A compact fusion reactor?

Here is Guy Norris:

Hidden away in the secret depths of the Skunk Works, a Lockheed Martin research team has been working quietly on a nuclear energy concept they believe has the potential to meet, if not eventually decrease, the world’s insatiable demand for power.

Dubbed the compact fusion reactor (CFR), the device is conceptually safer, cleaner and more powerful than much larger, current nuclear systems that rely on fission, the process of splitting atoms to release energy. Crucially, by being “compact,” Lockheed believes its scalable concept will also be small and practical enough for applications ranging from interplanetary spacecraft and commercial ships to city power stations. It may even revive the concept of large, nuclear-powered aircraft that virtually never require refueling—ideas of which were largely abandoned more than 50 years ago because of the dangers and complexities involved with nuclear fission reactors.

Here is another account.  It is suggested that the reactor can fit on the back of a truck.

In response to such speculative reports I usually say: “If it’s true, why isn’t the price of oil down?”  But these days the price of oil is down!  I am not suggesting this is the reason, but at least I can no longer say “If it’s true, why isn’t the price of oil down?”.  I have to say something else, so if this is true — which I cannot judge — there is no great stagnation (any more).

For the pointer I thank various MR readers.

Comments

I really like the picture of it mounted on a DeLorean.

I remember reading a Sunday Comic about how fusion energy would be arriving real soon now in about 1967.

It's always forty years away - that's one of the great invariants of physics.

"The energy source of the future, and it always will be."

Perhaps. But people used to describe replacements for CRT displays as "the technology of the future" also. And so they were, until they weren't.

The point is that fusion has been quoted as that for much longer with seemingly little progress (I know this is not true, but progress that the public would care about).

To be fair, they're only saying they might get there in a decade or so. So the price of oil, or the markets, are unlikely to respond given this kind of uncertainty. Plus, people expect PR releases.

Maybe LM executives have some stock options coming due soon.

In 2013, LM was promising a reactor in 4 years - http://www.dvice.com/2013-2-22/lockheeds-skunk-works-promises-fusion-power-four-years

Well, it doesn't sound crazy, but they haven't built anything yet, so who knows.

Also, in what world does "the potential to meet, if not eventually decrease, the world’s insatiable demand for power." make any sense?

For fusion to be competitive its capital cost is going to have to be a lot less than conventional nuclear power. The fact that fusion's fuel is cheaper only makes a small difference as the cost of fuel is only a small portion of the total cost of fission. As I doubt its capital cost will be much lower than conventional nuclear power I doubt fusion will have much effect on energy prices. But maybe they'll be great for missions to Pluto.

The other factor is the (non) cost of disposing of the waste. Of course, this is heavily subsidised, so we can expect some vicious push-back from the existing industry if this turns out to have legs (which would be a surprise)

Push back from the existing nuclear power industry? Despite what they might show in zombie movies, I don't think a corpse can push that hard.

The real question is why nuclear plants cost so much to build. Is it driven by the safety and regulatory precautions, or the sheer technical challenge of constructing a working fission power plant in the first place?

In other words, if I wanted to throw caution to the wind and construct a fission plant that had a dangerously high chance of melting down each year, how much could be saved in the initial construction?

I'd be inclined to think that the acceptable risks in fusion would be higher, and therefore perhaps the capital costs need not be so high. Perhaps the costs could come down further as the technology improved.

In any case, fusion could still be a game-changer (if not a TGS-ender) at any cost between nuclear and solar.

Yes, you could build a nuclear reactor cheaply and it would still be extremely safe. The costs are mainly paperwork proving a materials "Pedigree", you have to have an audit-able paper trail from the time a metal is mined, through the manufacturing process and into final installation. There are very few accidents at non nuclear plants even without these controls. Just to give you an example, I needed a connector for two electrical cables in a battery charger. The part cost $30, to get the identical part with the paperwork certifying it was "safety grade" was an additional $10,000.

