Published on Aug 22, 2012
Google Tech Talks
November 9, 2006
ABSTRACT
This is not your father's fusion reactor! Forget everything you know about conventional thinking on nuclear fusion: high-temperature plasmas, steam turbines, neutron radiation and even nuclear waste are a thing of the past. Goodbye thermonuclear fusion; hello inertial electrostatic confinement fusion (IEC), an old idea that's been made new. While the international community debates the fate of the politically-turmoiled $12 billion ITER (an experimental thermonuclear reactor), simple IEC reactors are being built as high-school science fair projects.
Dr. Robert Bussard, former Asst. Director of the Atomic Energy Commission and founder of Energy Matter Conversion Corporation (EMC2), has spent 17 years perfecting IEC, a fusion process that converts hydrogen and boron directly into electricity producing helium as the only waste product. Most of this work was funded by the Department of Defense, the details of which have been under seal... until now.
Dr. Bussard will discuss his recent results and details of this potentially world-altering technology, whose conception dates back as far as 1924, and even includes a reactor design by Philo T. Farnsworth (inventor of the scanning television).
Can a 100 MW fusion reactor be built for less than Google's annual electricity bill? Come see what's possible when you think outside the thermonuclear box and ignore the herd.
Because the battle between approaches to fusion power is not so much on the grounds of how easy they are to build, but how likely they are to generate more power than they consume.
I think there's a distinction being missed. There are two "creative" fusion ideas with real potential: reciprocal magnetodynamic-electrostatic confinement (polywell) and ultradense plasma (aka pinch/focus fusion). Dr. Bussard, Convergent Scientific Inc, the Iranians, and a few small universities are responsible for most polywell research. Most commercial startups, including Tri Alpha, are pursuing ultradense plasma. It's not surprising: ultradense plasma devices are smaller, thus cheaper to experiment with, but also more speculative/dependent on poorly-understood physics, making them a better fit for the private sector. Polywells on the other hand have some decent theoretical backing and the main trouble is trying to build one large enough -- IIRC, Bussard obtained some results which, if extrapolated, imply net power in a $100M device.
Looks like they achieved >5ms confinement of a 10^7 degree plasma.
For meaningful energy production, they would need say ~10s confinement of 3*10^9 degree plasma (300x the current temp -same as the temperature difference between the Earth and the surface of the Sun ). It's an interesting development, but I'd call the outlook very uncertain.
I've watched both the 'Should Google Go Nuclear?' and Microsoft Research videos on this topic.
Both are pretty good. The Google video has Bussard in it and was made not long before he died.
The Microsoft paper talks about, well a lot of things, but I recall one of the key research items was cusp confinement. The problem with using magnets to control a plasma is the plasma will reject the magnetic field as its density increases. The wiffle ball shape of the resulting plasma of the polywell allows the field lines to penetrate the plasma even at higher density.
In short: Tri Alpha demonstrated plasma confined for 5ms. So their concept works.
Their next step is burning D-T fuel (needs 10x temperature increase). Their goal is burning H-B fuel which requires much higher temperatures, but has numerous advantages.
Update: "Tri Alpha is backed by Sam Altman, among other things." -> not at all.
I've always wondered if this is truly the case. If the tech is proven to work, it'll leak eventually, and make its way across the world. Some countries might enforce patents on it, but most wouldn't (especially those with the greatest energy demands, India and China).
Actually they explicitly are not doing D-T fuel, staying with H-B fuel. And what they claim is that given a 10MW power source they could maintain a small ball of plasma indefinitely. So hook this up to a nuclear power plant and you have the worlds most expensive light bulb :-).
Since temperature scales with energy input, they scale up their box to produce the necessary temperatures and start collecting energy. However, as the article points out, not only is there little experience with plasmas at those temperatures, there are challenges with building things at that scale. So not a slam dunk.
That said, if we have a Fusion power plant that requires a Fission power plant to 'start' it will be a challenging thing to commercialize. Once you have a number running you could presumably 'grid start' (use excess power on the grid to start your new plant) but that would make cold start pretty hard. And if the scaling is accurate, you couldn't really turn these things on/off easily so they would have the opposite challenge of renewables which don't provide good base load support as one which can only provide baseload.
It seems all of the problems have the same solution. Store the excess power at night to use during the day. Then the stored power allows for a cold start. For the initial start the energy storage device can be filled over time from the grid.
