What the hell happened to optimism and trust in science? We used to be able to say "In 10 years there'll be a man on the moon" without having ever launched anyone in space. Now people doubt that in 50 years we will have a marginally better battery technology...
Fusion would be there already if only we had believed in it.
It's strange, you have to be educated somewhat to make a comment this ignorant.
Where exactly do you think the internet came from that you are using right now at this moment? Where do you think a multitude of inventions and technologies came from?
A huge chunk of that DoD budget is specifically for research and development.
It's bizarre it's like people learn just enough to solidify their world views then turn off.
Just so you realize, the DoD budget does pay for a lot of research. About 15% is directly going into research, and there's about another 10%-ish that's research disguised as procurement. And that research isn't all weapons system--I've been funded by the DoD to do compiler research.
Far less than 25% of the DoD budget is research. Quite a bit could charitably be called R&D with a heavy emphasis on development, but even then it’s not 25%.
That said, they do a lot of great research and some of it even stays unclassified.
I assume your referring to: “RESEARCH, DEV, TEST & EVAL“ which yes includes the word research, but that’s far from the only thing involved. Which is why it sticks on Dev Test and Eval.
Some of that money involves setting up assembly lines.
If we only believed more in the philosopher's stone!
There are fundamental problems in achieving economically viable fusion. Nice and clean reactions that power the sun are way too slow, and need immense volume. Faster reactions produce neutrons and still require temperatures several times higher than inside Sun.
Where economy and safety are not the primary concerns, fusion has long been achieved: in bombs.
It is because we believed in it that chemistry was invented! The worst case by pursuing an impossible thing is that you acquire a ton of knowledge in the process.
If ITER fails, what we will have learned about superconducting magnets manufacturing and high-temperature plasma will surely be useful.
Optimism is really, really hard. It requires you to risk looking foolish. Our emotional needs to feel connected work against us staking a position too far away from the norm.
But few great things have been accomplished without the risk of looking foolish, at least at the outset.
What? I constantly see people being optimistic based on nothing more than misplaced wishful thinking. Optimism is the default, officially approved thing you're supposed think about fusion. It's being skeptical that requires effort, since it requires doubting what the media and authority figures tell you.
I’m still optimistic about science in general, but sometimes I question the wisdom of our resource allocation: Tech companies making selfie sharing apps make enough money to fund an entire space program several times over, but instead reinvest it in making the next generation of selfie sharing apps.
And because all the money is in selfie sharing apps, that’s where a large percentage of the brightest minds go to spend their career (as opposed to something less financially responsible, like nuclear fusion engineering, astronomy, or space flight)
That's because that's what society wants. People pay for apps or watch ads to get their selfie app. SpaceX lands 3 boosters back on earth and fewer people watch than daily active Snapchat users.
It turns out bright minds love science, but they also love a decent standard of living and not having their career jeopardized be every time Congress tries to pass a budget.
And isn't that a bit of a dig at our astronomers and fusion engineers?
This chart always gets trotted out whenever this comes up.
You cannot know what it will cost to invent something before it is invented. You cannot know how many deadends you will have to go down before you find one that works. In fact, it may be that the invention depends on knowledge or technology that we don't have, and maybe won't have for many many years.
For example, today fusion researchers are using computers to simulate plasma flows, gain a better understanding of plasma, and design better reactors... something that was not possible in the mid-70s when that chart was created.. and would not be possible to do by hand (as was done in the 70s). Hundreds of billions of dollars have gone into creating computers capable of doing that... where is that in your chart?
And to preempt this line: saying they could buy a supercomputer is not a rebuttal. A Cray-2 was the 2nd fastest computer in the world from 1985-1990 (10-15 years after your chart). It's the equivalent to an ipad 2 (a 2011 32-bit mobile low-power processor). I'm using a computer many times faster than that just to type this post.
The chart refutes the idea that experts promised fusion power on a certain timespan, got the resources they requested as necessary, and then failed to deliver. That false narrative is the basis of the incorrect aphorism "fusion is 30 years away and always will be".
The chart is not intended to be definitive proof that we would have had fusion power if we had spent the money (lv's last throwaway line notwithstanding.) No one knows what would have happened. To estimate that, we must use our prior based on how long/expensive other major tech developments have been. That's very different than pretending we know fusion's track record.
Incidentally, NASA is currently spending huge amounts of money on computer simulations for the equipment that will return humans to the moon half a century after they were last there. That doesn't mean those simulations are necessary to get to the Moon, as indeed they weren't possible in 1969.
In addition to what you said, I think our perception of the Manhattan and Apollo projects is skewed by survivorship bias [1]. These projects kept hitting milestones, so they kept getting additional funding, until their successful final delivery. It's quite likely that many, many more projects (laser weapons, scramjets, ekranoplanes) experienced a funding flatline like the fusion famous graph, simply because their milestones proved elusive. It's better to deliver results than excuses and delays.
Very interesting graph, thanks for sharing. I guess the point of the graph is that the progress in fusion has been steady and followed an exponential low for 3 decades, in a way that matches the Moore law.
If I were a politician in 2019 with the right to sign funding and I see this graph, I cut all funding right away. Here's why:
* based on this graph, a simple extrapolation by only 2 years show the achievement of the commercial reactor in 2005. Where's the commercial reactor 14 years later? We are instead only 6.5 years away before pushing a certain button at ITER, which has some type of significance that's not entirely clear
* at least one point on this graph is "cooked": the JT-60U that appears to cross the breakeven point at some point prior to 2000. Here's what wikipedia [1] says about this:
"During deuterium (D–D fuel) plasma experiments in 1998, plasma conditions were achieved which would have achieved break-even—the point where the power produced by the fusion reactions equals the power supplied to operate the machine—if the D–D fuel were replaced with a 1:1 mix of deuterium and tritium (D–T fuel)."
What's the point of the "figure-of-merit" graph then? It projects when we'll get to breakeven or commercial, only to find there that you need to make another tiny change (i.e. substitute D with T) that will take 30 more years?
Berlin has been building an airport for fifteen years now. Building large things takes time, especially if it's supposed to be an international collaboration.
ITER is far behind that growth curve because its cost is beyond what a university or national research lab is filling to fund. To unlike a lot of other projects before it, it is this massively political beast. Involving politician by itself introduces at least a decade of delay.
The triple product is the correct figure of merit and JT-60U did great there. But it was a plasma experiment and did not have the shielding that is necessary to run it with D-T fuel that produces a lot of neutron during fusion. But we know a lot more about neutron fielding than we know about magnetically confined fusion. Improving the plasma handling and the triple product is much harder than adding shielding.
> The triple product is the correct figure of merit and JT-60U did great there.
Look. If you have a graph where you show a certain point called "figure-of-merit" being above a so called break-even point, every layperson in this world will assume that you get more energy out than you put in. If you then say that it's breakeven for a different reaction, they'll either question your sanity, or they'll accuse you of bait-and-switch. If you don't understand why you get your funding cut, you'll think life is unjust or cruel. Other people will nod sympathetically when you'll complain of the unfairness of this world , but secretly they'll disagree with you.
The money comes from them, you know? One way or another, they are in charge of funding, and the way to help them make better decision is to be truthful, not to massage the message (you want to tell me it's really ok to compare a figure of merit for a reaction with the breakeven point for another reaction? * )
So, here's what Zubrin has to say about nuclear fusion and ITER [1]: "While, driven by international rivalry, the world's national fusion programs did advance forcefully between 1960 and 1990, the decision to consolidate all of them into a unified global effort to build the International Experiment Reactor (ITER) has caused nearly all progress to screech to a halt since the 1990s."
Why did this happen? Could it be because laypeople in charge of funding listened to scientists lobbying for a cool (and expensive) magaproject?
But you can point out that the actual funding defeated even the optimistic predictions about the cost.
I am merely pointing out that fusion did not happen, not because it is inherently difficult or because it is a mirage, but because we did not fund it appropriately.
> where is that in your chart?
Way off the right. Fusion was supposed to have been achieved before Windows 95 was out.
Those projections were based on a scenario where tokamaks performed REALLY WELL, and a crash program to develop them was pursued.
But the scaling of tokamaks turned out to be much less favorable than that scenario assumed, and even if it had been favorable that effort was ignoring engineering considerations like limits on wall power density.
When that scenario was revealed to be a mirage, funding declined. Not sure why anyone would have expected otherwise.
The change you're observing is a response to the fact that we've implemented most of the low-hanging fruit. What we're left with are much more difficult problems.
Does anyone know the risk of meltdowns for fusion? I know the energy released is several orders of magnitude larger then that of fission(what nuclear power plants do now), so I would assume a meltdown would be several orders of magnitude worse.
If we can create energy through fusion safely it really is unlimited carbon neutral energy. There is a limit, but the natural end of the planet will happen first.
This is great news! I had no idea we were this close to fusion.
