ITER: World's largest nuclear fusion project begins assembly | Hacker News Reader

ITER: World's largest nuclear fusion project begins assembly

bbc.co.uk

669 pointsgoodcanadian449 comments

I am going to need some help on this one as my knowledge on physics is limited: What makes scientists believe that the solution to nuclear fusion power is only a matter of scale? I desperately want nuclear fusion to work because nuclear power is the only realistic way to solve the looming energy crisis of the 21st century while still maintaining the same standard of living for everyone. It's also 500% more important that nuclear fusion works because most people around the world think that nuclear fission is scary (weaponization concerns even though fission power is different from bomb-making, radioactive waste, reactor meltdowns, etc.), even though it's the only realistic method we have today of going carbon-neutral.

However, I looked up the article on "fusion power" (https://en.wikipedia.org/wiki/Fusion_power) and it says "but to date, no design has produced more fusion power output than the electrical power input, defeating the purpose."

Can anyone help explain what I am missing, or what is not explained well? My common-person impression is if a laboratory experiment cannot even produce desired outcomes, what makes people think that an engineered, faulty-prone system will? The way I see it is that researchers produce the proof-of-concept, and engineering will attempt to reproduce that at scale. Isn't this preemptive? Or, from the article, it seems that it is necessary to build this thing in order to get any conclusive research results.

This is a lab experiment, just a huge one.

Scaling factor in fusion is pretty well understood science. Last JET tokamak was the experiment that confirmed this part.

Fusion science has evolved to the point where it is necessary to bring the fusion engineering on the same level.

The simple answer is that the surface only goes up a factor of 4 if you increase the radius by a factor of 2, but the volume goes up by a factor of 8. So the ratio of (energy containing) volume to surface (where energy get's lost) if getting better with increasing size.

There is of course much more complicated answers, all the way up to full-device simulations using gyrokinetic codes like Gene or Gkeyll that take millions of core hours and predict that ITER will work (and that predict performance of existing devices correctly).

Good answer. It's referred to as the Lawson criterion for those wishing a deeper dive:

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

Regarding simulations, plasma physics can in a very real sense be viewed as the predominant driver for supercomputing research and funding. Ratio of fusion budget allocations to total global energy demand remains woefully pitiful. On order of something like 30B / 100T or 0.03%.

This is the main issue and I don't want to make it any more complicated. But I do want to explain something because people hear arguments between fusion scientists. Within the field no one really doubts that ITER is going to work (there's a few, of course, but the majority don't doubt). The thing they are arguing over is if it is needed. The big reason for ITER is not to generate power but to demonstrate that sustainable fusion is achievable. The idea is that if there is a demonstration people will stop shying away from it and investing would skyrocket (a government would have to be dumb not to substantially invest in PROVEN fusion devices). The counter argument is that there is enough funding and that they will beat ITER to the punch with small scale reactors (much closer to the path for devices that would provide actual energy to our grid). My take is "don't put all your eggs in one basket."

Without giving a strong opinion either way on whether this particular project is a good investment or not :

Science, as an institution imo undervalues good marketing and good PR. For profit enterprises are under no illusion that they can succeed without good marketing and public perception.

If large expensive projects like this are needed to increase public awareness and support for basic scientific research then I am all for them. Society is willing to spend billions on projects that give no tangible return and are all perceptions (for example, sports stadiums) so if some of that instead is spent promoting Science that’s a good thing.

Any actual discoveries in fundamental science that come out would be a bonus too, but getting people excited about research is an end unto itself.

I think a problem with ITER for PR isn't the cost. It's time. I learned about it and got excited more than a decade ago. It'll still take years of time for it to be finished.

I'm not really sure how to express this, but the huge amount of time just diminishes the impact of it for me.

One could also argue that it would be developed faster if they had more money. It wouldn't be a linear relationship, but there is a definite relationship between the two. But fundamentally I agree. It is hard to stay excited about something to come for decades.

I have similar feelings about the James Webb Space Telescope too.

Think of it like the construction of a legendary cathedral. The prolonged commitment required is part of its impressiveness.

With a legendary cathedral, the layperson can walk up to it at any time and easily perceive its ever-waxing size and majesty.

This is something a fusion reactor doesn't have (I guess they could do tours or something maybe?).

They do have a YouTube channel showing construction.

There are people whose imagination ITER exceeds.

This is why I'm particularly upset about SSC's[0] failure to be built. There are many advantages to housing the smartest scientists in the world within the borders of your country, and those benefits aren't limited to scientific output. After all, a big part of these multi-national projects is the culture war.

And let's be real, which is worth more? A super collider that runs for decades or the F35 (which cost more, but we'll pretend they are the same and that America would be the only one funding a multi-national facility)?

Fun (perspective) fact: any of the top 30 billionaires in the world could personally pay for one of these colliders.

[0] https://en.wikipedia.org/wiki/Superconducting_Super_Collider

My software startup was a huge beneficiary of the SSC cancellation, it freed up a nuclear physicist to write graphics software for us, which became one of our biggest products.

I’m in favor of pursuing basic research, but there is always a cost. When you take a group of your smartest people and focus them on non-economic activities, the economy loses something.

That's incredibly short-sited. I'm glad for you that your startup was able to make money, but do you really think that the long-term economic value created is higher by making yet another piece of graphics software than from fundamental research?

I would argue this entirely the other way; whenever the smartest people are sucked away from foundational work to make money "up the stack", it makes us all the poorer.

Pie in the sky when you die....

https://www.comparably.com/companies/cern/salaries

Says the man living in a world plagued by short termism, where companies forsake their future to improve their quarterly report, where we have condemned generations to suffer climate change just so that we could have 15% cheaper electricity and enjoy oversized cars.

This going further down this path only lead to misery

Given that interest rates in developed countries are generally lower than in all of history, and mind-boggling amounts of surplus capital are being deployed by companies like Amazon, Softbank, Tesla, etc., it seems to me that the world is less "plagued by short termism" than any time in the last 13 billion years.

I'm not debating your assessment in an absolute sense, or the direction that resources are deployed, I'm just saying relatively speaking, the world has been approaching an asymptote, so how far does it have to go to be enough? At what point does some disaster that we weren't expecting happen that proves we went too far in being future oriented? Like, I dunno, an actual plague?

Edit: I meant this to express something more than just another "well actually" comment - I think COVID-19 really raises existential questions about whether people in the wealthy countries have gone too far in a future orientation, as if everything was certain and we and our civilization were all immortal. It may not make a bit of difference for some, but I observe it's changing a lot of people's outlooks. Something that has stuck in my mind is how plastic bags were banned in my area just before the pandemic hit, and whether or not there was an actual repeal, everyone has brought them back. Is this something to lament, to cheer, or just to observe and contemplate? If everything is bad, is anything?

I am struggling to make any sence of your post.

COVID demonstrates lack of being prepared and future orientated, we did not have PPE stockpiles and did not takw actions that would contain the virus to save the economy now. The more future oriented you are, the safer you are.

You're talking about predicting the future correctly and making the right tradeoffs. Being future oriented means trading off short term goals for long term ones; preparing for the future says nothing about whether we are prepared in hindsight.

You can use "future oriented" to mean being prescient or omniscient, but I think that's a useless way to define the word, because it's not a thing that exists.

The US invasion of Iraq was a drastic pivot of the US towards a more future oriented foreign policy, and I assume we would agree it was a huge mistake, right?

WeWork was a tremendously future oriented company, that wanted to be everything to everyone, but it wasn't, and won't be, the next Amazon.

It would be bad for society for similar projects to suck up all the resources in preference to short term needs.

"it is difficult to make predictions, particularly about the future" - Mark Twain

> The US invasion of Iraq was a drastic pivot of the US towards a more future oriented foreign policy

Which aspects of the invasion of Iraq were future oriented? Thinking it was "mission accomplished" in 2003? Having no exit strategy? Creating a power vaccum, which was eventually filled by Islamic State? Aquiring oil, in a world which is trying to divest from it, and which is being damaged by its use?

>Which aspects of the invasion of Iraq were future oriented?