Fusion is one of those areas that would benefit greatly from Public R&D. It's really sad that we've spent almost nothing on fusion research for the purposes of producing affordable electricity in the last 20 years.[1][2][3]
1. The NIF is and always has been a sideshow for the Nuclear Weapons development research that goes on there.
The way the climate is going, we desperately need good energy sources. If you compare historically to the space race, the manhattan project, or the spending on any major conflict you'll see that's a rounding error -- e.g. America's DoD burned through the entire cost of ITER (which is shared among 6 nations + EU) roughly every week circa 2009.
the question is "are we ready?". With nuclear energy we were on the brink of nuclear war during various periods of time, and the natural limitations of the technology help to limit the access to it. The success of fusion in non-Tokamak devices, something like inertial confinement and mixed schemas, brings the risk of a new class of weapons, something along neutron bomb (or boosted fission) without fission primary, which thus wouldn't be subject to technological limitations rooted in the fission.
We already have enough firepower with standard nuclear warheads to destroy civilization a few times over. I doubt a bigger bomb would fundamentally change the equation at all.
Honestly, I doubt this is something to worry about. Fusion happens through confinement; it's inherently unstable -- the more energy production, the stronger the confinement is needed (which is provided by a self-regenerating power source in a fusion reactor). This is everything you don't want from a bomb. If the reaction starts getting too hot, your confinement breaks and then the reaction stops. Scaling up the reactor doesn't change this, and it's not like it's viable to fly a reactor with a 100MW startup power source and drop into places anyway.
A fusion bomb necessarily needs a transient, extreme ignition, which I believe can only be achieved with a fission bomb (even with the truly massive fission energy building an H-bomb took a while!).
you're talking about Tokamak like confinement, [electro]static type. That isn't weaponizeable. I'm not sure that electrostatic can produce a space engine too. Regular electric generation at best. No, i meant inertial confinement.
>A fusion bomb necessarily needs a transient, extreme ignition, which I believe can only be achieved with a fission bomb
Inertial confinement doesn't really need much energy. It needs huge power, huge energy density (notice that power != energy). Fission primary is the best way to achieve that of course, yet Sandia Z-machine or LNL NIF lasers achieve power enough for ignition too while not spending any noticeable amount of energy (NIF uses old lasers so it consumes more energy than it would if it was built with today's lasers). Both devices - Z-machine or a NIF-like with the old lasers replaced by modern solid state ones - are already of the size on the scale of 10-20 shipping containers and are capable of burning something like a grain size pellets which creates neutron flux deadly in the area of the size like that LNL NIF building. To me it seems like there are really great chances that either of them would be miniaturized enough to produce a space engine as well as a weapon (to which our civilization doesn't look ready).
Most of the funding I've seen has gone into maintenance for existing projects, nuclear weapons research, and the ITER boondoggle. But if you have any sources of projects that have received billions of dollars that have been about power creation, I'm 100% open to changing my tune.
Check out the US DOE budget [1]. For fusion energy alone they spent ~$400 million per year for the last several years. In addition they spent ~$1.3b on nuclear physics and high energy physics. From my experience as an undergrad lab assistant I can tell you that many of the funded physics experiments are indirectly related to fusion research because they study the byproducts of our best understood fusion reactor, the sun.
For anyone who reads this comment and is confused about what we are talking about.
On page 17, Fusion research for the NIF of ~480M is outlined. We could debate for a long time about how this is more of PR move, and not realistic. In the parent, this is where I talk about Nuclear Weapons research. For this to make power you literally need to drop millions of pellets of gold plated deuterium one after the other to create 'power', all the while firing the lasers like 100X faster than they can currently be fired. You decide if that's attainable.
On page 44, the supposed ~400M a year is outlined and in the discussion notes. "The FY 2014 Budget Request funds U.S. contributions to the ITER project for long-lead procurements required in construction of the facility; the majority of these contributions will be spent on in-kind hardware sourced from U.S. industries, national laboratories, and universities" Which is basically a retelling of my comment about the ITER 'boondoggle'. There is also a gov't sentence on how they money will go to existing projects, which is really lip service to Fusion research and mostly used for Public Education (very good!)
General research on the 'sun' is very far afield to say that it applies to harnessing fusion to create electricity.
As a percentage of annual budget and or GDP it is a very small amount. And in the grand scheme of is an even smaller amount given the predicted costs of climate change.
If he pays taxes then it's partially his money, and he has a right to vote on how it's used and tell people his opinion.