It depends on what you mean by worse - as I understand it, a fusion-induced meltdown at least wouldn't release massive amounts of radiation. Massive amounts of heat and destruction to the immediately-surrounding infrastructure? Could be. But not fallout, like caused all the really big problems post-Chernobyl.
As far as I know there is no risk for a meltdown. The amount of energy/fuel in the reactor at any given moment is pretty low. Unlike with a fission reactor they have to keep feeding the reactor constantly or it will stop.
Armchair opinion: I believe the fuel isn't radioactive, and neither is the reaction. So while it could blow up, it's not going to have a ton of radioactive fallout.
Sort of like a fusion bomb, the fusion bomb doesn't produce radioactive fallout, but the fission bomb necessary to start the fusion bomb does.
D-T fusion, which ITER will use, produces neutron radiation, which will cause neutron activation in reactor components, rendering them radioactive: https://en.wikipedia.org/wiki/Neutron_activation
Aneutronic fusion is possible, but requires collision energies much higher than what ITER is designed to create.
In theory you can chose construction materials that don't do anything too horrible when you bang them with neutrons in the construction of your fusion reactor. I'm not sure how well this is achievable in practice in the construction of a fusion reactor but I'm pretty sure they can avoid using anything that creates byproducts as nasty as uranium does.
EDIT: I do wonder if there's a danger of nuclear proliferation with using fusion reactors to breed plutonium from natural uranium, though.
Yes and no. Depending on the fusion process, the fuel can be radioactive: DT for example, Tritium, Hydrogen-3 is a beta emitter. But it certainly isn't as bad of a thing especially compared to Uranium and friends.
EDIT: see other comment, DT produces neutrons which can be bad, I still think the amount nastiness is small compared to say Chernobyl.
> But it certainly isn't as bad of a thing especially compared to Uranium and friends.
Tritium is something like 10 orders of magnitude more radioactive than uranium (in terms of its specific activity). Uranium itself is so low in radioactivity that the bigger risk when handling it is heavy metal poisoning. The by-products of fissions are much more radioactive, though.
If you set of a thermonuclear bomb in the middle of the air you'll get a bit of fallout from the primary stage that sets it off, the secondary, fusion, stage will produce a bunch of neutrons but no fallout, and the third, depleted uranium, stage will absorb all the neutrons from the fusion and produce a fair amount of fallout. A lot of neutrons will escape and interact with the air but none of the elements you find in the atmosphere turns into anything much worse than deuterium or carbon-14 when you whack it with a neutron so overall its fine. You can replace the third stage with lead and lose half the explosive force but gain an almost entirely clean explosion.
If you set it off near the ground many of the escaping neutrons will interact with things that get very angry when activated by a stray neutron. A big bomb detonated near the ground, to crack open a bunker for instance, can send stupendous amounts of fallout into the air and kill people hundreds of miles down wind of acute "vomit now and die in a week" radiation poisoning.
If your supply of Deuterium and Tritium was jammed wide open you, would stop the fusion (either by switching of the heating if you are in startup or by messing with the magnetic confinement to make the plasma hit the diverter). At that point all you do is pumping low density, relatively low radioactivity gases into a large gas proof vacuum vessel. No explosion, no meltdown, no release of radioactivity to the containment building let along the environment. As far as worst case scenarios go it is pretty harmless. The entire ITER site will never store more than 4kg of tritium, most in the form of metal-hydrites.
Too much fuel would estinguish the fire, similar to too much fuel in a car engine will cause it to stall.
The high temperatures necessary for fusion are only attainable at low masses, if too much fuel enters the reactor, it cools down and becomes inert. The fuel itself is not radioactive and safe for consumption in reasonable amounts (heavy water inhibits biological processes if it becomes too much in concentration, but unless you drink straight from the heavy water tanks for days, this is essentially not an issue)
Fusion reactors cannot melt down, if the containment fails, the plasma will be cold by the time it reaches the walls. Wikipedia has a decent rundown of the various failure modes and resulting effects.
Meta, but given frankbreetz seems be just asking a question doesn't mean they have to be downvoted for being ignorant. I don't think they mean to misinform.
I think the situation is reversed for fusion as it is for fission. Fusions requires fine control and energy input to keep going. Where fission you need fine control and constant energy removal to keep the process from running away.
Also when the plasma in a TOKAMAK escapes and hits the wall it strips off heavy metal ions which radiates all your beam energy away.
Let’s say ITER works exactly as planned at some point in the future (delayed or not). How far are we from commercial fusion power as a reasonable source for things?
Far enough away that even if there are no issues with scaling, we're going to face a climate change catastrophe, before we can replace fossil fuels with fusion.
We need to take drastic action on power generation in the next decade. There isn't a snowball's chance in hell that fusion will make it in time. I can't over-emphasize how dire our current situation is - and that's with us currently living in a world where 3 cent/kwh solar power exists.
If we're gambling on fusion, at this point, the best-case outcome for it would be that it could be brought online sometime between us hitting 'drastic climate catastrophe' and 'civilization ending climate catastrophe'. To me, that doesn't sound like a great situation to be in.
There is no civilization ending climate catastrophy on the horizon.
Wipe out 90% of the population and we're back to what, the 1800s? That level of death requires most people to just let the rising ocean drown them as they stand there plus a bunch more inland to walk in and drown themselves as well.
Massively distruption to the status quo is what climate change will bring. Civilization ending? Not even close.
One million refugees caused political crisis in Europe. What do you think will happen if half of Africa starts moving north because agriculture becomes impossible? And that's just the easy part. What happens when important ecological systems collapse? For example if phytoplankton dies off at a massive scale?
I upvoted you for adding some perspective often lacking. However you are still kind of wrong. You are right that humans will continue to live and survive.
But our civilization as we know it may very well collapse. Yes some sort of civilization will continue to exist. But this will be on par with the dark ages: not a complete destruction of human civilization but an enormous regression.
We will most likely loose a lot of our technological know-how and abilities. Events like this don't simply cause a gradual decline in standards of living. Rather they tend to cause massive outbreaks of wars, unrest, migrations, refugees, destruction etc.
Climate change in the bronze age caused nearly all major bronze age civilizations to collapse. But with this collapse we had major migration of people, wars, invasions, collapse of trade and many more things. What followed was large population declines as well as losing the ability to write and maintain organized rule, construct buildings etc.
Did all humans disappear? Of course not, that is why civilization reappeared again eventually at a later time. The same will likely happen with us. But how long will the dark ages last?
In many ways our resilience is weaker than theirs. Our trade networks are far more extensive and important than the ancient ones. What happens e.g. if the middle east, Africa and South America are turned into turmoil and we can no longer get resources extracted from either of those places?
Our economies may suffer a lot too, causing the breakout of widespread riots and unrest, overthrow of governments. If modern trade breaks down, there will be widespread famines even in the west, because many countries simply are not self sufficient. And if they are, their food production depends on a large number of imports. If they don't get hold of artificial fertilizer, fuel etc food production will collapse.
This is what we saw after the fall or Rome. Economies became far more regionalized/local. In a collapsing world economy, countries will become significantly poorer if forced to rely exclusively on local production of most essential things.
> Wipe out 90% of the population and we're back to what, the 1800s?
1. Except without access to easily accessible sources of oil, minerals, and energy.
2. If you think that rising sea levels are the problem [1] with climate change, I'm not sure you understand it, at all. This seriously undermines your points. The actual threat is disruption of agriculture, and ocean ecosystem collapse.
This fantasy is almost more intellectually insulting than straight up climate denial.
What kind of dream world do you inhabit where you think it’s easier to travel to other stars or terraform planets than simply not “shit where you eat” on this planet? Slow down our economic growth? Stop dumping trash? Nah, that’s too hard. Let’s build fucking rockets to other solar systems instead and invent entirely new life support systems to survive in places without any food, water, or air at all. Sure, that’s way easier than planting a garden.
Even if this were possible, the technology almost certainly wouldn’t be ready by the time civilization fully collapsed at the end of the century.
"What kind of dream world do you inhabit where you think it’s easier to travel to other stars or terraform planets than simply not “shit where you eat” on this planet? Slow down our economic growth? Stop dumping trash? Nah, that’s too hard."
A lot of people do see the concept of some entity strictly controlling the entire world as unattainable, regardless of whether it would be desirable or not.
Seriously.
So, yeah, travelling elsewhere is easier, because it requires less coordination. Space travel doesn't seem as rooted in the insolubility of political problems.
Not trying to convince you this is right, just that the opposite of your opinion is "obvious" to other people who think differently than you.
My whole point is that space travel doesn't presuppose a universal decision maker, who would decide things like "it's better not to bring our problems to other planets".
To not have space travel, everyone must agree not to do it.
To save the environment, everyone must agree to do it.
Anything that everyone must agree to do probably won't happen and seems "difficult" to me.