The planning of it, which, you know, happened before "mission accomplished" and all the things that went wrong. It has been widely reported and asserted that the planning started considerably before 9/11, too.

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

> Formation 1997

That's 6 years before the invasion, so slightly longer than a single election cycle, and significantly shorter than the duration of the war itself. I wouldn't exactly call that 'future oriented', especially since 3 of those years were spent under the Clinton administration; Bush/Cheney/Rumsfeld pulled the trigger pretty quickly once in power.

It's also incredibly short-term compared to its contemporaries, like Russia's "Foundations of Geopolitics" or China's "Peaceful Rise".

I don't see the term "future oriented" as having any firm meaning in a vacuum, as a somehow absolute judgment; it's meaningful in a relative sense and that's how I used it.

After 9/11, most Americans became more future oriented than immediately before; both advocating for the invasion and addressing "root causes" in different ways are different forms of that. Either was more focused on the future than maintaining sanctions and occasional airstrikes.

> After 9/11, most Americans became more future oriented than immediately before

I really can't tell if you're trolling at this point.

Declaring war against an intangible form of political violence, as a knee-jerk reaction to a specific incident, which ends up perpetuating that form of violence, is a classic example of short-term thinking.

We have plans to invade everywhere, buildings full of plans.

“Future” oriented thinking would be Phase IV Occupation plans. That filing cabinet was empty (by design) and “It’s the UNs problem now” is not very future-oriented strategery.

Cool your boots...

The claim was that there is a trade off - which is a pretty common sense assertion... The comment you to which you are responding even asserted that basic research does have value - and does not assert that the trade offs go one way or another.

The implication of the comment as I read it was that we should at least critically think about those trade offs rather than just writing a blank check to whatever basic science scientists are requesting to pursue.

Your comment, in contrast, claims that the tradeoff always is worth it - that 'whenever' money is not diverted to basic research, then we are all poorer. You have no evidence for this claim.

In fact - no one really understands much about how best to evaluate this tradeoff. A helpful contribution to the discourse would provide us with tools to think about how to evaluate it, rather than just the endless tribalistic drum beating for one side or the other.

Funny, the comment from the start-uper, to me, sounds like a implicit explanation that basic/fundamental research is OK as long as it has less priority than economy (in the sense of money making money).

> The claim was that there is a trade off - which is a pretty common sense assertion...

It also happens to be wrong.

If more projects that require nuclear physicists get funded, more people can become nuclear physicists. We're nowhere near the point that we're going to run out of potential nuclear physicists who could be trained if there was more demand for them.

At some kind of abstract level there are a finite number of people in the world so anybody who does something can't be doing something else, but there are more than enough people for whom the "something else" is either unemployment or some net-annihilating occupation like divorce attorneys or advertising that we shouldn't have to worry about people spending their time doing something actually beneficial.

>> The claim was that there is a trade off - which is a pretty common sense assertion...

>> It also happens to be wrong.

Wow. The startup guy gave an actual example of the tradeoff. Am I correct that you see that not as a tradeoff but as a loss?

Sure, because the trade off wasn't real. There is a whole market full of people who could've done that work, it didn't have to the exact person who did it who happened to be one of the ones who would have gone to the other project, and the market expands when there is more demand because supply comes to meet it.

So the trade off wasn't between physics and software. We could've had both of those at the "cost" of fewer divorce lawyers or lower unemployment, neither of which are actually costs.

There is a theoretical point where having people do physics is less productive than having them do anything else they might've done, but we're nowhere near it. We still have people working in advertising only to cancel out the work of other people working in advertising.

Gee I sure wish I could have found one of those easily available graphic software geniuses, it would have saved us the $30,000 we spent getting ours a work visa.

Resource allocation is still a hard problem. We might be better as a society if we reallocate all the physicists to HR until that is solved.

Nobody said you get to pay below market rate. See how many appear if you offer a salary above it.

Or, that particular person could have been sucked up in to a huge bureaucracy and produced nearly nothing their entire career.

Hard to say.

I think you're over simplifying things with the example. There's 2 major factors that I see.

1) It is unknown what economic impact that the SSC would have had. It could produced more value or less than that which you and similar industries contributed to from the benefit of these scientists. A search shows that without accounting for the impact of the fundamental science, CERN's revenues exceed the costs by ~3bn euros[0] (report seems to also ignore ambiguous contributions like spin-offs). So the companies that benefited from the closure would have to at have near this revenue.

2) Scientific projects are fundamentally long term economic investments. These scientific projects have returns on investment of larger than 20 years and sometimes upwards of 100. These are impossible ROIs for companies, but great for civilizations.

These factors are huge and we're not even talking about the local economy boom because and influx of smart people which typically leads to higher quality schools in the area and tech startups that grow because subcontracting, consulting, and spin-offs. We're also ignoring the cultural impact as labs serve as a form of cultural exportation, which is important in the global game played between countries (the economic value of which is difficult to quantify), and this is what you are directly responding to.

It is fairly difficult to answer these questions, so I think your answer is too simplified.

[0] https://www.sciencedirect.com/science/article/pii/S004016251...

Regarding CERN, it's probably too obvious but I can't help but mention Tim Berners-Lee's work.

A nuclear physicist working on graphics software sounds like underemployment to me.

I agree.

"The economy" seems to favour the current generation of Einsteins working on optimizing ad impressions for Google, spreading fake news for Facebook and sucking unimaginable amounts of money into Bezos' pocket, the consequences be damned.

We could use a lot less of this, and a lot more Einsteins in research laboratories and universities.

At least some of the products in tech are actual products, I'm more sad and frustrated by a generation of Einsteins working in finance...

10-ish years ago I would have agreed with you. Right now, we are at a point where most tech giants are not making the world any better IMHO.

Do you think that a nuclear physicist writing graphics software rather than doing nuclear physics was a net win for society in the long term?

Well, depends on the nuclear physicist and on the graphics software

Maybe this person had better career successes in the SW field than in the Physics field

yes

Lets we free up all these physisics and engineers working on long-term projects that may or may not succed and send them to trade stocks on wallstreet. Who needs real progress

> the economy loses something

like…graphics software?

I think the trade-off is more than worth it to better understand the nature of our universe.

You seem to imply that basic research has no long-term economic benefit, or maybe that the long-term benefit does not outweigh the short term cost. That seems like a pretty difficult implication to defend.

No implication was given, just that things have costs.

Its pretty presumptuous to assume your startup has economic value, and the SSC didn't.

Plenty of completely economically worthless tech startups.

Well, ours was profitable and sold for $95M, kind of a definition of economic value.

That startup was already sold?

Now you have economic value it's time to have other kind of values. And with that amount of money you definitely can.

(of course, this is the comment of someone who painfully gathers 3000 bucks a month, that is who has 0 firepower except a few donations here and there, a few signature, etc. Not everybody is blessed with "the good idea").

I mean, that nuclear physicist also wasn't farming.

What you mention about billionaires paying for it piques me especially. For someone like Gates, Bezos, Zuckerberg or any number of other ultra rich individuals, the PR value of simply stating that they will themselves pay for such a collider up to the cost of its proposed budget (maybe so long as some organization or government promises to cover any cost overruns, or overruns beyond a certain percentage) would be enormous.

Not to mention that they would be paying for an immense and unambiguously good scientific gift to posterity that would be free of all the complications, interest conflicts, mismanagements and so forth that come with building a mega-charity like the Gates Foundation (the conspiracy theories around that alone have been absurd, despite all its excellent health programs for the developing world).

It's always hard for a person to say what they'd really do if they were in other shoes, but if I had the sort of fortune that a Bezos or Zuckerberg has and especially with the relative youth of these guys, it wouldn't be hard to convince myself to make such an investment.

The Large Hadron Collider cost somewhere in the neighborhood of $10 billion dollars. Bezos alone saw his net worth grow by a multiple of that in 2020 alone. This huge cost would barely dent anything else they're doing or funding and in no way dent their personal quality of life.