A different group has a pro forma projection that their fusion technology could rival the cost of coal fired power plants.
http://motls.blogspot.com/2014/10/fusion-dynomak-new-compact-rival-of.html?m=1

Another group apart from Lockhead is touting a second approach (not priced out economically).
http://news.sciencemag.org/physics/2014/10/z-machine-makes-progress-toward-nuclear-fusion

Lockheed is touting a short development time horizon and a very small size.
http://motls.blogspot.com/2014/10/lockheed-martin-promises-fusion-plants.html#more

All of this, however, is pure vaporware until somebody manages to build a fusion reactor that can break even in terms of output to input energy. Nobody has ever done this and the only examples of it in nature involve stars or very large gas giants on the brink of becoming stars that don't quite make it. Until Lockheed can demonstrate a reactor that can break even, its ten year time horizon is coming entirely from the marketing department and not the engineers.

Also, while tiny fusion reactors would definitely be cool, improvements in battery technology could be almost as revolutionary. If you can invent as high energy density, low economic cost battery (or battery equivalent) - and the improvements relative to the status quo don't have to be all that huge - a cost reduction of a factor of 65-75% per unit of energy stored with only very modest improvements in energy density would be enough to shift the U.S. motor vehicle fleet from predominantly petroleum based to predominantly electric, then it really doesn't matter how big your power plant is.

The improvement in photovoltaic cell cost per watt of power produced needs to be on the same order of magnitude to make it more than a niche product. If you want to do something like turning an intermittent power source like wind or solar, into a base 24/7 power source like coal or nuclear power, you don't even have to care about energy density, all you need is a battery that isn't too leaky (i.e. that doesn't lose too much energy while it is stored) and is really, really cheap.

Existing nuclear fission reactors can be costly, but this has a lot to do with artificial regulatory costs, a lack of any permanent approved high level nuclear waste storage site, and a lack of approved mass producible designs. High level nuclear waste is nasty, but in the volume of waste produced by coal fired plants is many orders of magnitude greater than a fission reactor and is a lot more noxious than people give it credit for being. Coal may not be as scary, but actually kills far more people per billion kilowatt hours of energy generated. Also, the price of coal is as much a function of the cost of the next cheapest alternative as it is driven by the cost of actually mining coal and producing electricity with it (which externalizes a lot of the costs involved in accidental injuries and pollution). The supply of coal is so great relative to demand (which is low because it is so polluting and pretty much is now confined to electricity generation), that there is a lot of economic room to reduce the spread between the input based cost of making and using it and the current price. Proliferation concerns and political fear are at least as much drivers of cost of nuclear power as genuine technological challenges.

The smallest old school fission reactors used in practical applications are those used in nuclear submarines which at ca. 8000 tons displacement, are not particularly big as ocean going maritime vehicles go. Commercial freighters and oil tankers and cruise ships are all much larger. The latest model nuclear fission reactors are small enough to generate only enough power, for example, for a single village in Alaska of a few thousand people, perhaps just ten megawatts or so, and the fuel replacement cycle is measured in decades.

> turning an intermittent power source like wind or solar, into a base 24/7 power source like coal or nuclear power, you don’t even have to care about energy density, all you need is a battery that isn’t too leaky (i.e. that doesn’t lose too much energy while it is stored) and is really, really cheap.

If you have the right geography, http://en.wikipedia.org/wiki/Pumped-storage_hydroelectricity

I agree that sentence was weird. Perhaps they mean that there is some probability it will blow up and kill us all. That would reduce demand for power.

Well, if it significantly reduces the need for oil/coal/natural gas, we will need less energy to drill/refine/pump/transport.

It makes sense if you put "conventional" in there. But there's no way the insatiable demand for power stops. If that can fit inside a truck, it can also fit in the back of a truck that can transform into a robot and shoot lasers out of its eyes.

Like the tacocopter, I assert the only reason we don't have trucks that turn into eye-laser-shooting-robots is because of over regulation. I hereby demand the next President and Congress make it a point to abolish all robot inhibiting regulations ASAP!