How is it an intractable political problem when a large majority of the US, and a larger majority of the world, is in favor of strong governmental action to stave off climate disaster?
A political revolution against the corruption of our democracy by extractive industries is
much, much easier than going to other planets. It might get a bit bloody or require you to march in the street, but it will sure as hell be easier and cost fewer lives in the long run.
It is cowardly to assume that our political situation can’t change; it’s merely a matter of what kind of action you’re willing to take.
There are many ways that a majority can fail to get what they want even in a democracy. For instance, we have recently seen notable examples such as Donald Trump being elected President of the US, and the so far ineffectual efforts in the UK to exit the EU.
> What kind of dream world do you inhabit where you think it’s easier to travel to other stars or terraform planets than simply not “shit where you eat” on this planet? Slow down our economic growth?
I don't think any expansion steps looked less hard at the time.
DT fusion would be terrible for spacecraft, since 80% of the energy comes out as neutrons, and those neutrons need to be captured and thermalized in a blanket to make more tritium. A fission rocket would have better performance.
According to Wikipedia operation is supposed to start in 2033 with input from ITER design. No idea what’s right but I believe they want to build it in parallel with ITER, not after. Not sure how realistic this is.
That's completely unrealistic. The location and schedule of DEMO have not yet been agreed on. For that matter, whether DEMO will be the next major fusion reactor built has not been decided. The US position is that an intermediate facility, the Fusion Nuclear Science Facility, should be built and operated first, to develop and mature systems that DEMO would need (for example, fusion blankets cannot be developed without a strong source of fusion neutrons).
One cloud hanging over all this is that ITER is going to use up most of the world's tritium. With fission in trouble, especially heavy water moderated fission, there will likely not be a continued source of this isotope, and making it intentionally would be extremely expensive ($100M/kg and up.)
ITER has a design gross fusion power of about 400 MW.
If you compute the thermal power/volume ratio of the reactor, it's about 50 kW/m^3.
In comparison, the thermal power/volume ratio of a commercial PWR fission reactor is about 20 MW/m^3 -- 400 times higher.
A commercial tokamak would have a somewhat higher power, but not enormously higher. If based on ITER, the power density would be two orders of magnitude worse than a fission reactor. And since the complexity of a fusion reactor is far higher than a fission reactor core, the cost ratio of the reactors (for a given power output) would be even larger.
Since fission reactors are having desperate trouble being competitive today, how is making the reactor core orders of magnitude more expensive going to lead anywhere useful?
I believe your assumptions are faulty. Nothing about ITER suggest that it's volume or performance is indicative of what future reactor design will look like.
Stronger super-magnets, different designs could lead to smaller reactors.
As to why is it useful? A zero emissions powerplant, with minimal environment impact that is immune to core meltdown. Sure we don't have the tech to build it profitabily. Yet. But doesn't mean we never will.
I support the effort, but in terms of solving global warming problems this is IMHO a total dead end. The progress is far too slow. It takes far too long time to build these reactors. They are expensive and complicated.
Say we get commercial reactors in 2050. If these are anything like building fission reactors, we will start getting these commercial reactors online in 2060. That is very late.
Within that time we can massively expand solar power and wind power. As renewable energy start dominating the grid, ultra low spot prices on power will spur a huge cottage industry of businesses relying on buying this power cheap, storing it and selling it later. Alternatively factories will be built to take advantage of it. E.g. only running power hungry processes in periods of low power prices.
Sure. But you will need power after 2060. Just because you went neutral doesn't mean you suddenly don't need extra energy.
ATM fission, with its Chernobyl and Fukushima is still safer than solar and wind. And overall better for environment.
Fusion in theory promises even more - low environment impact plus no risk of disaster.
I don't see variability as a positive trait. From what I remember, renewables, really suck for the grid, as it wasn't designed for too much variability or diffusion in mind.
Given the rate of decline of cost of renewables, a fission reactor anything like ITER by 2060 will be museumware.
I don't think DEMO is ever getting built, btw. I suspect ITER is not going to have a very happy operational experience either (disruptions, maintainability). If ITER cannot effectively control disruptions they won't ever operate it with DT.
The idea that "renewables suck for the grid" is not correct. The very low levelized price of renewables means there's a lot of economic headroom to deal with intermittency.
> Given the rate of decline of cost of renewables,
Don't mistake current trends for future prognosis. Even if costs are falling at the moment, there is no reason future fusion technology will be museumware.
At worst we'll discover new data on plasma behavior. At best we have a possible tech for colonizing the Milky Way.
> I don't think DEMO is ever getting built,
It's ok. I don't think we'll transfer to renewables before it's too late anyway.
> The very low levelized price of renewables means there's a lot of economic headroom to deal with intermittency.
So you're saying that economics will change grid design? That grids designed for single way of transmitting energy will somehow reconfigure to allow consumers to stream energy to the source?
The trend of decreasing costs for PV has followed an empirical experience curve for decades, with an exponent of about -0.2 (giving cost as a function of cumulative production). There are perennially assertions this trend has ended, but those assertions have never been correct.
Is there any reason to think they will be correct now? I assert no: PV modules are still well away from the costs of the material and energy inputs.
I'll add that fusion economic studies have to posit large learning effects as well, so the "N-th of a Kind" plant is much cheaper than the first. And there, this assumption is dubious, since the closest technology (fission) has not shown good learning effects. That's why fission is failing now -- any technology that doesn't improve eventually gets killed as the competition does.
> It's ok. I don't think we'll transfer to renewables before it's too late anyway.
And you think the transition to fusion would be FASTER? My mind boggles at your double think.
> So you're saying that economics will change grid design? That grids designed for single way of transmitting energy will somehow reconfigure to allow consumers to stream energy to the source?
It's bizarre you think dealing with intermittency would require impossible, or even particularly difficult, changes. If you notice, grids are already becoming smarter, with power markets dispatching on short term prices. Storage and dispatchable demand fit nicely into this system. The market is going to force things in the right direction, which is renewables/storage and not high cost baseload sources.
> And you think the transition to fusion would be FASTER? My mind boggles at your double think.
No? I just think fusion tech is the future. A clean, abundant, uninterruptible, and independent source of energy that can be used on Earth or on Alpha Centauri.
I don't really expect to see them in my lifetime.
> It's bizarre you think dealing with intermittency would require impossible, or even particularly difficult, changes. If you notice, grids are already becoming smarter, with power markets dispatching on short term prices. The market is going to force things in the right direction, which is renewables/storage and not high cost baseload sources.
Again, that's not my experience. I've seen some countries move towards renewables, but nothing on the required scale. As for smart grids are you talking about decentralized grids, where consumers can send energy upstream or just more efficient energy grids?
And as for the markets, those depend on policies. If Trump says, we're making a renewables tax and tariffs on solar panels, those same markets could turn on renewables.
> I've seen some countries move towards renewables, but nothing on the required scale.
This is a form of a tired and invalid argument that one often sees. It's basically "nothing can ever happen for the first time." It also ignores the details of just what has been happening.
Renewables have declined in price very quickly. The timescale for this decline has been shorter than the lifespan of installed generating infrastructure. This is very unusual for power technology, so if one is not careful one may be misled, as you apparently have been. We're in a weird situation where that installed base is obsolete. It's still there, churning away, but only because its capital cost is written off. And some is being ripped out because it can't even earn its operating costs now.
What you want to look at is not this set of stranded assets, but the choices being made for new capacity. That is becoming overwhelmingly renewable. General Electric, for example, is in a world of hurt right now because the demand for gas turbines is collapsing. And nuclear fission or coal with steam turbines? Dead technologies.
Let me guess: you implicitly assume that cost is proportional to volume, because that is roughly true for a fission reactor, which is filled with fuel and control rods.
But in a tokamak reactor, you have a large container with all the expensive engineered stuff placed around its surface, so cost should be roughly proportional to surface area, or volume^(2/3).
Fusion reactors are MUCH more complex than fission reactors. So, just comparing volume is being very favorable to fusion. There should be a complexity multiplier in there as well, but that's difficult to quantify.
This complexity also means keeping that fusion reactor running would be a nightmare. Complex things break, and the fusion reactor will be so irradiated that hands on maintenance will be impossible.
I will add that fission power plants are uncompetitive even if you ignore the cost of the reactor itself.
> Fusion reactors are MUCH more complex than fission reactors. So, just comparing volume is being very favorable to fusion.
That's a non sequitur. Your volume comparison can, at best, tell you how cost scales with reactor power. If the exponent is wrong, it serves no valid purpose at all.
Complex things being more expensive than simple things is a non sequitur?
Surely your apparent position, that something made of superconductors is going to be just as cheap as something made of steel, is much more supportable! /s
> Complex things being more expensive than simple things is a non sequitur?
No. The statement "Fusion reactors are MUCH more complex than fission reactors." followed by "So, just comparing volume is being very favorable to fusion." is a non sequitur.