I'm not at all arguing that they owe something like this to the world, or trying to paint any kind of anti 1% resentment, just pointing out that paying for this doesn't even seem like such a bad idea for people like them and their image for posterity (something always important to most billionaires).

The top four or five billionaires in the world could even cover most of ITER's much larger budget without permanently or seriously denting their fortunes.

The discussion of billionaires building "CERNs" came out of a discussion I was having with a friend as a way to quantify how rich these people are in terms more understandable. Graphics like pixel wealth[0] just show scale, but doesn't put things in perspective. But a Forbes analysis estimates the cost of finding the Higgs at around $14bn (rounding up)[1]. I also tried using Super Computers, but at around $300m a piece the number was too small (e.g. Elon is 100 Auroras and Bezos is 1000, ball parks). Number of CERNs became a much more approachable metric but also has the fault that these take over a decade to build (meaning the # of CERNs is likely an underestimate).

I do agree that this would be a great PR move. Especially as we've seen Bezos grow his wealth by $65bn in the last 4 months[2] or Elon by $44bn (more than doubling) in the same time frame[3]. But the practicability is also a little naive considering that these numbers do not equate to liquid cash that they have available. Though someone like Gates has much higher liquidity than Bezos or Elon.

But it is interesting. Maybe if my metric catches on someone will build one (or another mega science project) :D

[0] https://mkorostoff.github.io/1-pixel-wealth/

[1] https://www.forbes.com/sites/alexknapp/2012/07/05/how-much-d...

[2] https://www.forbes.com/profile/jeff-bezos/?list=billionaires...

[3] https://www.forbes.com/profile/elon-musk/?list=billionaires#...

Regarding mkorostoff's 1 pixel wealth page. Taking wealth from the richest is very difficult. You'll have to find a way to make them agree with it. If you forcibly take money from people then you're just an angry mob. The only thing that would appear fair is higher taxes and maybe a wealth tax [0] but a 100 billion dollars buys you a lot of lobbying power. The worst case scenario is that the rich just hide their wealth.

[0] One could even make a concession and let the wealthy choose among a list of approved charities/projects to spend the taxed money on.

> You'll have to find a way to make them agree with it. If you forcibly take money from people then you're just an angry mob.

Not really. They use the "mob"'s roads and banks to make their business work and store their wealth. They use the "mob's" courts and police to protect themselves and their property. We live in a society, if your worry is that they are too powerful for them to participate in that society (e.g. because their power would let them hide their wealth away), then that just makes the ridiculousness of their wealth all the more apparent.

Wow. You think that the Gates Foundation has complications and conflicts of interest, and you think that a mega-science project would be free of those?

I mean, first of all, mega-science is just as prone to grift and mismanagement as any other type of mega-project, and far more so than vaccine programs or malaria treatment. And just imagine, what would the absurd conspiracy theory nuts say about a supercollider?

Of course, hey, wait a second, Gates and/or the Gates Foundation already do this (e.g. TeraPower, e.g. CFS), so your whole line of reasoning that it'd have considerable PR benefit is proven false. Fortunately the Gates Foundation isn't very motivated by PR.

You're right that something like a super collider could produce its own major opportunities for graft and corruption, but because it's a single project with a single construction to production timeline and a much more specific budget proposal, it doesn't quite compare to the endlessly fluid, highly organic and always adaptive nature of many long running billionaire foundations.

So no, I didn't quite prove the idea of its PR value false. Conspiracy nuts will always find a reason to consider anything made by some people or groups as suspect or outright nefarious, no matter how straightforward it is. but foundations oriented towards education, medical research and healthcare for the developing world are easily much more prone to that kind of narrative manipulation than a fixed, specific and extremely complex scientific device that will be used for a more limited range of possible things over a certain period of time.

The same applies to the differences in opportunities for graft and corruption between the two: Dishonest contractors and consultants could find ways to milk a privately funded super collider project during its construction and even during administration, but the scope of their opportunities would be fairly limited beyond a certain point, especially if the (presumably not stupid) billionaire funding it is keeping an eye on details.

Imagine on the other hand how many opportunities something like the Gates Foundation could offer for bad financial administration down the decades after its creator and main benefactor dies of old age or etc. Most of the collider's budget as covered by a billionaire in my scenario above will be a single sunk cost spent during a very limited period of time. The endowment of a giant foundation with flexible goals can on the other hand be managed (or mismanaged) ambiguously for decades.

The two things are not the same either administratively or philosophically.

> (the conspiracy theories around that alone have been absurd, despite all its excellent health programs for the developing world)

I mean, super science mega charity is going to have worse conspiracy theories. It's going to be the Gates foundation stuff mixed with the LHC stuff.

Like he's opening a portal to hell to mine energy from or something. No wait that's DOOM.

I can actually argue for the F35. F35 and Carrier groups ensures open oceans and secure maritime trade for all. 90% of global trade is over the sea and is worth Trillions of dollars.

There were times when countries would actively block maritime routes and steal cargo. A well integrated global economy, with open and secure maritime trade is beneficial for everyone and particularly the USA.

That would be done by frigates, destroyers and corvettes. Carrier groups are needed to dominate conflict theaters and other countries by projecting power.

Totally agree on open trade.

And as far as the F-35 is concerned, I think it was an attempt to be the end-all NATO, and global, fighter craft. Didn't work out that good, so far at least.

> Science, as an institution imo undervalues good marketing and good PR. For profit enterprises are under no illusion that they can succeed without good marketing and public perception.

Not to mention, this could be an attraction. Sell tickets for guided tours. I would gladly give money to have a guide walk me through that. Granted, at $100 a ticket, and 10,000 people a year, it'll take 20,000 years to recoup the investment.

Therefore, they need to make ITER romantic to have millions of people flock to rekindle an affection as powerful as the sun, and fuse humans as ITER fuses atoms.

You need to create a ritual. A couple would go, each holding a cup of water in their hand. They would then pour the water into the system as individual cups, and ITER would fuse atoms forever.

Now you can have millions of couples instead of two people like me.

Apparently, there are 80 million people per year who visit the frigging mall in Dubai. Many also visit the Eiffel tower, which was built for an exposition by the way.

Man, you’re just giving out ideas like this for free?

They also have to change the name. ITER --> GLITTER, to go with the fusion --> sun --> star theme.

Not all that GLITTERs is gold.

They could have a theme park with lots of different reactor designs!

Well, I did visit ITER back in 2014, it was really nice, we went on a tour, saw the facilities, attended a presentation, saw models of the reactors, etc. It was all for free (but the train ticket). We were a small (6) group of students who asked if a visit would be possible, and we were pleasantly surprised by the answer. There was at least one other family with us, so it's not just students.

Making money isn't really the goal for ITER, though. PR and education is. The tour left us with a pleasant feeling about ITER, but we weren't the ones to convince. I didn't see that kind of tour advertised in any brochure.

Might be better if, when trying to promote your passion, you don’t bash others.

I’d argue getting people excited about sport, and society as a whole playing more of it, is an unbounded good.

I even know some research scientists who play sport recreationally, and follow professional-level sport.

Plenty of room and resources for both!

I absolutely agree with you. Sports are good.

But it's an arbitrary good, because it's determined culturally. The rules of soccer could be changed at any time, and it would probably be just as exciting. Soccer is fun because we all agree it's fun, and no other intrinsic reason.

There's no reason we can't get excited about a monument to science the same way. If we have to spend billions just to get excited about science (without producing actual research), we should do it, just as we spend billions on sports just for fun and not because it produces any physical goods.

>>> small scale reactors

I believe this is the DoE position. Leverage AI/ML to design smaller magnetic confinement. With the goal toward commercialized products and partnerships.

JET was on track to breakeven. And I think most of the researchers involved believe they would have gotten there if the funding hadn't been cut. I think what they want is simply long term commitments. The sort of thing that is risky in fiscal budgets that can change year to year ;)

Final Report of the Committee on a Strategic Plan for U.S.Burning Plasma Research (2018)

http://brookhaventech.com/wp-content/uploads/2018/12/Burning...