Ah, so TGS is simply a result of insufficient cheap energy?

In all fairness I think Tyler was conceding that such an innovation would be a world changing innovation that will have ramifications that are positive for the lowest on the economy ladder for generations to come. Like harnessing electricity did. We will now have much more low hanging fruit.

This would be a method for energy production, not for moving energy from one point to another. Anyway, inasmuch as more energy correlates with lower hanging fruit... sounds like a "yes."

"Ah, so The Great Stagnation is simply a result of insufficient cheap energy?"

Pretty much.

Dude, where's my flying car?

Being suppressed by Big Rubber.

How about:
Being covered up by Big Rubber
Contracepted* by Big Rubber
?
* - props to the flexibility of the english language

sorry, forgot that all blog inputs aren't wysiwyg

Being covered up by Big Rubber

Contracepted* by Big Rubber ?

* – props to the flexibility of the english language

In other questions, WHAT is driving oil prices so much the last couple weeks?

Cynically, a quid pro quo to the middle eastern states for the US to go after ISIS -- drive energy prices down to add further pain to Russia.

Same thing moving dollar up, stocks down and treasuries up. Expected fed tightening --> deflation and recession

Higher US light oil production and a response from Saudi Arabia to increase production to preserve their market share, something of a finger in the eye of the rest of OPEC because other OPEC members are unwilling to cut their output to maintain the price level.

Nothing to do with the fed. It's the same oil supply constraints moving around as it usually is. Durr.

It's also because demand is expected to decrease because of economic slowdown in China and Europe. Less demand combined with more supply.

The Sci AM article didn't seem to talk about the process so I don't know if this is similar or not but I saw a show sometime within the last year (Discovery, Sci Channel wherever) where some people were working on laser based fusion. The device was in a lab room but clearly could have been smaller than the prototype they were working with.

All that is needed is well focused lasers, a few hydrogen atoms -- to make helium -- and a few terra watts of power (IIRC) to fuse the hydrogen atoms. Once started the energy produced is supposed to be sufficient to then power the lasers and still have enough energy remaining to power a large city.

High startup cost, low marginal unit cost?

As for the price of oil I don't see why it necessarily needs to decline -- that's the standard 101 view but there is some monopoly and cartel aspects not to mention the regulatory side. If the industry has 10 more years oil prices might go up.

The "Guy Norris" link is much more informative and they have a video.

The laser based method has always seemed like it was just a few mechanical hurdles away, but those hurdles have been much higher than anyone had hoped. This is based off of the toroid/tokamak design which relies on magnetic fields and a semi doughnut shaped plasma.

There are three or four main approaches. The laser one, the tomak, and one or two that split the difference.

I don't see what's special about this approach. It's still magnetic compression of a hot gas. There were pinch devices similar to this one over 60 years ago. What makes this one different?

Probably the superconducting coils are more feasible.

They had superconducting magnets 50 years ago, and they don't conduct any better today than back then. You might be able to cool them with liquid nitrogen today, rather than liquid helium, but that's mostly an operating expense issue.

I think they use liquid He, at least that's what they do at the LHC and presumably ITER. But the fact is big superconducting magnets are standard equipment in a way they weren't a few decades ago. I don't really know what has changed, the electronics for energising them perhaps?

On second thought, maybe it's that Helium fridges are better nowadays.

Historically these have been built, but require more power to run than they produce. The efficiency has consistently increased to the point where even when I studied this stuff 10 years ago some people were claiming that if they built one from scratch they could get to a level where the net power production was positive. Unfortunately, you can't build one of these things from scratch every time you discover a small improvement. I am not sure, but my guess is that this is the sum of a lot of minor improvements that finally add up to something that Lockheed believes is a viable product, or at least a worthwhile project. Please note that they are going public now because they are looking for investment partners, as this will cost well into the billions of dollars.