> And I wasn't talking about an exponent there.
Your compared power/volume ratio, implying that cost is proportional to volume (otherwise, the comparison is pointless).
I pointed out that the different operating principles of fission and fusion reactors suggest cost proportional to volume^(2/3) for the latter (which of course implies that there is a size beyond which a fusion reactor will be more cost efficient than a fission one).
The 2/3 is known as an exponent. I believe exponents are part of the K-12 math curriculum, and definitely propaedeutic to nuclear engineering.
> No. The statement "Fusion reactors are MUCH more complex than fission reactors." followed by "So, just comparing volume is being very favorable to fusion." is a non sequitur.
You seem to be implying that comparing by volume is unfair to fusion. So, instead, let's compare by mass. The mass of ITER is 23,000 tonnes, for an average density of about 2.8 g/cm^3. I don't have the figure for the mass of a PWR reactor vessel in front of me, but it's mostly empty space (filled with water and fuel in operation, which are not part of the capital cost). So, comparing by volume alone may actually ALSO be giving fusion an advantage.
> I pointed out that the different operating principles of fission and fusion reactors suggest cost proportional to volume^(2/3) for the latter (which of course implies that there is a size beyond which a fusion reactor will be more cost efficient than a fission one).
I don't think scaling data for cost of fusion experiments supports this.
BTW, as a fusion reactor gets bigger, at a fixed geometry and magnetic field strength, the mass of the material needed to support the magnets (to resist the outward force the magnetic field puts on the coils) is proportional to the volume of the reactor, not to the surface area of the plasma. And you want the reactor to be at as high magnetic field as possible, since the fusion power density goes as B^4. If wall power loading prevents one from taking advantage of this the reactor is too big and will have inferior power density, compared to a smaller reactor at higher B. This is why concepts like MIT's ARC have power density an order of magnitude higher than ITER (but still very much less than a PWR).
> The 2/3 is known as an exponent. I believe exponents are part of the K-12 math curriculum, and definitely propaedeutic to nuclear engineering.
> You seem to be implying that comparing by volume is unfair to fusion
I am pointing out an obvious logical fallacy. If it really isn't obvious, try substituting "Tesla cars" for "fusion" and "bars of gold" for "fission reactors". Here's what you get:
"Tesla cars are MUCH more complex than bars of gold. So, just comparing volume is being very favorable to Tesla cars."
Which of course is utter nonsense. A Tesla Model 3 occupies roughly 10 m^3 (order of magnitude, mind you); that volume of gold would weigh about 19.3e4 kg, which at the current price of $45.6e3/kg would be worth about $8.8 billion.
Clearly, the MUCH more complex Tesla Model 3 costs quite a bit less than the same volume in gold.
> So, instead, let's compare by mass.
That would be equally pointless. Instead, how about taking a moment to look up a serious economic analysis? Here is a fairly recent one, open access and all:
It compares existing power plants with four tokamak conceptual designs which "span a range from relatively near-term concepts, based on limited technology and plasma physics extrapolations, to a more advanced conception". Upon reading it, you will be happy to learn that it finds Total Cost Of Electricity (TCOE) produced by those fusion reactors to be slightly higher than the lowest cost alternative, fission reactors.
All other options considered (combined-cycle gas turbines, open-cycle gas turbines, coal plants, solar large photovoltaic plants, onshore and offshore wind plants) turn out to be more expensive.
It's amusing you believe those projections, especially when projections of the cost of fission power plants -- a much simpler and more mature technology -- keep coming in 3x over estimates, and when ITER blew through its initial lowballed cost estimates.
Think about the motivations for the people making estimates of the economics of fusion. They have little external motivation to be honest. The worst they could suffer if they were caught cooking the books was to lose their professional positions: but that is the same thing that would happen if fusion turns out to be seen as unaffordable and work is shut down. And really, a fusion power plant is so far in the future that they aren't likely to be caught on any but the most blatant dishonesty anytime soon. And who is being funded to check on their estimates?
So what might they do in that situation, even if they didn't flagrantly lie? They could tweak the assumptions going into the estimates, and keep tweaking them until they get results that are acceptable. If some source gives an input to their cost model that seems too good to be true, they might not question it too much. It's not "lying lying", they might tell themselves. They might even rationalize it's only a venial sin to bend things a bit, since fusion is so obviously a good thing to work on, right? Motivated reasoning can be a powerful thing.
All those failure modes are there, and there's incentive for them to occur, and there's little real world feedback to keep them in check. So you should take any estimates like the ones you linked to with more than a few grains of salt, especially when they fail a "smell test" by giving better economics than simpler, smaller fission.
In contrast, costs for renewables face the harsh test of actual experience, many times a year. Renewable cost estimates are, as a result, highly grounded in real evidence, and typically come in within 10% of the actual values. There is no room for lying -- to oneself or to others -- there.
The paper explains the methodology and lists references to the data. You can go through them and check the calculations yourself.
> Think about the motivations for the people making estimates
That's another logical fallacy, a classic one known as "ad hominem". Their motivations are utterly irrelevant to the question whether their results are correct.
That's now how it works. If they put together a big, complicated cost model, it's up to them to provide evidence it's valid. Otherwise, it's just an exercise in "believe us, we're honest". For example, they could apply the same model to fission reactors and show it accurately predicts what they end up costing. Oddly, I don't recall fusion cost modelers ever doing that. Hell, I don't recall fission cost models being terribly accurate.
You like this result because it's telling you what you want to hear, not because it's intrisically believable. More motivated reasoning.
Yes it is. They presented their work, their sources and conclusions. If you believe it's wrong, but can not explain why, all you have is a prejudice. If you want anyone to take you seriously, you need to provide a rational argument explaining exactly how they are wrong, not just why you want to believe that they are.
While there are many hopeful alternatives to “big fusion” which may promise to be cheaper, is it not true that ITER is the most credible path to commercial-scale fusion power?
I've been watching the progress of the Wendelstein 7-X (https://www.ipp.mpg.de/4550215/11_18) which has been making steady progress month after month, year after year. At the current pace they will demonstrate net power fusion before ITER does. Extrapolation is always risky though so it isn't something one should count on.
The Wendelstein 7-x is not designed to ever reach ignition. It's a (very large, very cool) research reactor studying stellarator geometry. Ignition will require a yet larger device, and barring any disasters, ITER should definitely get there first.
I understand that, these are the bits that I like about the project:
"Then, it will remain to be seen whether Wendelstein 7-X can also fulfil its optimisation goals during continuous operation – the essential advantage of stellarators." ... "Although Wendelstein 7-X is not designed to generate energy, the device is intended to prove that stellarators are suitable for use in power stations."
And from the ITER Goals statement (https://www.iter.org/sci/Goals) "ITER will not capture the power it produces as electricity, but as the first of all fusion experiments in history to produce net energy... it will prepare the way for the machine that can."
My take away from that, is that if both machines fulfill their goals, then building a power station that uses the technology would be the next step. I expect two things to be true; First, the 7-x will prove out its goals before ITER does, and second given that the stellerator design is significantly less complex than the tokamak design, it will be realized as a test power plant before the tokamak power plant is.
ITER is the big-scale, brute-force approach. People suspect there may be smart ways to do better with less, but until either succeed that's anyone's guess.
ITER is a tokamak. That's one approach. There is also the inertial confinement approach, that more efficient LED lasers have made an interesting contender. There is the stellarator approach, aka the "smart tokamak" experimented in Germany. Another one is the Z-pinch, used in Z machines, which seem to focus more on nuclear weapons research but that could potentially be a way towards usable fusion.
What makes ITER a likely candidate is solely its funding. I wish we funded the others to a similar amount.
And in each approach, there are also several ways of doing things.
It's questionable. ITER's design comes from the late 1990s and early 2000s.
Hight Temperature Superconductors (HTS) have changed the game. HTS tape has only been available for about a decade.
The cube of magnetic field is proportional to the energy gain in a Tokomak. See this video at at 46 minutes to get the equation, watch more to understand why people are now doing this.
Having worked on that tokomak, I can confirm that the Chinese are not playing around when it comes to funding long-term science initiatives. The US should be doing some serious soul searching about its priorities...where is the collective national vision of tomorrow?
Reading this article one thing that pops in my mind is: we have spent more money collectively to fund the LHC in CERN as we have funded ITER.
As much as extremely good science has came out of this experiment it boggles my mind that we have spent less in the search for the next breakthrough in energy generation, something that ALWAYS has catapulted technology and development of humanity.
The fusion effort is based on a set of assumptions that might have seemed reasonable in the 1950s when the effort got started, but are no longer realistic.
Fusion assumes the problem with nuclear power is that uranium is scarce, so the cost of nuclear energy would be dominated by fuel cost. In such a scenario, where reactors themselves are cheap, it makes sense to make a more complex expensive reactor to save on fuel. This was also the motivation for breeder reactors.