You don't need AI, just use stronger magnets[1]. ITER was conceived before high temperature superconducting magnets were discovered, so it has more trouble maintaining plasma stability.

1: https://www.youtube.com/watch?v=KkpqA8yG9T4&t=39m0s

MIT-affiliated company building a reactor using these magnets: https://cfs.energy/technology

These reactor concepts are still very large for their power output, compared to fission reactors.

I’d compare it to the human genome project. Look where we are now with full genome sequencing for the cost of a laptop.

I had a conversation with nomeone who worked on ITER who was of the name opinion, and my follow-up question to him was: "how much money would I need to raise to get us over the line". His answer was calm and sincere: "I think, something like 3 trillion US dollars".

That's prohibitively expensive for private markets. No one will take that bet. It needs to be proven to be possible first.

> investing would skyrocket (a government would have to be dumb not to substantially invest in PROVEN fusion devices).

Even if it is proven to work that's no proof that it's economically sustainable. The enormous complexity and required scale are huge barriers to adoption.

Most of the advantages of fusion reactors are already enjoyed by fission reactors, and it's not like the fusion reactors are going to have that much of an easier time managing public perception. "They're building a H-Bomb in your back yard!"

Worse, since fusion reactors aren't useful for making bombs government investment is basically guaranteed to be tepid. It's only advantage is saving the environment, which doesn't get votes, and environmental groups will likely be lukewarm on the plants just like they are with fission. Sure it doesn't release CO2, but /nuclear waste/ is a huge boogeyman, even when you're talking about low level incidentals like irradiated gloves and bits of piping.

> Even if it is proven to work that's no proof that it's economically sustainable.

If it is proven to work, then it removes one of the key unknowns: Does it work?

If it can produce more power than it consumes, then it's ahead of the smaller demonstration reactors that have already been built and demonstrated.

> The enormous complexity and required scale are huge barriers to adoption.

Hence they're building this one.

> Most of the advantages of fusion reactors are already enjoyed by fission reactors,

Say what? What do you think the advantages of fusion reactors are, and same question for fission reactors?

> and it's not like the fusion reactors are going to have that much of an easier time managing public perception. "They're building a H-Bomb in your back yard!"

'This reactor can be built far enough away that even if it explodes - and it can't - you won't notice. You won't notice it going off, and you won't notice cancer and birth defects, vast tracts of land cordoned off, water sources rendered toxic, etc for the next few hundred years. And we don't have to bury the waste product somewhere for a millennia or more.'

Sounds like a much easier sell.

> Worse, since fusion reactors aren't useful for making bombs government investment is basically guaranteed to be tepid.

How many power stations are owned by governments around the western world today?

It feels like most of the interesting things happening in high tech now are all privately funded. If / when this works, and there's money in it, government lack of interest due to inability to make bombs from it (and that all sounds dubious to me) is irrelevant.

Indeed, we know that a fusion reactor will work if scaled up to about 1.4 x 10^27 cubic meters...

> Gkeyll contains ab-initio and novel implementations of a number of algorithms, and perhaps is unique in using a JIT compiled typeless dynamic language for its implementation.

LuaJIT. Neat. I assumed this stuff was still all in FORTRAN.

There's a good talk here about why ITER is so large: https://www.youtube.com/watch?v=L0KuAx1COEk

The idea is essentially that to get the parameters needed to make net energy with tokamaks, you need either very strong magnets or a large device. At the time of ITER's design, they used the strongest magnets they could find and then made the thing big enough to get the energy gain they wanted.

The speaker of this talk argues that it's size that stalled progress in tokamaks, since they'd become so big that building them became a massive, multinational project.

I wonder how much the possible magnet strength affects the design of a tokamak. There are very clear limits on the field strength you can achieve with classical superconductors, and I know e.g. in Nuclear Magnetic Resonance those limits had been almost hit maybe a decade ago or so. But very recently spectrometers based on new high-temperature superconductors have been delivered, so the technology seems to be far enough for actual production use now. I'm not sure how big the possible increase in field strength will be. Right now it's 23.5 Tesla for the largest NMR spectrometer with conventional superconductors compared to 28 Tesla for the new ones with high-temperature superconductors (actually it's a hybrid with high-temperature superconductors on the inside, and conventional ones on the outside), but I suspect there is more room with the new ones for higher fields.

It would be interesting whether the availability of better superconductors would change the design of a fusion reactor much, and allow significantly smaller ones.

Yes that has a big impact. The quality of a tokamak (as meassured by the triple product density, temperature, and confinement time) goes up like B^3 (IF I remember correctly, I do plasma physics but not fusion stuff) and the fussion power goes up like B^4 or something like that. So larger magnetic fields help a lot. But back when ITER was designed we did not have such strong superconductors yet.

PS: Google found DOI 10.1088/0029-5515/56/6/066003 but I didn't read it carefully.

Is there a successor to iter already being worked out on paper that takes into account new developments?

There is things such as SPARC by Commonwealth Fusion Systems (CFS) using rare-earth barium copper oxide (REBCO) superconductors. Not quite the scale of ITER, but it should give us some first experience working with that material. If that works out the next step would be ARC [1].

[1] https://en.wikipedia.org/wiki/ARC_fusion_reactor

Yes - the sequence is ITER, DEMO, then PROTO:

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

https://en.wikipedia.org/wiki/PROTO_(fusion_reactor)

Yes, the video above is basically about this, and is really worth watching -- it's a fantastic talk.

I'm pretty sure that Commonwealth Fusion Systems[0] is the entity affiliated with MIT that has been doing work to prove out the recently available higher magnetic field strength superconducting materials and apply them to tokamak construction to bring size down dramatically. They had a bunch of press in 2018/2019 when they first got underway[1], and it looks like they've received a lot more investment over the last few years and likely made quite some progress since then[2].

0 - https://cfs.energy/ 1 - https://www.bostonglobe.com/opinion/2018/03/09/new-approach-... (op-ed in the Boston Globe by a Vice President at MIT) 2 - https://cfs.energy/press/

Cambridge Fusion Systems[1] is a private company spun out of MIT that is building a proof-of-concept reactor based on these new magnets within the next 5 years.

Breakthrough in Nuclear Fusion? - Prof. Dennis Whyte (2016) - https://www.youtube.com/watch?v=KkpqA8yG9T4

Timeline (in case you want to skip over some parts):

  00:01:00 - introducing Dennis Whyte, MIT department head for nuclear science
  00:04:24 - presentation starts
  00:06:00 - identifies breakthrough with REBCO magnets
  00:07:25 - explains deuterium-tritium fusion
  00:12:30 - basic metrics for reactor performance
  00:17:15 - energy output of other previous fusion experiments
  00:19:00 - examines ITER and the problems of its approach
  00:22:00 - problems solved by high energy magnetic fields
  00:28:15 - full scale reactor concept, teardown of REBCO magnets
  00:37:00 - design limits and margins
  00:39:00 - fixes plasma instabilities found in weaker magnetic chambers
  00:40:00 - maintainability, lifespan, component replacement
  00:45:00 - solution to neutron damage and energy capture
  00:50:30 - cost and profitability
  00:54:00 - full graph of field strength vs reactor scale (and thus funding requirements)
  01:01:50 - Q&A
  01:30:00 - question about the biggest risks

He gave another talk in 2019 with more numbers and even more confidence: https://www.youtube.com/watch?v=rY6U4wB-oYM

1: https://cfs.energy/technology

And then you go to the arxiv paper for ARC and learn: the power density is 40x worse than a PWR primary reactor vessel, and their projected cost is $29/W(e) (vs. < $1/W(e) for PV). Also, the vacuum vessel likely wouldn't survive a disruption (although maybe they've fixed that in the years since).

Compact high field tokamaks have better power density that ITER or DEMO would, but they still are very inferior to fission reactors. And fission reactors are far out of the running economically.

^ That is a _fantastic_ talk for someone who know's some physics get a quick grasp of the state of the field, the different approaches being taken from a set of first principles, and how to evaluate them.