Arguing that his idea would be spending "other people's money," is a non-sequitor. Following that logic nobody should have any say in how the government spends money, because no single person is contributing the majority of funding.
That graph is dug out whenever the feasibility of fusion is discussed. It's a prediction from 1976 – you should know how credible decade-long predictions on the progress of unproven technologies are.
How does this address the graph? It probably doesn't predict the current situation, but it certainly is consistent with it, that fusion funding is below the "fusion never" line, and look, we have no fusion.
I could be wrong, but the projections are extrapolations based on obviously overly optimistic assumptions about the a) the federal government's desire to fund large, multiyear scientific projects b) the ease of creating an economically scalable fusion reactor
You have a fantastic mindset in that you can call multiple billions of dollars "almost nothing"
Well, when you consider that the Iraq war cost to the USA has been at least 1 trillion (this is not just what the DoD spent), spending tens of billions on fusion research doesn't sound so bad.
NIF has nothing to do with Nukes. Zero, zilch, nadda. It doesn't even vaguely work anyway.
Public funding for fusion research in the US and Europe is a waste of money; it's simply providing employment for past generations of fusion researchers, and pouring money down a hole on technologies which will never work.
Tri-Alpha, on the other hand, has something new that might actually work. I went over the patents a few years ago (was going to do a blog on them; got busy); it's quite different from the approaches that get money via public funding, and unlike those approaches, it stands a chance of working.
The summer of 1999 on the way to E&M class I rode the same bus to the physics building as Dr. Hendrik Monkhorst. He is one of the founders of Tri Alpha Energy, and I remember chatting with him about about aneutronic fusion. It was very eye opening for a 19 year old undergraduate. I remember him espousing commercialization within ten years, which now seems like prototypical professorial optimism. It is an exciting milestone to see them have successful confinement for a solid length of time.
Of course. Nobody in power wants fusion, as most of the political class have vested interests in petrochemicals. We're all going to die in the next few decades. Such is life.
Oh I thought it was a reference to the fact that the Sun's (about 8 light-minutes away) been doing fusion at stable rate due to it's enormous gravity this whole time. Maybe I misunderstand what is meant by gravitic containment?
You seemed confused when you were talking about the time to get to LEO. I was just clarifying what I thought the "eight minutes away" comment referred to.
Perhaps a stupid question, but why does the plasma need to stay confined for long periods of time? If you can devise a process that performs a full power-generating cycle, then no matter how long (or short) that cycle takes, you will get net power out of it.
1. Ions in the plasma need to interact with each other for enough fusion reactions to occur. For this to happen, the plasma needs to be somewhat dense and the temperature needs to be very high. If the plasma escapes, the temperature and densities fall off rapidly, and the nuclear fusion rate drops quickly. (Nuclear fusion has an incredibly steep temperature dependence --- the triple alpha process, for instance, goes as T^28 or something like that).
2. When the plasma escapes, it interacts with whatever is containing it and can destroy it very quickly.
But I'm an astrophysicist, not a nuclear physicist, so I could very well be wrong. Plasma confinement is much easier in stars. Just let the gravity do the work for you. :)
Quite right. The figure of merit for a fusion plasma is called the 'triple product' since it's the temperature, density and confinement time multiplied together.
The problem in general with pulsed systems is that they take a lot of energy to get them from idle to active, and the research machines have miniscule efficiency. The claim is often that "we'll scale this up and it'll be more efficient" but there are always unknown unknowns. Nature is a cruel mistress, especially in the fusion business.
I forgot to address this in my earlier comment, but you don't worry about the plasma escaping -- it is very delicate, generally orders of magnitude less dense than air. Rather, keeping it alive is the difficult part. If the plasma hits the wall of a fusion device in an uncontrolled manner, it will dump all its energy into it. If the plasma picks up too many impurities, it will radiate energy away in the form of bremsstrahlung. It is hard to keep the ions hot enough to fuse while there are many phenomena conspiring against you.
Interesting question. I don't have the answer but it could be because it's costly to get the whole thing running. Once it's up, you want to keep it up to avoid all sorts of problems/costs. And that's not considering how long it takes to generate said power from the plasma.
Might be like an MRI machine. You could quench the magnet every other month, but at what cost and benefit?
Great infographic, but unfortunately it misses my favorite juxtaposition of temperatures:
Surface temperature of a red dwarf star (e.g. Wolf 359) 2500 C
Melting point of tungsten: 3400 C
I find the idea of making balloon-like objects out of tungsten and gas, with a density less than that of the star's photosphere, and floating them around on the surface of a star to be intriguing. It would be a great location to put a heat engine. A totally sci-fi idea, I know, but still interesting to think about.