Yep, in fact it's a little assuring that this thing is not actually revolutionary. The explanation that makes me believe that ITER is worth a few tens of gigabucks is "well we know how tokamaks work, they have scaling laws that say that if we build a big huge ugly cow of one one, then it will work."

Presumably this research effort is more less ploddingly sure than the tokamak approach, but on the other hand requires a smaller device and puts less money and effort at risk. So I'm glad their trying it.

I thought the main problem has always been plasma instabilities, in which the confinement of the plasma breaks down -- not because you don't have enough power to maintain the magnetic field, but because no amount of power is sufficient to prevent anomalies from developing and growing until they screw it all up. Are you telling me that plasma instability is a solved problem?

http://en.wikipedia.org/wiki/Plasma_stability

+1 I'd like to know how this reactor overcomes the instability.

Imagine that you had to balance a stick with a hinge in the middle on your finger. That's what plasma instability is like, and it has been an issue with the first fusion reactors from the 1950s until today.

The main instabilities one worries about in a Tokamak come in two flavors. One is current driven. Since the magnetic field used to confine a plasma in a Tokamak is partially created by running an electrical current through the plasma itself, this means that a big Tokamak has huge electrical currents flowing in it (full power ITER shots will be 15 million amps). That's a source of free energy and many of the worst instabilities tap into this source of energy. These tend to be big, catastrophic instabilities - when you trigger them, the entire configuration goes so unstable that the plasma smashes into the wall and all hell breaks loose. So they must be avoided, and this places some limits on how much current you can run. (The main ones are called kink modes and tearing modes, externa

The other flavor of instabilites are those that are driven simply by the big pressure gradients present (or density/temperature gradients specifically). Some of these are big, catastrophic instabilities that would throw the entire configuration into the wall, though this is pretty uncommon. Mostly, these instabilities stir up turbulence which causes heat and particles to leak out of the device - this is the main mechanism limiting the gradients in temperature or density you can achieve. Almost all of the instabilities we run into that drive this turbulence are very sensitive to the structure of the confining magnetic field.

Without digging too much into the details, we would crudely expect all of these instabilities to not be a big deal for a mirror configuration such as the one Lockheed martin is proposing. However, there are tons of other problems that are likely far worse (see my comment at the bottom here). Also, there are other instabilities that pop up when the plasma pressure approaches values comparable to the confining magnetic pressure (in a Tokamak, one might get to 5% - they're claiming something like 100%). Assuming they can solve all the problems associated with end-losses in a mirror machine (extremely unlikely but I can't rule it out given that they haven't released many details), then we would get to see what really happens when they heat that thing up to beta values of 100%. There are instabilities that we know about from astrophysical applications called the mirror and firehose instabilities (and even many others) which arise when the pressure is that high - these could be a showstopper (assuming they get past all the other issues).

Really, this is an old idea that has failed many times in the past. Perhaps they have solved some of the problems, but they haven't yet released enough detail to take it credibly.

All my lifetime, which is not inconsiderable in length, people have been promising cheap, clean fusion power in ten years time. It is the Brazil of the nuclear power world - it is going to be huge in the future. And always will be.

The only reason to take this seriously is that it comes from the Skunkworks. They do have a track record of actually doing stuff. Largely because they have a lot of money and very little red tape.

So we will have to see. I would not be selling my shares in the nuclear reactor companies just yet.

Precisely. The research into fusion is well-intentioned, but we get articles like this about every 2-3 years: some new breakthrough in fusion which will generate positive energy gains after about ten more years of research. These announcements have the same value as that email from the Nigerian prince wanting to share a million dollars with you. I'm not saying that the research is bogus, I'm saying that designing a workable fusion reactor is exceedingly difficult, and if you ignore every fusion article that gets printed, you'll be not much worse off than ignoring all of those Nigerian emails.