Also check out this UK company which is instead going down the path of using stronger magnets: https://www.youtube.com/watch?v=hWYnbYsp3i8

I didn't watch the talk, maybe it's the same one, but I recall watching a talk about Tokomaks that said thanks to the development of high-temperature superconducting wire ITER is already an obsolete design.

Awesome talk. Thank you for the link!

>looming energy crisis

Where do you see the crisis? Unless you want to provide everybody with an abundance of energy, people need food and shelter, some electricity for electronics and some transportation once in a while.

Of course, any amount of energy can be burned for simulations and bitcoins but that's simply limited by supply. There will be a crisis when bitcoins and simulations are more valuable than human lives, but that's not changed by offering more energy.

>people need food and shelter, some electricity for electronics and some transportation

Modern shelter, food and transportation means a lot of energy.

Modern shelter means operating energy for heating or air conditioning, plus embedded energy used to create its components. If the shelter is green, or "energy efficient", it means it is operating energy efficient, but it contains a huge amount of embedded energy in the insulation and heat exchange components.

Air conditioning doesn't necessarily have a worse bottom line, energy wise, than just heating. It's all about the time * temperature difference envelope. Air conditioning usually has a much lower temperature difference between the outside and the inside. If it's 40 C (as hot as it gets in places where masses of people live) outside, air conditioning has to drop by 15 C to get the inside to 25. If it's freezing (0 C) outside, heating has to heat by 25 C.

Double that for a modern workplace - office, warehouse or manufacturing, they all consume a lot of energy to keep the workers comfortable.

Modern high-speed transportation means a lot of energy too. We move crash-resistant 1500 kg structures at highway speeds around daily.

However, we need less energy by the year to maintain an identical standard of living, because we are making everything far more efficient. Our energy use has been extremely inefficient up until now. Modern heat pumps alone will revolutionize heating, requiring 2-3x less energy for heating.

Moving around massive cars by burning gasoline is a perfect example of a hugely inefficient and primitive system. Mere electrification requires many times less energy.

The US depends on massive amounts of energy because we outlaw city construction that enables car free living, but that will begin to change too if younger generations can ever wrest control away from the boomers.

This is an odd jump you are making. Most energy isn't used just for food, shelter and some transportation, and certainly not for "bitcoin and simulations".

By far most energy is used to produce things (including the transportation requirements to do so). And as it turns out, the hunger for producing and buying new things is very hard to temper. This hunger is evident from the fact that the ways we are measuring "economic success" strongly correlates with the amount of things we produce, or even with the speed at which we are increasing the amount of things we produce.

However, will it lead to a crisis? The hunger for new things simply dies down if they are not affordable. It's already happening. Energy is much more expensive than it used to be. Resources are more expensive so that new products have tighter margins and waste less materials.

People buy as much as they can. If they get less energy for their money, they buy less. Everybody had had-made clothes and ate organic food. Those times passed with hardly anybody complaining about receiving less.

What on earth are you talking about, shortage of energy would be a total ecobomic collapse, not just a crisis.

6 months of people staying at home are causing economic crisis.

Large cities literally depend on cheap energy to survive, to bring in food and move away the trash. If there is no power for a week, London and every other megacity turns into a mass graveyard

It's a crisis because it is happening abruptly. The price for hydrocarbons will rise much slower, like the increased consumption of the growing population of the world. Meanwhile, more renewable energy will be installed.

There will be energy in the future, even if there is not the abundance of nuclear fusion. So I don't see why there is a looming crisis.

> People buy as much as they can.

That reminds me of something I heard early on in the pandemic.

"The economy is collapsing because people are only buying the essentials".

> nuclear power is the only realistic way to solve the looming energy crisis of the 21st century while still maintaining the same standard of living for everyone

I assume by "energy crisis" you mean "how do we continue to generate energy while drastically decreasing climate effects"? Because if we're not worried about climate there's no looming energy crisis. There's plenty of oil, gas, and coal out there still.

And so if we're talking about climate stuff... is nuclear really the only way? Renewables continue to get cheaper and scale up; grid storage with batteries and pumped hydro is already a thing, and plenty of other storage methods are in development, it seems unlikely not a single one of them will turn out to be useful?

Plus, carbon capture for fossil fuel electrical generation is possible, just expensive, and not all fossil fuels are equally bad for the climate, so possibly carbon capture + cleaner fossils could be part of the future, too?

I think people have been saying "it has to be nuclear, so it has to be fusion" since way before we had the alternative sources (and storage) we now have, and are continuing to develop. Is it just a trope now, that maybe should be revisited?

(I find the possibility of fusion power really exciting and interesting. I'm just not sure it's necessary)

> There's plenty of oil, gas, and coal out there still.

I was under the impression we only had ~100 years worth of oil left, just that its not talked about because climate is more pressing and normal ecconomic pressures will fix the problem if we run out of supply. Is that wrong?

It's likely that we have more than that now thanks to newer methods of oil extraction like fracking. That said, hopefully within my lifetime the primary use of oil will be for plastics rather than power.

I think typically those predictions mean there's ~100 years of oil that's extractable with current prices and technologies, not that there won't be any oil left at all in 100 years.

It seems to me that it’s a question of long term energy needs, i.e. when civilization goes to the next level and wants to send laser-propelled light-sail equipped probes to other planets with 10% speed of light.

Ok, but that's not the question here :) The question is whether the idea that nuclear (and fusion in particular) is the only way to have climate-friendly energy generation and "avert a looming crisis" is outdated, and maybe no longer stands up when today's technology is taken into account :)

As I understand it renewables and batteries have their own environmental impacts that can't be ignored. When comparing nuclear with renewables the impacts start to show nuclear as at least a good contender as the future power source. But large scale fusion might drastically lower energy costs. That could enable things like carbon capture.

"Can't be ignored" meaning what precisely? This is a phrase that sounds like pure FUD, and despite spending a lot of time trying to quantify where this supposed environmental damage comes from, I haven't been able to find anything of substance.

Similarly, I don't think there's much reason to believe that fission nor fusion will be able to compete with renewables on cost. Even if fusion is a completely free source of heat, you still need to convert heat to electricity with a steam turbine and it won't be long before renewables are cheaper than a steam turbine and cooling infrastructure on a free heat source.

People have this strange faith in fission and fusion as being cheap, at some point in the future, but nobody can ever explain why it will be cheap. What is the mechanism that could drive this along?

Solar is the current cheap electricity that will drive carbon capture. There are even startups that have plans for using atmospheric carbon capture to generate synthetic fuels, that plan to be profitable with the current cost curves of solar and the rest of the industrial process.

The world is a very different place than 20 years ago when it comes to technology, and I think it's time to re-evaluate the potential promise of fusion as an energy source. I don't think it has much promise for terrestrial power, but as always, I would love to have my skepticism conquered.

Can't be ignored means we have not found a non destructive means of energy production yet. Waste from current renewables will build up. We still need to be on the look out for better technologies for the future. That is not to say they are not a huge leap forward but it does have to be considered.

I don't understand this. What problem do you foresee happening?

You may as well say that we can't build any more houses, or make more clothes, because of the waste. If the rate becomes a problem, in some way, we will recycle the necessary parts. But unlike, say, coal ash, the waste is easy to handle, and easy to repurpose if we find the need to.

> won't be long before renewables are cheaper than a steam turbine and cooling infrastructure on a free heat source.

That's an absolutely extraordinary claim. How could a wind turbine ever be cheaper than a traditional generator and turbine, since it consists of a generator + rotor + massive tower in a remote location with a very low utilisation rate?

You know you cannot just assume a trend will continue infinitely, especially if the conclusion is so non-sensical.

The steam turbine has to endure super heated steam, requiring very specific materials. Steam turbines also require boilers, cooling, etc. Wind turbines have the potential to require far less in terms of the input materials. And with solar, who knows how efficient we can get with that!

We have been optimizing steam turbines for more than a century, but are just barely getting started on optimization for wind and solar. So far, the rate of improvement hasn't start to slow at all, so I'm fairly confident that due to greater simplicity, wind and solar will end up being less costly than steam turbines. Predicting the future is tricky business, but in the 90s it would have been foolish to think that semiconductors wouldn't get better for more than just a few years. And we are in a very similar place with wind and solar techs now as we were with semiconductors in the the 90s.