Umm, actually, a "heat engine" needs both the source of heat and a place to dump the waste heat into, so the surface of a star would be a bad place to build one.
The only way of dumping heat would be to radiate it out to space, but "radiation" is a very inefficient way of losing heat. Unless I'm missing some really clever trick, soon your radiator will become about as hot as the surrounding gas, at which point the efficiency (= (T_hot - T_cold) / T_hot) drops to near zero.
You are correct that a heat pump needs both a heat source and heat sink. However, radiation is quite effective at transmitting heat at high temperatures. Black-body radiators emit as the fourth power of the temperature, although real objects never emit quite as efficiently as ideal black-bodies. It works pretty well for stars, though, since radiation is how they lose the vast majority of their energy.
If the photosphere above the heat engine was opaque, then the heat engine would not work. So it makes sense to keep the heat engine near or above the top of the photosphere, without going high enough to overheat near the top of the chromosphere.
It's mainly just a fun idea. I have no plans of trying to build one in the near future. :)
* I am a physicist, but that is no protection against being wrong. ;)
FYI David Brin explored this idea in the novel 'Sundiver' (1980)[1]. In the novel a laser is used to radiate heat out into space. This is the first book of the Uplift Trilogy. Incidentally, the second book, 'Startide Rising' is really great imo, and won a bunch of awards.
Aye, actually Sundiver is my favorite of David's books. I think his writing is better in later books, but I enjoy the audacious technical ideas and the classic 'closed room mystery' plot.
I met David earlier this year at the NASA NIAC symposium, and spent a nice afternoon hanging out with him and Joe Haldeman and his wife. Very nice people! We toured the Swampworks and launch sites at KSC, and talked about practical methods for moving planets. It was a very enjoyable day.
Very cool. You seem to be familiar with him and his work, so I'll ask this; I seem to recall that him and Vernor Vinge have sidelines doing 'scenario planning' for government agencies, is that something you've ever heard of? I've always wondered what that consisted of. Maybe I'm just searching for a reason to explain Vinge lack of productivety and imagined that?
I'm sorry. but I don't know much about that. However, you could ask them about it. Most sci-fi authors I have spoken with are very open to answering questions, as long as they are addressed in a respectful manner. I don't know about Vernor Vinge, but I know that David Brin has a website that you could email him at.
The concrete in the walls surrounding you contains Uranium. That alpha-decays from time to time. The system consisting of a single, former Uranium atom that just alpha-decayed, together with its alpha fragment, has a temperature in the tens of billions of Celsius/Kelvin.
(Yes, questions about the validity of defining temperature in such a system arise :)
If only any of these passes the lab stage to start with engineering... ITER is the only fusion machine 10-20 years away from demonstration, in 2015, and it's too big to fail.
This has the advantage of not needing a power recovery system and the actual power can be external (like solar) and the fusion engine would act more like a souped-up ion engine (with correspondingly ultra-low thrust).
https://www.youtube.com/watch?v=rk6z1vP4Eo8
Published on Aug 22, 2012 Google Tech Talks November 9, 2006
ABSTRACT This is not your father's fusion reactor! Forget everything you know about conventional thinking on nuclear fusion: high-temperature plasmas, steam turbines, neutron radiation and even nuclear waste are a thing of the past. Goodbye thermonuclear fusion; hello inertial electrostatic confinement fusion (IEC), an old idea that's been made new. While the international community debates the fate of the politically-turmoiled $12 billion ITER (an experimental thermonuclear reactor), simple IEC reactors are being built as high-school science fair projects.
Dr. Robert Bussard, former Asst. Director of the Atomic Energy Commission and founder of Energy Matter Conversion Corporation (EMC2), has spent 17 years perfecting IEC, a fusion process that converts hydrogen and boron directly into electricity producing helium as the only waste product. Most of this work was funded by the Department of Defense, the details of which have been under seal... until now.
Dr. Bussard will discuss his recent results and details of this potentially world-altering technology, whose conception dates back as far as 1924, and even includes a reactor design by Philo T. Farnsworth (inventor of the scanning television).
Can a 100 MW fusion reactor be built for less than Google's annual electricity bill? Come see what's possible when you think outside the thermonuclear box and ignore the herd.
Google engEDU Speaker: Dr. Robert Bussard