The observation that this comes from the Skunkworks is good to note, but we've seen similar fusion announcements from MIT and other notable places. These are smart and hard-working researchers, but I'm not holding my breath.
http://www.wbur.org/2014/02/06/mit-fusion-center-federal-funding

It is also worth noting that all of the fundamental laws of physics pertinent to nuclear fusion reactors are already known to high precision and have been part of the Standard Model of particle physics for a generation. This is all engineering and computation, not scientific discovery.

I'm interested to know how big Lockheed's investment is in this. It's one thing for governments to try out a multidecade moon shot on fusion, but it indicates something more immediate and plausible if a major, mainstream for-profit company is putting significant money on the line for fusion.

Fusion is the power source of the future and it always will be! *rimshot*
Sorry, someone had to post it.

Tyler,

In this case you could ask, "if this is true, why isn't lockeed stock up?"

It is - outperformed S&P since start of 2013 & from start of 2014 & over recent months.

The market cap of Lockheed is fifty billion dollars. Working fusion by itself would be worth more than 20 times that.

And the big deal is supposed to be compactness? Let's see controlled fusion first, even non-compact, even full-size.

Sure, it's a talking dog, but he never has anything interesting to say.

It's good that they're saying it's "only" ten years away versus the typical "fifty" or even indeterminate, but I'm a little skeptical. I hope they get their funding to do it, but we'll have to wait and see if they can get net positive power output at a commercial breakeven point with it.

It really would be useful, though. Having something other than coal and natural gas to run the baseline power grid would be nice, especially if the other growing power supplies are the more erratic solar and wind.

Fission plants are used for base load because you can't throttle a nuclear reactor core very quickly. In France, they run their reactors at about 40% of full power in the middle of the night, and they provide cheap electricity to their neighbors at night, because they need to keep the cores hot to meet daytime demand.

A fusion power plant wouldn't be anything like that. You would be able to throttle it up and down at will. You could throttle it much faster than a coal plant, but not quite as fast as a combined-cycle natural gas plant.

@MT - you still working on your fusion invention you mentioned a while ago or was that a joke?

As for this story, it's significant. The fact is: had the US government offered a huge prize to the first team to invent feasible fusion --say the cost of the Gulf wars--with smaller prizes for meeting milestones along the way to this goal (so the teams don't lose hope), we'd have fusion now. I am familiar with the invention industry and though most if not all inventors nowadays invent for almost free (i.e., for their base salary, which BTW is a shame and there should be a federal labor law against that, with prizes given to worthy inventors by the Fed government) their employers do not. Offering a prize, and/or a bigger, better patent reward, does accelerate innovation.

As for naysayers who say fusion news is often positive but nothing big ever seems to happen, well the same thing is said for cancer, for MS, for other diseases including virus-vector diseases, and in fact over the last 40 years the frontier has expanded on combating these diseases. Would you prefer 1970s technology if you were ill, or today's? Progress is being made, but if you want to raise the bar you have to spend the money. That simple. Instead of worrying about patent trolls, which are like mosquitoes to a large mammal--no big deal--society should be worrying how to increase innovation. Off soapbox now...

Not a joke. I can't make any progress right now because so much of my time is taken up every day taking care of my mother, who has Alzheimer's DIsease. I have the equipment and raw materials, but not the time. It'll either work or it won't. I don't know when I'll be able to run the experiment.

The important bit is it is not magnetic compression of a hot gas. Neither was the first man-made fusion, over 80 years ago. That experiment was what is today called beam-target fusion, but it was not a practical source of energy because among other things it took place at near-vacuum pressure. I believe any practical form must take place at atmospheric pressure and ambient temperature, at least for the bulk materials. (In my case there would be islands of reaction sites which could be interpreted as having extremely high temperature and pressure, but those sites would be at the nanoscale. At the macroscopic scale, the bulk material would seem like some highly energetic fuel which only required a certain exciting energy input to yield a much larger energy output. It would not be self-exciting, so no danger of a runaway reaction.)

I spent 20 years operating US Navy submarine reactors. We could go from self sustaining (drifting) to full power in under a handful of minutes. We had the identical reactors at the training sites and with trainees at the controls we would start at full power, shutdown, restart and go back to full power in a 4 hour shift.