Mist of this post is badly wrong, wind turbines are essentially a close cousin of steam turbine, and are massively more expensive per Mw of power. A single steam turbine can generate a GW and will weigh less than a wind turbine that generates a MW, a thousand time less!

We run steam turbines as high pressure because that gives us efficiency, not because we have to.

"he rate of improvement hasn't start to slow at all, so i am confident"

You have no evidence, so you are confident?

Also airdinamics are well understood, there are no miraculous effficiencies coming to wind turbines

You don't even address some of the core points: a GW of steam turbine is going to require massive cooling infrastructure that wind turbines do not require.

The learning rate will not stop overnight with wind turbines; there is more than enough research and industrial improvements that are in the works that we will start to see a slowing of the learning rate before it suddenly stops.

Also, airdynamics and not well understood as it relates to wind farms, and new research comes out all the time to improve efficiency of future designs.

For somebody who's claiming that the other person is not providing any evidence, you are operating without any links, only with certainty that I must be wrong about predictions that I admit are tricky to make.

The real issue with nuclear is not the waste, but the time it takes to build the reactors. We don't have enough time left before we irrevocably fuck up the climate. Solar and wind can be brought online much quicker.

Methanization is currently expected to be the critical storage component for CO2 neutral energy systems. A major driver of storage need in many places is not renewable fluctuations but seasonal fluctuations: You need more heating in winter/AC in summer. So make Methane and store the gas. Technologically this is all solved.

New opportunities to leak methane and we've seen it's often cheaper to ignore leaks, given the cost of it is incurred by the planet.

If you put the methanation plants next to the gas power plants you don't need leaky gas infrastructure. And methane leaks still seem easier to deal with than nuclear waste.

Which fossil fuels are "less bad" for the climate? They're all very bad.

Obviously old coal powerplants are the worst, but there's more evidence every year that the allegedly "clean" natural gas is nowhere as clean as advertised once you account for fugitive methane emissions that in practice are largely untracked. And flooding the market with cheap fossil fuel like methane as a "transition fuel" – another marketing gimmick – will only delay the transition to renewables for obvious economic reasons.

I'm not optimistic about big shots like fusion, but small modular nuclear reactors for example could be operating relatively soon if only we had the desire to go in that direction.

> Obviously old coal powerplants are the worst, but there's more evidence every year that the allegedly "clean" natural gas is nowhere as clean as advertised once you account for fugitive methane emissions that in practice are largely untracked. And flooding the market with cheap fossil fuel like methane as a "transition fuel" – another marketing gimmick – will only delay the transition to renewables for obvious economic reasons.

None of this makes nuclear, and fusion in particular, necessary in order to avoid a "looming energy crisis"? That's what we're talking about. If we wanted to have better emission tracking and/or carbon capture for natural gas we could probably do it, I'd guess, and it's still not clear natural gas generation is required at all in the face of improving renewables and storage.

You know, what's not clear at all is that renewables will ever be cheaper than fossil fuels outside of a minority of markets. You're putting a lot of weight on that prediction but despite all the recent progress on costs, if that was going to happen anytime soon, climate change would largely solve itself. Wouldn't that be nice.

Small modular nuclear on the other hand could be economically feasible in places that don't have the money to pay extra for renewables but still want to drop their GHG emissions.

I'm not claiming that renewables have to completely replace fossil fuels (see my reference to carbon capture), or that they even need to be cheaper (fossil fuels could become artificially expensive by, eg, carbon taxes). I'm questioning the assertion that nuclear is essential, which is what was claimed by the comment I replied to.

Nuclear might be nice to have, might be better for various reasons, might be lots of things. But is it absolutely required "to solve the looming energy crisis of the 21st century while still maintaining the same standard of living for everyone"? That's my question.

>will only delay the transition to renewables for obvious economic reasons.

If you want more renewables then you need a more flexible grid. All that crappy coal baseload clogging the grid needs to disappear and make space for renewables. Gas plants are cheap to build but expensive to operate. They will be mostly used when there is a temporary shortfall of renewables. They are also necessary for power to gas if you actually want to reach 100% renewables.

The fundamental difference between nuclear and any other source is energy density. It boggles ones mind that your can replace burning 14 thousand tonns of coal _every day_ with something that needs to be refueled once a year.

Sure, but energy density (which is essentially hypothetical in the case of fusion, because we don't have fusion power) has got very little to do with the assertion that nuclear energy (and fusion in particular) is the only way to avert a looming energy crisis?

nuclear energy density is derived from e=mc2, so fusion/fission are in the same ballpark.

Current accepted way is to build solar/wind and redesign grid to deal with intermittency. Solar and wind are even more energy diluted than fossil fuel, they take 500 times more space, more land out of nature/alternative uses.

Nuclear has already demonstrated that it can decarbonise industrial economy up to 80%, see France. and they did it just in 10 years.

This is bascally the reason i think this is the way out of crisis

All for nuclear, but the amount of land used is higely exadurated. It is actually totally trivial if you run the numbers.

In reality >90% of worlds valuable land use is farming. Factories, cities, houses, etc. all hardly matter.

Well, one example. I live in Singapore, and for it to become energy independent, it would need a square kilometer for a nuclear plant. To achieve the same with solar, you need to pave the whole country with solar panels.

Besides, to make these solar panels you need to spend energy too. and EROI for advanced solar is about 20, where for nuclear it is 80+.

I think people should stop using nature when they can get the same/better results without using it/using several orders of magnitude less of it.

I view renewables as energy farming, and city-states like Singapoure, lixemburg, etc. Are not very suitable for either form of farming :) for most countries the calculus is very different. UK has been developing sea based wind turbines quite succesfully, thats something that could work for singapore.

I view nuclear as a bit of a lost cause to be honest, they take long time to build, cost a lot upfront, you need a lot of expertise around them, and they would require a miracle of PR. It does not appear likely that countries that do not have an established nuclear industry will suddenly become nuclear powered. Thats just the way things appear to be heading, and its a shame.

Besides, don't underestimate the infrastructure needed to refine fuel, manufacture fuel rods, deal with waste, etc. If you outsource all of that, you are not terribly energy independent.

I think the only hope for nuclear industry is small modular reactors, and that's only if a massive amount investment comes through.

What I meant is ofcourse de-carbonise economy, Singapore can't be energy independent in meaningful sense as it does not have fuel sources of any kind. And solar/wind do not count as there could totally be still nights.

I frankly dont understand how an energy source that can shift from 100% to 5% capacity on its own whim, independent of its user's needs could get so much attention and considered viable replacement for something predictable..

There was not a single economy decarbonized on solar\wind, even for those that tried really, really hard (see Germany). And we have example where it worked, in short 10 years.

There are countries still that can build nuclear on time and on budget - see South korea for example. Wonder why they don't get all the orders for urgent decarbonisation needs.

> And so if we're talking about climate stuff... is nuclear really the only way? Renewables continue to get cheaper and scale up; grid storage with batteries and pumped hydro is already a thing, and plenty of other storage methods are in development, it seems unlikely not a single one of them will turn out to be useful?

Maybe one of them will work out. Do you want to bet the planet on that?

> Plus, carbon capture for fossil fuel electrical generation is possible, just expensive, and not all fossil fuels are equally bad for the climate, so possibly carbon capture + cleaner fossils could be part of the future, too?

Again, maybe.

Yes, there are a bunch of other things that might work out. But fusion is the possibility that's closest to proven. Surely it's worth trying? If it turns out we figure out a storage mechanism that's good enough to make solar viable (say), great. But let's not abandon one of our most promising approaches until we're sure.

To add to your list of currently available proven approaches, would an approach that decarbonized France's energy in just 10 years should also be considered?

It looks like proven track record of decarbonisation if there is one.

You're talking about fission? I think we should push it as far as it will go. At the same time I think there is some legitimacy to the waste and disaster concerns (even if mostly overblown) and so fusion offers enough potential advantages to still be worth pursuing.

