensive than fission. To put that in perspective, the start-up load of lithium-6 will cost about $1.80 a watt. The concrete in the floor will be another 15 cents. So just for the floor and one ingredient, you're already more expensive than a complete spinning-the-meter PV system.

> Face it, the only people buying solar
... is everyone on the planet. PV is the second fastest growing power source in history. Wind is the fastest. Numbers:

http://cleantechnica.com/2014/03/18/37-gw-solar-capacity-installed-worldwide-2013/
http://www.mercomcapital.com/global-solar-installations-to-reach-approximately-43-gw-in-2014
http://www.epia.org/fileadmin/user_upload/Publications/GMO_2013_-_Final_PDF.pdf

As a result of this activity, PV alone has gone from nowhere to a real bump on the graphs:

http://www.renewableenergyworld.com/rea/news/article/2013/02/100-gw-of-solar-pv-now-installed-in-the-world-today

100 GW of PV compared to about 370 GW of fission, before many of them were turned off. It took about 40 years to get to that point with fission, so PV is on track to surpass it quite rapidly.

> about $2 billion per year total.

Which, at current prices, gets you about 2 GWp worth of wind or 1 GWp of PV.

So if we continue for another 30 years and a miracle occurs and we get a working design, we would have already installed 120 GWp of wind turbines for that cash.

Wind turbines actually work, which is why they're the fastest growing source of electricity ever. Fusion almost certainly will never work, economically at least, and certainly won't be available for decades. And even if it does, by the time we have them, wind turbines will be even cheaper, grid distribution will be a solved problem, and we'll probably have a day's worth of storage in all our homes.

So how much should we be dumping into this latest idea to centralize the grid? Explain your answer without resorting to miracles or handwavium.

> The target of ITER is not break even but Q=10

That's still break-even. And that's not what he said anyway, he said "achieves break-even (but no power tapping)", in the context that "DEMO is built and demonstrates power to grid".

Do you deny that ITER is not tapping power? Let's see...

> DEMO is then supposed to actually convert this excess heat into electricity

It appears you agree, and are simply quibbling over one possible interpretation of the statement "achieves break-even". Note that any Q >1 is "achieves break-even", so the original statement is perfectly correct.

> About the ELM's: of course the target of ITER is to overcome that issue and make the process reliable

As was the purpose of every design before it; to overcome [insert problem here] and make the process reliable. Given the 65-year string of failures in this regard would it be terribly surprising if ITER didn't manage to fix ELMs?

> But what about, say, HiPER?

I wrote the wiki article on HiPER (check the history if you don't believe me). The lead researcher has moved to LLNL, and the fast ignition method turned out to be a dead end. HiPER still exists on paper as what would best be described as a laser development effort, but for all intents it's dead. The entire fast ignition field has moved on to another holy grail, although there's continuing effort in Japan as their experiments were furthest along.

Simply put, laser-based ICF cannot ever work economically. We have suspected this since the 1960s. There was a brief period during the early 1970s when it appeared the driver energies were low enough and isotopic smoothness was not all that critical, so we might be able to build one. By the 1980s it was clear both of these were not true, and that you needed extremely powerful highly smoothed laser systems, along with extremely expensive highly machined holoraums.

What that means is that even if you get the energy output to be higher than the input, and we're several orders of mag away, the amount of *money* you burn is higher than what you get back out. Every time. And we know enough about the instabilities of the implosion process to say that that's just the way it has to be:

http://matter2energy.wordpress.com/2013/04/21/fusion-the-power-of-wishful-thinking/