But this isn't how it worked out.
Fuel is a small part of the cost of nuclear fission power. The cost of building and maintaining the power plants, and disposing of them at the ends of their lives, turned out to dominate.
In that scenario, making larger, more complex, more expensive reactors just so one can burn deuterium and lithium is totally bass-ackwards.
The collapse of the fission breeder program should have been a sign that fusion wasn't going to work either. They were both based on the same incorrect idea of how things were going to go.
Except that fusion reactor have a bunch of advantages over all fission designs, including breeder and molten core Thorium, that you conveniently ignore.
The big one is: ramp up time is measured in minutes not days. This is a perfect addition to a grid that gets a large fraction of it's power from (somewhat) variable renewable energy.
The next big one is: There is no proliferation risk. No spent fuel that contains isotopes suitable for building fission bombs. And now legitimate reason for uranium enrichment to process your own fresh fuel. In other words there is no reason not to give this technology to North Korea or Iran.
The third important one: No radioactive waste that needs to be contained long term. The main "ash" is helium-4 which is not radioactive. The neutron flux will activate parts of the reactor vessel, but the halflifes are generally below a century. General rule of thumb is that things need to be contained 10 halflifes. Humanity has demonstrated successfully that we can build buildings that last 1000 years. This is a big contrast to the waste from fission plants that needs to be contained at least 100000 years, longer than we have had buildings.
I ignore these putative advantages because they are of minor importance, compared to the major downside of cost. If fusion is too expensive, all those other things do not matter.
Can someone help me understand the engineering and economics of distributing energy from a fusion plan?
Let's assume that in 2035 ITER reaches its goal and can sustain fusion at a cost of ~$25B per plant w/ $1B in yearly operational costs. Now what? Do we then build a nuclear-plant-style pressurized water turbine? How many turbines can we place around a fusion plant? How far can we distribute this power?
According to this[1] paper, fusion power becomes profitable at $175/Mwh. Nearly every other type of energy source, including wind and PV is cheaper[2]
I thought that one of the main benefits is that it is closer to zero emissions. Wind isn't always blowing, and the sun isn't always shining. Solar/wind take up a lot of surface area.
That is only true out till (a litle past) Mars orbit. For deep space mission solars performance in terms of Watts per weight quickly becomes pretty bad. The 1/r^2 drop in solar flux eats your lunch there.
Fission is fine out there. And DT fusion has no advantage over fission anywhere in space.
But I tell you what: let's let space programs pay for fusion, if that's going to be where it gets used. They all seem to have much more immediate tech development needs closer in.
I don't think the intent for ITER is to be an economically viable nuclear power plant, but rather to be a prototype, a proof-of-concept that will enable the next generation of smaller, more widespread, and above all cheaper energy-producing fusion reactors.
Fusion has advantages over Wind and PV; it works like a nuclear power plant. That means you can build a large plant where the power is needed and then provide it continously without stopping. Wind and PV can't do that unless you add buffer technology, but even that has limits.
Fusion can easily provide 100% of our baseline power needs and because it's powered by the most common elements in the solar system, it can easily be harvested (even from solar wind, which should deliver more than enough to power humanity with fairly little effort). Fusion is essentially the key to unlocking unlimited power (insert emperor palpatine here).
With only 100'000 tonnes of water, filtered for deuterium, the US could cover it's annual electricity needs entirely. The Mississipi provides this much water over 6 seconds.
60 seconds are enough to cover the entire world's annual energy needs.
Fusion has the disadvantages of nuclear. It will be very expensive. Nuclear is not competitive even if the nuclear island is FREE; the non-nuclear thermal power plant part is too expensive. Since fusion reactors themselves are unlikely to cost negative dollars, fusion power plants won't be competitive either.
I don't think that is true for baseline power provision, only for the peak demands during the day alternatives are viable.
For baseline, Fusion doesn't compete with PV and Wind, it competes with Batteries powered by PV and Wind, which is a whole different price calculation.
That plus the fuel is common and cheap to obtain, unlike Uranium or some of the rare earth metals for PV.
On a levelized basis, nuclear is grossly uncompetitive.
This means that when the wind is blowing, or the sun is shining, nuclear cannot sell for close to what it needs to make ends meet.
Cheap batteries are just going to be the final nail in the coffin. At the current rate of decline of battery prices I expect most operating nuclear reactors to shut down within a decade.
Oh, and please don't repeat the "PV needs rare earth metals" lie. They don't use rare earths, or for that matter need any rare elements at all. The only somewhat rare element silicon PV uses is silver for front contact wires, but that can be substituted for with copper (with a migration barrier layer to prevent reaction with silicon) if silver gets too expensive.
I don't think your expectations are realistic and underestimate the cost of baseline power. Baseline means this is how much the grid is pulling 24/7 without any pause.
You'd cycle batteries fairly fast if that was the only option which would increase the price, plus you'd now operate two powerplants; PV and battery.
Fusion and other thermal plants are more suited to providing baseline load. A fusion reactor that is running at a predictable minimum of 30% would cost far less than any PV/Wind/Battery system every could.
I think you're committing the common error of thinking that baseload demand requires baseload supply.
But baseload supply was just the most economical way to supply baseload demand. I emphasize was. We are moving into a new situation where intermittent sources are much cheaper than those old reliable baseload sources, when they are available. If this cost discrepancy becomes too large, the optimum mix of sources will change abruptly, to some combination of intermittent sources and various forms of storage.
The levelized cost advantage of renewables has become very large. Right now, they pair with dispatchable sources (gas, primary hydro), but storage is beginning to undercut that. Baseload nuclear is already grossly uncompetitive; in the US to compete with the current gas/renewable mix new nuclear would require a CO2 tax of $300/ton or higher. Even existing nuclear is struggling to meet its operating costs here.
142 comments
[ 2.7 ms ] story [ 248 ms ] thread"Zeta Hotter Than the Stars" "To Cheap to Meter " :-)
What the hell happened to optimism and trust in science? We used to be able to say "In 10 years there'll be a man on the moon" without having ever launched anyone in space. Now people doubt that in 50 years we will have a marginally better battery technology...
Fusion would be there already if only we had believed in it.
In comparison, the most recent DoD budget (in the USA) has ballooned to three-quarters of a trillion. It's clear we're not interested in the future.
Where exactly do you think the internet came from that you are using right now at this moment? Where do you think a multitude of inventions and technologies came from?
A huge chunk of that DoD budget is specifically for research and development.
It's bizarre it's like people learn just enough to solidify their world views then turn off.
That said, they do a lot of great research and some of it even stays unclassified.
Some of that money involves setting up assembly lines.
There are fundamental problems in achieving economically viable fusion. Nice and clean reactions that power the sun are way too slow, and need immense volume. Faster reactions produce neutrons and still require temperatures several times higher than inside Sun.
Where economy and safety are not the primary concerns, fusion has long been achieved: in bombs.
If ITER fails, what we will have learned about superconducting magnets manufacturing and high-temperature plasma will surely be useful.
But few great things have been accomplished without the risk of looking foolish, at least at the outset.
And because all the money is in selfie sharing apps, that’s where a large percentage of the brightest minds go to spend their career (as opposed to something less financially responsible, like nuclear fusion engineering, astronomy, or space flight)
And isn't that a bit of a dig at our astronomers and fusion engineers?
You cannot know what it will cost to invent something before it is invented. You cannot know how many deadends you will have to go down before you find one that works. In fact, it may be that the invention depends on knowledge or technology that we don't have, and maybe won't have for many many years.
For example, today fusion researchers are using computers to simulate plasma flows, gain a better understanding of plasma, and design better reactors... something that was not possible in the mid-70s when that chart was created.. and would not be possible to do by hand (as was done in the 70s). Hundreds of billions of dollars have gone into creating computers capable of doing that... where is that in your chart?
And to preempt this line: saying they could buy a supercomputer is not a rebuttal. A Cray-2 was the 2nd fastest computer in the world from 1985-1990 (10-15 years after your chart). It's the equivalent to an ipad 2 (a 2011 32-bit mobile low-power processor). I'm using a computer many times faster than that just to type this post.
The chart is not intended to be definitive proof that we would have had fusion power if we had spent the money (lv's last throwaway line notwithstanding.) No one knows what would have happened. To estimate that, we must use our prior based on how long/expensive other major tech developments have been. That's very different than pretending we know fusion's track record.
Incidentally, NASA is currently spending huge amounts of money on computer simulations for the equipment that will return humans to the moon half a century after they were last there. That doesn't mean those simulations are necessary to get to the Moon, as indeed they weren't possible in 1969.
[1] https://xkcd.com/1827/
I am expecting that if China announced a working fusion reaction tomorrow, US would put the funding and have one in only 5-6 years.