I am all for pursuing fusion, for pushing borders of our knowledge.

At the same time, concerns are way overblown about fission. You know, it is regulated to 1/10000 amount of radioactive increase that is known to ever cause harm.

This is causes the designs to be order of magnitude more expensive then they could be. For example, for BWR, fukushima, there is no graphite inside, and reaction is going on only when water is present. There is no reason to have super expensive air-tight building around it.

Moreover, in fukushima case, the fact that it was airtight actually caused it to explode, as hydrogen formed inside it. Should there be ventilation, there would not be chance for it to explode and disperse material further than otherwise.

Secondly, unneeded evacuation actually caused ~1k deaths when they were removing patiends from life support equipment, for example.

As for the waste - it is again purely political problem, you can totally reprocess it, as they do in France and they end up with a few slabs of glass for 30 years of powering Paris. Even if you dont reprocess it first, it takes ridiculously little space, and never harmed anyone.

Compare coal - 14000 tons of coal daily. Just imagine that, and compare this to the fact that it is possible to store _all_ nuclear waste right _on site_, ready for politicians to come to their senses. :)

To look at it this way, most dangerous nuclear plant is the one that does not get built, as it will be for significant part will be replaced with something burining dinosurs.

> To look at it this way, most dangerous nuclear plant is the one that does not get built, as it will be for significant part will be replaced with something burining dinosurs.

Entirely true. But unfortunately there are a lot of nuclear plants not getting built at the moment. Tragic as it is, developing working fusion is probably an easier way to change that than political persuasion.

I guess we can agree to disagree here. What will probably happen is that Russia and China continue to build up nuclear, and export it to whoever will take it.

even the most active proponents of solar/wind will continue to lag way behind in decarbonisation, as the bigger percentage of renuables on the grid the more difficult it is to manage.

at some point optics of the situation will catch up, and political persuasion will happen this way.. in my opinion it will happen much faster than cheap fusion

It's simply not true that nuclear is the only way. Renewables plus methanization can definitely do it at a price not much above what we currently pay. It's not clear at all that nuclear is cheaper, though it would require less restructuring of infrastructure.

I'm not familiar with how you are using methanization in this context. Are you referring to deliberately pumping methane into the atmosphere?

You generate hydrogen using excess renewables, then turn that into methane and store it to burn in conventional gas power plants. Look up various versions of Power2Gas. It's not economical right now because we have few times when renewable production exceeds demand but is a natural component of a fully renewable power grid.

Or, just store the hydrogen.

Hydrogen is really hard to store, though.

EDIT: should have qualified my statement with "in man-made objects"

No, actually it's pretty easy, in underground caverns (cheapest is solution mined cavities in salt; cost per kWh of capacity about 1/200th the per-energy cost of batteries.)

Interesting

Storing methane is cheaper, and you can use it for heating, it brings down costs when used in addition to hydrogen storage:

https://arxiv.org/pdf/1801.05290.pdf

" 4.7. Methanation scenario - In the Methanation scenario the conversion of hydrogen to methane is allowed, which can then be fed into the natural gas network for use both in the heating and electricity sectors. Since the carbon dioxide required for the methanation is captured from the air, the methanation has a low overall efficiency (60%), but the resulting methane is extremely valuable to meet the peak heating demand.Despite the costs of the methanation equipment, total system costs reduce by 11% compared to the Heating scenario. In the heating sector, a substitution of heat pumps with gas heating can be observed in Figure 9. Significantly reduced CO2pricesand average marginal prices for electricity and heating are also seen in Table 3. Furthermore, the benefit of transmission reinforcement is weakened, since the methanation allows the use of cheap gas storage to smooth synoptic and seasonal variations of renewables. Optimal transmission reduces the total systems costs by only 17%, compared to 25% in the Heating scenario,and the optimal transmission volume is also lower. The total volume of synthetic methane produced with no transmission is 708 TWhth, compared to 795 TWhth from natural gas. With optimal transmission the volume of synthetic methane reduces to 263 TWhth as transmission smoothes more synoptic variations of wind."

I'm also a fusion power fanboy but

>nuclear power is the only realistic way to solve the looming energy crisis of the 21st century while still maintaining the same standard of living for everyone

This is just not true, there's no way you can say this. Solar costs are going down massively. Hydro is dirt cheap already. So renewables can absolutely be part of an energetic transition in the near future, while fusion is at best many decades away. So while I think fusion energy has the potential to transform energy generation, and by extent everything about our life, it's wrong to assume renewables aren't probably our safest bet in the near future.

Also there's something amusing about "Here's my sure assessment. Anyway I checked the wiki page on fusion power and".

>This is just not true, there's no way you can say this. Solar costs are going down massively. Hydro is dirt cheap already. So renewables can absolutely be part of an energetic transition in the near future, while fusion is at best many decades away. So while I think fusion energy has the potential to transform energy generation, and by extent everything about our life, it's wrong to assume renewables aren't probably our safest bet in the near future.

We need nuclear to meet baseload because the storage requirements (for PV particularly, by far the largest renewable) would be absurd without it. But since peak load can be several times average and nuclear plants can't be spun up in a day, we also need storage, and with storage around renewables can be cheaper than nuclear. It's not an either-or question; both should be used.

Climate agreements have largely avoided the thorny question of providing nuclear power to the developing world, but if they're to achieve a prosperous standard of living in a sustainable future based on foreseeable technology, this has to be addressed.

> We need nuclear to meet baseload because the storage requirements (for PV particularly, by far the largest renewable) would be absurd without it.

The problem with statements like this is that when they're proven wrong (they aren't always, but when they are), it's often because of some massive underlying shift that makes a bunch of assumptions wrong, leading to a wrong prediction.

If we envision a system where not only delivery, but storage is centrally managed, then yes, there's a massive amount of energy storage required, and that's hard to justify and invest in for large companies.

If instead you assume that maybe electric cars will act like large battery reservoirs, and stuff like the powerwall will also be used to supplement it, then we end up with a massive amount of battery storage already distributed to different endpoints, an d paid for by individuals instead of a few massive companies.

Whether that's all that likely, or even possible at a huge scale because of required rare materials is a question, but that's an entirely different type of scenario than "energy companies invest in massive batteries to leverage solar/wind for efficiently", and the type of thing that's hard to predict and because of that often overlooked. That doesn't mean stuff like that doesn't happen all the time. In fact, I would say there's a major shift like that every decade or so, we just don't necessarily notice them unless we look at them.

The internet itself was a major thing. Just relating to the internet, there have been major advanced every few years. The rollout of new major advancements is unevenly distributed and often over the span of a decade, leading to it being hard to notice them. Just this last year, the massive increase in remote work will likely cause a major shift in the economics of many markets, and change how many predictions would play out.

Bringing this back to energy, consider that it seems like every year California is having massive fires, and is bankrupting its public utility provider to the point that the state is prepared to take it over if it gets much worse. At that point, if the state decides it needs to actually replace a lot of infrastructure that PG&E needs to maintain, maybe pushing for some partially distributed model starts to make sense.

These are all things that go into making predictions about energy really hard, since we're at an inflection point where a lot of stuff that used to work is not working very well, and new technologies are just at the cusp of being useful.

*useful and financially competitive

Important distinction here: nobody actually cares or wants power that is baseload. In some cases it can actually be a bad thing.

The property you really want is “dispatchable”. There when you want it. Not when you don’t.

Yes sure you can break it down like that. At larger scales (e.g. interstate transmission lines) you have to aggregate and then becomes baseload + surge again. Individual users want dispatchable, cities or counties becomes baseload +.

>nobody actually cares or wants power that is baseload.

That doesn't make sense. At all hours of the day there is some demand on the grid:

https://www.eia.gov/todayinenergy/detail.php?id=42915

If you only have to store for the fluctuations vs. having to store all night you end up with massively less storage.

Electric cars won't do it either. There are less than 300 million registered vehicles in the US and a car battery holds about 100 kWh. That's 30 billion kWh and overnight demand may be as high as 5 billion kWh/night. You'd need an extremely high compliance rate to pull that off.