The submarines aren't as efficient as the large commercial ones (we don't recover nearly as much steam plant heat). Commercial plants are designed to change power slowly (or not at all) but to operate very efficiently at full power.

"In this case you could ask, “if this is true, why isn’t lockeed stock up?”"

I don't think the market cares much about what happens 10 years from now.

Tell that to people who buy zero-coupon bonds, or just about any bonds for that matter.

I have some zero coupon bonds that will mature in twenty years. I have had them for ten years already. There is a market for long term investments.

Bonds have value outside of their coupon and maturity. You could just up and sell them at any time you wish.

They do, its just discounted. What's the discount rate on something like fusion? Call it 30%, which is probably low, and that $1 ten years from now is worth $0.02 today. If it is 50%, also probably low, that dollar in ten years is worth $0.00098 today.

There's a new DOE fusion program:
https://arpa-e-foa.energy.gov/#FoaId4c902711-2bf4-409d-bd46-8d9c57ed83d1

Funding in the small millions range for three years. You will be hearing more about small fusion reactors.

The "five-to-ten years" timeframe makes me highly skeptical about how feasable this technology really is.

Would this be worthy of being voted on at SciCast?

Going through the info at Lockheed Martin I got the impression that they kept on talking about magnetic containment only, nothing on how the plasma is generated or how fusion is induced, seemed like they are talking about better shovels, not about the gold nuggets.

Sigh, my bet is still on the careful development and scale-up process being followed by ITER and not some whiz-bang stuff. If the current development trajectory holds, after ITER we'll build a demonstration-scale fusion plant explicitly for electricity production, gain operational experience, and THEN we'll get commercial designs deployed. Earliest that's happening is 2040, maybe 2035. You just can't leapfrog scale-up without consequences, and you see it over and over throughout history.

So for once I can intelligently comment on a Marginal Revolution article. (I have a Ph.D. in applied plasma physics and fusion energy; I worked on the "conventional" fusion reactor design, the tokamak). Lockheed hasn't released many details of their concept (at least, not enough details that it can actually be evaluated in technical detail), but it looks like it's a combination of a magnetic mirror and a levitated dipole. The magnetic mirror was studied in detail in the 1960s and 1970s and didn't work out (due to [detailed plasma physics reasons]) and the levitated dipole has a fundamental flaw as a power-producing reactor in that the superconducting magnets are inside the neutron shielding - neutrons destroy the magnets.

It's tough as a scientist to be able to comment on things like this, because it's "science by press release", i.e. there's a big media hype but the actual researchers don't release enough technical details to actually evaluate it. One wants to remain cautiously optimistic, but with fusion in particular, we've been down this road many, many times. Thus I predict that the most likely outcome is that as they scale their device up, they'll find that the confinement (a measure of how well the device holds a fusion plasma) unexpectedly drops off due to some different types of turbulence turning on at higher temperatures / higher pressures... and it will quietly go away.

I hope that I am proven wrong.

Geoff , I hope you are proven right, cheap powerful energy will not coexist well with chaos theory in this corner of the universe.

@wwjbd - you hope he is right about being proven wrong? So logically you're pro-power? But I do see your point, and you are wrong. Just because ignorant people exist who are wasteful, doesn't mean the rest of us should concede. That's a recipe for the status quo. Doing that, we all become just slaves to the middlemen of society, the Walmarts and such that do nothing but buy low and sell high to the masses. Which btw the law is geared to facilitate such middlemen.

@Geoff what do you think of http://lawrencevilleplasmaphysics.com/ ?

They seem to be developing a car sized device too.

Step right up, folks, and see Little Egypt do her dance of the pyramids.
She walks, she talks, she crawls on her belly like a reptile.
Just one thin dime! One tenth of a dollar! Step right up, folks!

Normally I'd dismiss this kind of announcement out of hand, but since it's the Skunkworks and they've had some impressive previous achievements, I'll mentally categorize it as a 5% probability.