If I were a politician in 2019 with the right to sign funding and I see this graph, I cut all funding right away. Here's why:
* based on this graph, a simple extrapolation by only 2 years show the achievement of the commercial reactor in 2005. Where's the commercial reactor 14 years later? We are instead only 6.5 years away before pushing a certain button at ITER, which has some type of significance that's not entirely clear
* at least one point on this graph is "cooked": the JT-60U that appears to cross the breakeven point at some point prior to 2000. Here's what wikipedia [1] says about this:
"During deuterium (D–D fuel) plasma experiments in 1998, plasma conditions were achieved which would have achieved break-even—the point where the power produced by the fusion reactions equals the power supplied to operate the machine—if the D–D fuel were replaced with a 1:1 mix of deuterium and tritium (D–T fuel)."
What's the point of the "figure-of-merit" graph then? It projects when we'll get to breakeven or commercial, only to find there that you need to make another tiny change (i.e. substitute D with T) that will take 30 more years?
[1] https://en.wikipedia.org/wiki/JT-60#JT-60U_(Upgrade)
The triple product is the correct figure of merit and JT-60U did great there. But it was a plasma experiment and did not have the shielding that is necessary to run it with D-T fuel that produces a lot of neutron during fusion. But we know a lot more about neutron fielding than we know about magnetically confined fusion. Improving the plasma handling and the triple product is much harder than adding shielding.
Look. If you have a graph where you show a certain point called "figure-of-merit" being above a so called break-even point, every layperson in this world will assume that you get more energy out than you put in. If you then say that it's breakeven for a different reaction, they'll either question your sanity, or they'll accuse you of bait-and-switch. If you don't understand why you get your funding cut, you'll think life is unjust or cruel. Other people will nod sympathetically when you'll complain of the unfairness of this world , but secretly they'll disagree with you.
So, here's what Zubrin has to say about nuclear fusion and ITER [1]: "While, driven by international rivalry, the world's national fusion programs did advance forcefully between 1960 and 1990, the decision to consolidate all of them into a unified global effort to build the International Experiment Reactor (ITER) has caused nearly all progress to screech to a halt since the 1990s."
Why did this happen? Could it be because laypeople in charge of funding listened to scientists lobbying for a cool (and expensive) magaproject?
[1] https://www.amazon.com/dp/B07HDSSKHJ/ref=dp-kindle-redirect?...
( * ) in all fairness, you only argued that the figure-of-merit it a good metric, which I don't contest.
I am merely pointing out that fusion did not happen, not because it is inherently difficult or because it is a mirage, but because we did not fund it appropriately.
> where is that in your chart?
Way off the right. Fusion was supposed to have been achieved before Windows 95 was out.
But the scaling of tokamaks turned out to be much less favorable than that scenario assumed, and even if it had been favorable that effort was ignoring engineering considerations like limits on wall power density.
When that scenario was revealed to be a mirage, funding declined. Not sure why anyone would have expected otherwise.
The change you're observing is a response to the fact that we've implemented most of the low-hanging fruit. What we're left with are much more difficult problems.
http://www.ddprofusion.com/US10354761.pdf
Sort of like a fusion bomb, the fusion bomb doesn't produce radioactive fallout, but the fission bomb necessary to start the fusion bomb does.
Aneutronic fusion is possible, but requires collision energies much higher than what ITER is designed to create.
EDIT: I do wonder if there's a danger of nuclear proliferation with using fusion reactors to breed plutonium from natural uranium, though.
EDIT: see other comment, DT produces neutrons which can be bad, I still think the amount nastiness is small compared to say Chernobyl.
Tritium is something like 10 orders of magnitude more radioactive than uranium (in terms of its specific activity). Uranium itself is so low in radioactivity that the bigger risk when handling it is heavy metal poisoning. The by-products of fissions are much more radioactive, though.
If you set it off near the ground many of the escaping neutrons will interact with things that get very angry when activated by a stray neutron. A big bomb detonated near the ground, to crack open a bunker for instance, can send stupendous amounts of fallout into the air and kill people hundreds of miles down wind of acute "vomit now and die in a week" radiation poisoning.
Fusion is about hitting a precise balance point. It's why it's so hard.
The high temperatures necessary for fusion are only attainable at low masses, if too much fuel enters the reactor, it cools down and becomes inert. The fuel itself is not radioactive and safe for consumption in reasonable amounts (heavy water inhibits biological processes if it becomes too much in concentration, but unless you drink straight from the heavy water tanks for days, this is essentially not an issue)
Also when the plasma in a TOKAMAK escapes and hits the wall it strips off heavy metal ions which radiates all your beam energy away.
We need to take drastic action on power generation in the next decade. There isn't a snowball's chance in hell that fusion will make it in time. I can't over-emphasize how dire our current situation is - and that's with us currently living in a world where 3 cent/kwh solar power exists.
If we're gambling on fusion, at this point, the best-case outcome for it would be that it could be brought online sometime between us hitting 'drastic climate catastrophe' and 'civilization ending climate catastrophe'. To me, that doesn't sound like a great situation to be in.
http://www.helsinkitimes.fi/finland/finland-news/domestic/16...
https://arxiv.org/pdf/1907.00165.pdf
Wipe out 90% of the population and we're back to what, the 1800s? That level of death requires most people to just let the rising ocean drown them as they stand there plus a bunch more inland to walk in and drown themselves as well.
Massively distruption to the status quo is what climate change will bring. Civilization ending? Not even close.
But our civilization as we know it may very well collapse. Yes some sort of civilization will continue to exist. But this will be on par with the dark ages: not a complete destruction of human civilization but an enormous regression.
We will most likely loose a lot of our technological know-how and abilities. Events like this don't simply cause a gradual decline in standards of living. Rather they tend to cause massive outbreaks of wars, unrest, migrations, refugees, destruction etc.
Climate change in the bronze age caused nearly all major bronze age civilizations to collapse. But with this collapse we had major migration of people, wars, invasions, collapse of trade and many more things. What followed was large population declines as well as losing the ability to write and maintain organized rule, construct buildings etc.
Did all humans disappear? Of course not, that is why civilization reappeared again eventually at a later time. The same will likely happen with us. But how long will the dark ages last?
In many ways our resilience is weaker than theirs. Our trade networks are far more extensive and important than the ancient ones. What happens e.g. if the middle east, Africa and South America are turned into turmoil and we can no longer get resources extracted from either of those places?
Our economies may suffer a lot too, causing the breakout of widespread riots and unrest, overthrow of governments. If modern trade breaks down, there will be widespread famines even in the west, because many countries simply are not self sufficient. And if they are, their food production depends on a large number of imports. If they don't get hold of artificial fertilizer, fuel etc food production will collapse.
This is what we saw after the fall or Rome. Economies became far more regionalized/local. In a collapsing world economy, countries will become significantly poorer if forced to rely exclusively on local production of most essential things.
1. Except without access to easily accessible sources of oil, minerals, and energy.
2. If you think that rising sea levels are the problem [1] with climate change, I'm not sure you understand it, at all. This seriously undermines your points. The actual threat is disruption of agriculture, and ocean ecosystem collapse.
It’s one of those things that enable sending spacecraft to other stars. D-T fusion is still baby steps on that path, but worthwhile none the less.
Unlocking fusion, allows us to tap into a near infinite source of fuel. On a planetary scale.
What kind of dream world do you inhabit where you think it’s easier to travel to other stars or terraform planets than simply not “shit where you eat” on this planet? Slow down our economic growth? Stop dumping trash? Nah, that’s too hard. Let’s build fucking rockets to other solar systems instead and invent entirely new life support systems to survive in places without any food, water, or air at all. Sure, that’s way easier than planting a garden.
Even if this were possible, the technology almost certainly wouldn’t be ready by the time civilization fully collapsed at the end of the century.
A lot of people do see the concept of some entity strictly controlling the entire world as unattainable, regardless of whether it would be desirable or not.
Seriously.
So, yeah, travelling elsewhere is easier, because it requires less coordination. Space travel doesn't seem as rooted in the insolubility of political problems.
Not trying to convince you this is right, just that the opposite of your opinion is "obvious" to other people who think differently than you.
To not have space travel, everyone must agree not to do it.
To save the environment, everyone must agree to do it.
Anything that everyone must agree to do probably won't happen and seems "difficult" to me.
A political revolution against the corruption of our democracy by extractive industries is much, much easier than going to other planets. It might get a bit bloody or require you to march in the street, but it will sure as hell be easier and cost fewer lives in the long run.
It is cowardly to assume that our political situation can’t change; it’s merely a matter of what kind of action you’re willing to take.
I don't think any expansion steps looked less hard at the time.
One cloud hanging over all this is that ITER is going to use up most of the world's tritium. With fission in trouble, especially heavy water moderated fission, there will likely not be a continued source of this isotope, and making it intentionally would be extremely expensive ($100M/kg and up.)