Fission reactors last, what, 40-50 years? After that, the metal in the reactor core is weakened, through exposure to the neutron flux of fission.

What I've never understood: how can fusion be 'limitless', if the same neutron dynamics weaken the steel in the Tokomak??

BTW, fusion: still forty years away… ;)

What you appear to have here is a PR campaign related to a proposal for government-funded research.

There is no published research and no discovery being claimed. There is only an idea of how an experimental fusion-engine could be built much smaller and cheaper than the one (ITER) being built in France.

A lot of commenters seem impressed that Lockheed has put significant own money into this idea, but I don't see any evidence of that.

And in other news, Tom sez: "if man could fly, he'd have wings!". Nuff said.

Shouldn't we be looking at the price of coal or something instead?

At current international prices the fuel cost of a kilowatt-hour of electricity generated from coal is under 1.5 cents. Now that's not the only cost of coal. Operating an maintaining a coal power plant is likely to cost at least another 1.5 cents a kilowatt-hour and then there are the cost of capital and externalities (which are very high) on top of that.

By the way, my comment above on coal was meant to be a reply to Saturos in case she or presumably he is interested.

Rather than trying to reinvent the wheel and building another fusion reactor, why not use the energy from the existing one: The Sun?

It's inconveniently far away and half the time is blocked from you by the planet that you are presumably on.

Seriously, dude, what are you trying to accomplish here? Like, holy shit nobody in this comments thread has ever heard of solar power? There is plenty of money and effort going into improving solar power enough to make it a major part of our energy generation. Lockheed Martin's efforts are not detracting from that.

"There is plenty of money and effort going into improving solar power enough to make it a major part of our energy generation. Lockheed Martin’s efforts are not detracting from that."

There is of course a similarity between the two: Decades of promises and decades of failing to live up to the promises.

In the 70's I was convinced that solar power would provide a significant chunk of the US energy budget. By the 80's I was doubting and by now I just laugh at the thought. 40 years go by and Solar is up to ... 0.23% of US energy generation.

In both cases, solar and fusion, I think we should let the track record speak for itself, and be properly dubious until some sort of provable (and economically sound) success occurs.

Rooftop solar now provides 6% of total electricity use in South Australia and is the cheapest source of electricity available to households throughout Australia. And this would still be the case even without subsidy or feed-in tariffs.

According to this source Australia as a country only gets 0.2% of it's electricity from Solar power.

http://en.wikipedia.org/wiki/Solar_power_in_Australia

JWatts, no it doesn't say that.

Point of use solar does present an interesting challenge for fusion or any utility scale generation. It's not uncommon now for electricity produced by point of use solar to cost considerably less for consumers than purchasing electricity from the grid. So much so that even if the wholesale price of electricity was zero it still would not be competitive with electricity produced by PV panels on rooftops. This is going to make problems for fusion which will be faced with low average electricity prices during the day and existing pumped storage that can save its capacity for the evening and so help keep prices down then. But if they can fit fusion reactors in the back of a truck, maybe it can become point of use energy for large industries, providing both heat and electricity.

On the other side of the door, scientists in the Laboratory for Organogenesis and Neurogenesis are working on something that has fired the imagination of science fiction authors for many years. They are at the cutting edge of an emerging field: rebuilding the body by growing tissues and organs from stem cells. They hope to develop the next generation of therapies for a variety of debilitating human diseases, and unravel the mysteries of brain development.

I wonder if this would also fit the "TGS over" requirements.

From: http://txchnologist.com/post/97816566470/the-deep-dive-the-man-who-grew-eyes

LMT isn't the only one building fusion reactors. I saw this in the NYTime a few days ago:

Days earlier, Y Combinator, known for aiding web and mobile-app start-ups like the social news site Reddit and the game maker Omgpop, took part in a $1.5 million early investment in Helion Energy, which is developing an engine powered by nuclear fusion.

http://www.nytimes.com/2014/10/13/business/venture-capitalists-return-to-backing-science-start-ups.html

It's science by press release, which trades at a deep discount. It's coming from an organization which AFAIK has a good reputation for seriousness, so the discount isn't as deep as it could be for other sources, but still, it should be a deep discount.