If you compute the thermal power/volume ratio of the reactor, it's about 50 kW/m^3.
In comparison, the thermal power/volume ratio of a commercial PWR fission reactor is about 20 MW/m^3 -- 400 times higher.
A commercial tokamak would have a somewhat higher power, but not enormously higher. If based on ITER, the power density would be two orders of magnitude worse than a fission reactor. And since the complexity of a fusion reactor is far higher than a fission reactor core, the cost ratio of the reactors (for a given power output) would be even larger.
Since fission reactors are having desperate trouble being competitive today, how is making the reactor core orders of magnitude more expensive going to lead anywhere useful?
Stronger super-magnets, different designs could lead to smaller reactors.
As to why is it useful? A zero emissions powerplant, with minimal environment impact that is immune to core meltdown. Sure we don't have the tech to build it profitabily. Yet. But doesn't mean we never will.
Say we get commercial reactors in 2050. If these are anything like building fission reactors, we will start getting these commercial reactors online in 2060. That is very late.
Within that time we can massively expand solar power and wind power. As renewable energy start dominating the grid, ultra low spot prices on power will spur a huge cottage industry of businesses relying on buying this power cheap, storing it and selling it later. Alternatively factories will be built to take advantage of it. E.g. only running power hungry processes in periods of low power prices.
ATM fission, with its Chernobyl and Fukushima is still safer than solar and wind. And overall better for environment.
Fusion in theory promises even more - low environment impact plus no risk of disaster.
I don't see variability as a positive trait. From what I remember, renewables, really suck for the grid, as it wasn't designed for too much variability or diffusion in mind.
I don't think DEMO is ever getting built, btw. I suspect ITER is not going to have a very happy operational experience either (disruptions, maintainability). If ITER cannot effectively control disruptions they won't ever operate it with DT.
The idea that "renewables suck for the grid" is not correct. The very low levelized price of renewables means there's a lot of economic headroom to deal with intermittency.
Don't mistake current trends for future prognosis. Even if costs are falling at the moment, there is no reason future fusion technology will be museumware.
At worst we'll discover new data on plasma behavior. At best we have a possible tech for colonizing the Milky Way.
> I don't think DEMO is ever getting built,
It's ok. I don't think we'll transfer to renewables before it's too late anyway.
> The very low levelized price of renewables means there's a lot of economic headroom to deal with intermittency.
So you're saying that economics will change grid design? That grids designed for single way of transmitting energy will somehow reconfigure to allow consumers to stream energy to the source?
Is there any reason to think they will be correct now? I assert no: PV modules are still well away from the costs of the material and energy inputs.
I'll add that fusion economic studies have to posit large learning effects as well, so the "N-th of a Kind" plant is much cheaper than the first. And there, this assumption is dubious, since the closest technology (fission) has not shown good learning effects. That's why fission is failing now -- any technology that doesn't improve eventually gets killed as the competition does.
> It's ok. I don't think we'll transfer to renewables before it's too late anyway.
And you think the transition to fusion would be FASTER? My mind boggles at your double think.
> So you're saying that economics will change grid design? That grids designed for single way of transmitting energy will somehow reconfigure to allow consumers to stream energy to the source?
It's bizarre you think dealing with intermittency would require impossible, or even particularly difficult, changes. If you notice, grids are already becoming smarter, with power markets dispatching on short term prices. Storage and dispatchable demand fit nicely into this system. The market is going to force things in the right direction, which is renewables/storage and not high cost baseload sources.
No? I just think fusion tech is the future. A clean, abundant, uninterruptible, and independent source of energy that can be used on Earth or on Alpha Centauri.
I don't really expect to see them in my lifetime.
> It's bizarre you think dealing with intermittency would require impossible, or even particularly difficult, changes. If you notice, grids are already becoming smarter, with power markets dispatching on short term prices. The market is going to force things in the right direction, which is renewables/storage and not high cost baseload sources.
Again, that's not my experience. I've seen some countries move towards renewables, but nothing on the required scale. As for smart grids are you talking about decentralized grids, where consumers can send energy upstream or just more efficient energy grids?
And as for the markets, those depend on policies. If Trump says, we're making a renewables tax and tariffs on solar panels, those same markets could turn on renewables.
This is a form of a tired and invalid argument that one often sees. It's basically "nothing can ever happen for the first time." It also ignores the details of just what has been happening.
Renewables have declined in price very quickly. The timescale for this decline has been shorter than the lifespan of installed generating infrastructure. This is very unusual for power technology, so if one is not careful one may be misled, as you apparently have been. We're in a weird situation where that installed base is obsolete. It's still there, churning away, but only because its capital cost is written off. And some is being ripped out because it can't even earn its operating costs now.
What you want to look at is not this set of stranded assets, but the choices being made for new capacity. That is becoming overwhelmingly renewable. General Electric, for example, is in a world of hurt right now because the demand for gas turbines is collapsing. And nuclear fission or coal with steam turbines? Dead technologies.
But in a tokamak reactor, you have a large container with all the expensive engineered stuff placed around its surface, so cost should be roughly proportional to surface area, or volume^(2/3).
This complexity also means keeping that fusion reactor running would be a nightmare. Complex things break, and the fusion reactor will be so irradiated that hands on maintenance will be impossible.
I will add that fission power plants are uncompetitive even if you ignore the cost of the reactor itself.
That's a non sequitur. Your volume comparison can, at best, tell you how cost scales with reactor power. If the exponent is wrong, it serves no valid purpose at all.
Surely your apparent position, that something made of superconductors is going to be just as cheap as something made of steel, is much more supportable! /s
And I wasn't talking about an exponent there.
No. The statement "Fusion reactors are MUCH more complex than fission reactors." followed by "So, just comparing volume is being very favorable to fusion." is a non sequitur.
> And I wasn't talking about an exponent there.
Your compared power/volume ratio, implying that cost is proportional to volume (otherwise, the comparison is pointless).
I pointed out that the different operating principles of fission and fusion reactors suggest cost proportional to volume^(2/3) for the latter (which of course implies that there is a size beyond which a fusion reactor will be more cost efficient than a fission one).
The 2/3 is known as an exponent. I believe exponents are part of the K-12 math curriculum, and definitely propaedeutic to nuclear engineering.
You seem to be implying that comparing by volume is unfair to fusion. So, instead, let's compare by mass. The mass of ITER is 23,000 tonnes, for an average density of about 2.8 g/cm^3. I don't have the figure for the mass of a PWR reactor vessel in front of me, but it's mostly empty space (filled with water and fuel in operation, which are not part of the capital cost). So, comparing by volume alone may actually ALSO be giving fusion an advantage.
> I pointed out that the different operating principles of fission and fusion reactors suggest cost proportional to volume^(2/3) for the latter (which of course implies that there is a size beyond which a fusion reactor will be more cost efficient than a fission one).
I don't think scaling data for cost of fusion experiments supports this.
BTW, as a fusion reactor gets bigger, at a fixed geometry and magnetic field strength, the mass of the material needed to support the magnets (to resist the outward force the magnetic field puts on the coils) is proportional to the volume of the reactor, not to the surface area of the plasma. And you want the reactor to be at as high magnetic field as possible, since the fusion power density goes as B^4. If wall power loading prevents one from taking advantage of this the reactor is too big and will have inferior power density, compared to a smaller reactor at higher B. This is why concepts like MIT's ARC have power density an order of magnitude higher than ITER (but still very much less than a PWR).
> The 2/3 is known as an exponent. I believe exponents are part of the K-12 math curriculum, and definitely propaedeutic to nuclear engineering.
That's very special, I'm sure.
I am pointing out an obvious logical fallacy. If it really isn't obvious, try substituting "Tesla cars" for "fusion" and "bars of gold" for "fission reactors". Here's what you get:
"Tesla cars are MUCH more complex than bars of gold. So, just comparing volume is being very favorable to Tesla cars."
Which of course is utter nonsense. A Tesla Model 3 occupies roughly 10 m^3 (order of magnitude, mind you); that volume of gold would weigh about 19.3e4 kg, which at the current price of $45.6e3/kg would be worth about $8.8 billion.
Clearly, the MUCH more complex Tesla Model 3 costs quite a bit less than the same volume in gold.
> So, instead, let's compare by mass.
That would be equally pointless. Instead, how about taking a moment to look up a serious economic analysis? Here is a fairly recent one, open access and all:
https://www.sciencedirect.com/science/article/pii/S036054421...
It compares existing power plants with four tokamak conceptual designs which "span a range from relatively near-term concepts, based on limited technology and plasma physics extrapolations, to a more advanced conception". Upon reading it, you will be happy to learn that it finds Total Cost Of Electricity (TCOE) produced by those fusion reactors to be slightly higher than the lowest cost alternative, fission reactors.
All other options considered (combined-cycle gas turbines, open-cycle gas turbines, coal plants, solar large photovoltaic plants, onshore and offshore wind plants) turn out to be more expensive.
Think about the motivations for the people making estimates of the economics of fusion. They have little external motivation to be honest. The worst they could suffer if they were caught cooking the books was to lose their professional positions: but that is the same thing that would happen if fusion turns out to be seen as unaffordable and work is shut down. And really, a fusion power plant is so far in the future that they aren't likely to be caught on any but the most blatant dishonesty anytime soon. And who is being funded to check on their estimates?
So what might they do in that situation, even if they didn't flagrantly lie? They could tweak the assumptions going into the estimates, and keep tweaking them until they get results that are acceptable. If some source gives an input to their cost model that seems too good to be true, they might not question it too much. It's not "lying lying", they might tell themselves. They might even rationalize it's only a venial sin to bend things a bit, since fusion is so obviously a good thing to work on, right? Motivated reasoning can be a powerful thing.
All those failure modes are there, and there's incentive for them to occur, and there's little real world feedback to keep them in check. So you should take any estimates like the ones you linked to with more than a few grains of salt, especially when they fail a "smell test" by giving better economics than simpler, smaller fission.
In contrast, costs for renewables face the harsh test of actual experience, many times a year. Renewable cost estimates are, as a result, highly grounded in real evidence, and typically come in within 10% of the actual values. There is no room for lying -- to oneself or to others -- there.
The paper explains the methodology and lists references to the data. You can go through them and check the calculations yourself.
> Think about the motivations for the people making estimates
That's another logical fallacy, a classic one known as "ad hominem". Their motivations are utterly irrelevant to the question whether their results are correct.
If you think their work is in error, show how.
That's now how it works. If they put together a big, complicated cost model, it's up to them to provide evidence it's valid. Otherwise, it's just an exercise in "believe us, we're honest". For example, they could apply the same model to fission reactors and show it accurately predicts what they end up costing. Oddly, I don't recall fusion cost modelers ever doing that. Hell, I don't recall fission cost models being terribly accurate.
You like this result because it's telling you what you want to hear, not because it's intrisically believable. More motivated reasoning.
Yes it is. They presented their work, their sources and conclusions. If you believe it's wrong, but can not explain why, all you have is a prejudice. If you want anyone to take you seriously, you need to provide a rational argument explaining exactly how they are wrong, not just why you want to believe that they are.
"Then, it will remain to be seen whether Wendelstein 7-X can also fulfil its optimisation goals during continuous operation – the essential advantage of stellarators." ... "Although Wendelstein 7-X is not designed to generate energy, the device is intended to prove that stellarators are suitable for use in power stations."
And from the ITER Goals statement (https://www.iter.org/sci/Goals) "ITER will not capture the power it produces as electricity, but as the first of all fusion experiments in history to produce net energy... it will prepare the way for the machine that can."
My take away from that, is that if both machines fulfill their goals, then building a power station that uses the technology would be the next step. I expect two things to be true; First, the 7-x will prove out its goals before ITER does, and second given that the stellerator design is significantly less complex than the tokamak design, it will be realized as a test power plant before the tokamak power plant is.
ITER is the big-scale, brute-force approach. People suspect there may be smart ways to do better with less, but until either succeed that's anyone's guess.
ITER is a tokamak. That's one approach. There is also the inertial confinement approach, that more efficient LED lasers have made an interesting contender. There is the stellarator approach, aka the "smart tokamak" experimented in Germany. Another one is the Z-pinch, used in Z machines, which seem to focus more on nuclear weapons research but that could potentially be a way towards usable fusion.
What makes ITER a likely candidate is solely its funding. I wish we funded the others to a similar amount.
And in each approach, there are also several ways of doing things.
Hight Temperature Superconductors (HTS) have changed the game. HTS tape has only been available for about a decade.
The cube of magnetic field is proportional to the energy gain in a Tokomak. See this video at at 46 minutes to get the equation, watch more to understand why people are now doing this.
https://www.youtube.com/watch?v=L0KuAx1COEk
Tokomak Energy and Commonwealth Fusion Systems among others are looking at smaller reactors that use High Temperature Superconductors.
Also have a look at Helion Energy.
Also note that the Chinese are starting to put serious money into fusion research.
https://www.reuters.com/article/us-china-nuclearpower-fusion...
As much as extremely good science has came out of this experiment it boggles my mind that we have spent less in the search for the next breakthrough in energy generation, something that ALWAYS has catapulted technology and development of humanity.
Why?
Fusion assumes the problem with nuclear power is that uranium is scarce, so the cost of nuclear energy would be dominated by fuel cost. In such a scenario, where reactors themselves are cheap, it makes sense to make a more complex expensive reactor to save on fuel. This was also the motivation for breeder reactors.
But this isn't how it worked out.
Fuel is a small part of the cost of nuclear fission power. The cost of building and maintaining the power plants, and disposing of them at the ends of their lives, turned out to dominate.
In that scenario, making larger, more complex, more expensive reactors just so one can burn deuterium and lithium is totally bass-ackwards.
The collapse of the fission breeder program should have been a sign that fusion wasn't going to work either. They were both based on the same incorrect idea of how things were going to go.
The big one is: ramp up time is measured in minutes not days. This is a perfect addition to a grid that gets a large fraction of it's power from (somewhat) variable renewable energy.
The next big one is: There is no proliferation risk. No spent fuel that contains isotopes suitable for building fission bombs. And now legitimate reason for uranium enrichment to process your own fresh fuel. In other words there is no reason not to give this technology to North Korea or Iran.
The third important one: No radioactive waste that needs to be contained long term. The main "ash" is helium-4 which is not radioactive. The neutron flux will activate parts of the reactor vessel, but the halflifes are generally below a century. General rule of thumb is that things need to be contained 10 halflifes. Humanity has demonstrated successfully that we can build buildings that last 1000 years. This is a big contrast to the waste from fission plants that needs to be contained at least 100000 years, longer than we have had buildings.
Let's assume that in 2035 ITER reaches its goal and can sustain fusion at a cost of ~$25B per plant w/ $1B in yearly operational costs. Now what? Do we then build a nuclear-plant-style pressurized water turbine? How many turbines can we place around a fusion plant? How far can we distribute this power?
According to this[1] paper, fusion power becomes profitable at $175/Mwh. Nearly every other type of energy source, including wind and PV is cheaper[2]
[1] https://sci-hub.tw/https://www.sciencedirect.com/science/art...
[2] https://en.wikipedia.org/wiki/Cost_of_electricity_by_source
But I tell you what: let's let space programs pay for fusion, if that's going to be where it gets used. They all seem to have much more immediate tech development needs closer in.
Then there is a reasonable hope that we can widen the margins - the second is the DEMO reactor project.
Fusion can easily provide 100% of our baseline power needs and because it's powered by the most common elements in the solar system, it can easily be harvested (even from solar wind, which should deliver more than enough to power humanity with fairly little effort). Fusion is essentially the key to unlocking unlimited power (insert emperor palpatine here).
With only 100'000 tonnes of water, filtered for deuterium, the US could cover it's annual electricity needs entirely. The Mississipi provides this much water over 6 seconds.
60 seconds are enough to cover the entire world's annual energy needs.
For baseline, Fusion doesn't compete with PV and Wind, it competes with Batteries powered by PV and Wind, which is a whole different price calculation.
That plus the fuel is common and cheap to obtain, unlike Uranium or some of the rare earth metals for PV.
This means that when the wind is blowing, or the sun is shining, nuclear cannot sell for close to what it needs to make ends meet.
Cheap batteries are just going to be the final nail in the coffin. At the current rate of decline of battery prices I expect most operating nuclear reactors to shut down within a decade.
Oh, and please don't repeat the "PV needs rare earth metals" lie. They don't use rare earths, or for that matter need any rare elements at all. The only somewhat rare element silicon PV uses is silver for front contact wires, but that can be substituted for with copper (with a migration barrier layer to prevent reaction with silicon) if silver gets too expensive.
You'd cycle batteries fairly fast if that was the only option which would increase the price, plus you'd now operate two powerplants; PV and battery.
Fusion and other thermal plants are more suited to providing baseline load. A fusion reactor that is running at a predictable minimum of 30% would cost far less than any PV/Wind/Battery system every could.
But baseload supply was just the most economical way to supply baseload demand. I emphasize was. We are moving into a new situation where intermittent sources are much cheaper than those old reliable baseload sources, when they are available. If this cost discrepancy becomes too large, the optimum mix of sources will change abruptly, to some combination of intermittent sources and various forms of storage.
The levelized cost advantage of renewables has become very large. Right now, they pair with dispatchable sources (gas, primary hydro), but storage is beginning to undercut that. Baseload nuclear is already grossly uncompetitive; in the US to compete with the current gas/renewable mix new nuclear would require a CO2 tax of $300/ton or higher. Even existing nuclear is struggling to meet its operating costs here.