Also FWIW, even if it's everything we could hope for from the most optimistic reading of the press release, it's not everything you could dream of in an energy source. In particular, (1) this kind of fusion is just going to create heat (not e.g. some marvellous direct-to-electricity trick) so to use it to drive anything interesting like an engine (not just a space heater) will require big expensive machinery and (2) fusion produces neutrons to go with each joule of energy, and those will tend to transmute the materials nearby into somewhat-hazardous waste as the reactor is operated, so it's not likely to be as headache-free compared to fission power as you might like to think.

Don't get me wrong, it could be true and if true it would be great, but it's not the way to bet and even if it's great it won't be Star Trek Federation propaganda footage hand-wavey perfect side-effect-free greatness, it'll share in some measure the disturbing nuclear-energy-scale side issues that we associate with fission power (and we may not have the political technology to deal with them any more sanely than we deal with the fission power issues today).

They're basically just trying to build a magnetic mirror configuration (http://en.wikipedia.org/wiki/Magnetic_mirror). This is an old idea but a lot of effort was put into it by the magnetic confinement community (it was never anywhere near as successful as the Tokamak). The first issue you run into is that you lose particles directly onto your walls because the magnetic field lines don't close in on themselves (as is the case in a Tokamak or Stellarator). The mirror configuration tries to limit this but has never been particularly successful, especially since you're always going to be losing some of the most energetic particles. I don't see anything particularly novel about their approach - perhaps they just didn't do their homework on what has been tried in the past.

Their claims about being able to build a very compact reactor are rather bogus. In principle, one might expect a mirror machine to have excellent stability properties compared to say a Tokamak - this is why they claim they can make the reactor much smaller (since they would hypothetically be able to confine a much hotter plasma with a given magnetic field strength). However, because of how badly mirror machines have failed in the past (for more boring reasons - mainly the end losses) we haven't ever really managed to see if it really would be a stable configuration when the confined plasma pressure is comparable to the confining magnetic field pressure. I highly doubt it would be stable. The sorts of instabilities one typically worries about in A Tokamak would not be an issue, but there are other instabilities that pop up when the pressure is that high (firehose, mirror, etc.) We observe many astrophysical plasmas under these conditions (very high plasma pressure relative to magnetic field pressure) and they are wildly unstable.

Even if everything I just wrote is wrong, they will still have to solve some of the most crucial engineering issues that a Tokamak faces: how to find inner wall materials that can survive the heat and neutron loads without falling apart, and how to breed tritium. These are completely unsolved issues for a Tokamak reactor (though they do not look impossible), which would be much bigger and have much lower heat/neutron loads. A tiny reactor like this would have heat/neutron loads on the walls that are order of magnitudes higher than what a Tokamak reactor would face, and as I said, we're still not sure if we can find materials that will survive those conditions for long enough to make a reactor economically worthwhile.

As someone working on magnetic confinement research, I found this to be really poorly thought out. Maybe it's still worth attempting, and I wish them luck, but it's very frustrating to read another one of these ridiculous articles (and to see it get so much press, but that's unavoidable I suppose...)

From a engineering point of view, IF the CFR succeeds in creating fusion to generate electricity, it would have to couple to current large size steam turbine, elec generators and cooling towers, the saving in size is meaningless.

The only optimal use for the CFR with advantage in small size would be for plasma canon, which would be in the core business of LM.

Just a mere PR effort by Lockheed Martin, which lost its relevance long ago, when it shut down Skunkworks Burbank and completed the SR-71 program.

According to me the cost of the technology are much their begining statges, but when technology get improved their cost automatically goes down. And fusion reactor costs must depend on the energy and powers it has.

Comments for this post are closed