But seasonal productivity fluctuations are the bigger issue with solar and wind. You might get way more power in July than you need and way less in January -- wind is also seasonal, but the peak month varies by region. You're not going to store months and months worth of electricity in cars and even grid storage facilities would become cost-prohibitive, and you can't reset a nuclear reactor on a weekly basis, but one or two starts a year might be achievable if you design for it. Currently that's not legal:

http://ansnuclearcafe.org/2013/09/03/why-dont-we-mothball-nu...

https://model.energy/

Go there and model (with real historical weather data) how much a solar/wind/battery/hydrogen system would cost to deliver steady power. Then compare against new nuclear. Sorry, nuclear.

(The hydrogen part is essential in some places, like northern Europe, and its impact is not fully appreciated by many nuclear fans.)

> We need nuclear to meet baseload because the storage requirements (for PV particularly, by far the largest renewable) would be absurd without it.

This is not true. With a properly designed solar/wind/battery/hydrogen system, a steady stream of power can be delivered more cheaply than what you could get from a new nuclear reactor.

This wasn't true even ten years ago, but it's true now, and many (such as yourself) have not updated your priors.

Hydro is extremely restricted because of geographical and environmental constraints. Solar is also quite constrained by lack of large-scale storage and environmental concerns, and just by the weather. Same goes for wind.

Keeping in mind that electricity production has to increase significantly to absorb the shift from ICE to EV vehicles I would tend to agree that there is no way to meet demand without nuclear even if renewables are, and should be, pushed.

Being part of does not mean being sufficient.

I will challenge your assertion that energy storage is some kind of “deterministic barrier” that needs to be crossed before decarbonizing our grid.

Solar is still negligible in terms of penetration in many places in the world (even though it and wind are the cheapest new build primary energy today, so expect this to rapidly change). As well, you can add quite a bit of solar to a grid before curtailments become necessary.

Add that solar and wind pair together (when it’s not sunny, it’s often windy and visa versa). Many grids today (Uk, Denmark, Germany, etc) can have decent penetration of renewables with 0 energy storage.

Add in large scale grid interconnectivity (sunny in Nevada, windy in Idaho, Hydro in pacific north west) it will always be sunny and windy somewhere.

> I will challenge your assertion that energy storage is some kind of “deterministic barrier” that needs to be crossed before decarbonizing our grid.

This is clearly not my assertion.

My bet is that large scale storage is a much, much easier problem to solve than power-generating nuclear fusion.

Yes, but that does not change anything to the point of the importance of nuclear (fission) power.

That depends on how quickly you think the storage problem can be solved. If it can be solved in 10 years then it makes little sense to build more nuclear (fission) capacity, because it will be obsolete by the time it is built.

Storage is only part of the problem, though a big one. The area that needs to be covered with solar panels and windmills is another one.

> The area that needs to be covered with solar panels and windmills is another one.

I don't really see that as a huge problem. There's loads of free area for solar panels on rooftops. Sure, it's more expensive to build solar there. But that's not even an engineering problem, it's simply a matter of allocating funds to it. And most of the cost is labour. Which is arguably a bit of a bonus in a world where we're constantly worrying about there not being enough jobs.

Hydro is actually pretty bad in terms of big safety incidents (Banqiao: 171000 dead, Machchu: 5000 dead, South Fork: 2208 dead) and solar/wind have trouble beating nuclear on metrics like death/watt because you need lots of infrastructure per watt. Picture a few large cities with contractors running around every roof tending to solar panels and compare to a few experts at a single nuclear power plant. For the contractors-on-roofs, the slip & falls add up even though they'll never get an HBO mini-series.

Anyway, I tend to agree that going forward solar + storage is probably workable. The storage part isn't proven yet but I have faith we'll figure it out. There are lots of promising options under investigation and the proven fallbacks aren't that horrendously expensive, all things considered.

It's just a pity we stopped building nuclear 40 years ago because it was viable all the way back then. Heck, we got to 20% nuclear! Compare to 2% solar today. If we had merely continued building nuclear at the same pace instead of stopping in the 1980s our grid would be 100% low-CO2 today instead of maybe 30 years from now if we hurry. But that didn't happen. We made the super-mature and responsible decision to fill our atmosphere with CO2 instead and now we get to live with that decision. So it goes.

Hydroelectric projects also cause an increase in mercury levels specifically methylmercury. You'd think hydroelectric would be a safe method of generating power with flooding being the only bad effect. An Inuk guy in Labrador I follow on Twitter (@AndersenAngus) is trying to raise awareness about methylmercury due to flooding of land for hydroelectric dams.

Here is some info on it via the Harvard School of Engineering and Applied Sciences: https://www.seas.harvard.edu/news/2016/11/human-health-risks...

"Microbes convert naturally occurring mercury in soils into potent methylmercury when land is flooded, such as when dams are built for hydroelectric projects. The methylmercury moves into the water and animals, magnifying as it moves up the food chain. This makes the toxin especially dangerous for indigenous communities living near hydroelectric projects because they tend to have diets rich in local fish, birds and marine mammals such as seals. "

We are just so bad grasping the entire chain of effects of our actions, and the environment is incredibly complex. We need massive automated testing of environmental impact; systems that surveil ground, air, biodiversity over time all around the globe and particularly near projects like dams. It's no longer possible to just shrug it off, hope the planet can tolerate it and subconsciously accept some human loss in the name of progress.

> Picture a few large cities with contractors running around every roof tending to solar panels and compare to a few experts at a single nuclear power plant. For the contractors-on-roofs, the slip & falls add up even though they'll never get an HBO mini-series.

Most solar power will not be distributed power, but large-scale solar power plants. All large-scale solar power plants I've read about are on the ground, which makes sense since, unlike with wind power, increasing the height of the panels on a solar power plant gains nothing.

> increasing the height of the panels on a solar power plant gains nothing

Well, not nothing, since less atmosphere means less energy dissipated before light hits the panels, but certainly it’s a better tradeoff cost-wise to keep them low.

It gains the space underneath. Flat roofs (of large commercial buildings, for instance) sound like a good place for them. I don't understand why they are put on the ground instead so often.

Yeah, I keep coming back to this "$1 billion idea" of building a solar plant on big shed roofs. Quite literally buy a reasonable sized property in the country(with/near rail) and cover it in sheds.

Once the solar energy output is established then the shed can be used for productive work. Preferably with some level of energy storage but a lot of productive sheds don't need much overnight electricity.

Sounds like a pretty good idea. You could run it as effectively two separate business, one is a power utility and the other rents storage or workshop space.

Because it's difficult to install them at scale like that. For example, if you want to do a big solar project, you have to negotiate with the owner of every building you want to put panels on, you have to find a place in every building for inverters, you have to install metering, you may have to upgrade the grid to handle the extra power going out, then you have to have an account and customer service and payments for every building, and if the panels need to be maintained, you have to schedule a time and send out a crew.

Then you also can't practically do certain things like install trackers, because they break more often and it's too expensive to maintain them if you have to have a crew go out there instead of your on-site maintenance people just go over to the broken tracker/panel.

Overall the costs of rooftop solar are just more expensive than a central solar farm. Plus, you can site the solar farm where there is better sunlight and land is cheap.

Ultimately this is the problem I have with solar: It puts an effective cap on our energy use. Solar will never be more than 100% efficient, and there is only so much land that it is reasonable to use, so if we say, wanted to expand our worldwide energy usage 10x, it's not really feasible.

> Ultimately this is the problem I have with solar: It puts an effective cap on our energy use. Solar will never be more than 100% efficient, and there is only so much land that it is reasonable to use, so if we say, wanted to expand our worldwide energy usage 10x, it's not really feasible.

Isn't this exactly why we should be installing solar on rooftops. It might be more exensive, but it's a bettet use of space. Plus most of the extra cost is labour, and aren't we constantly worrying about how their aren't enough jobs. Seems like a win-win to me.

It does not out any kind of cao on our energy usage, the amount of land that needs to be covered in solar panels to supply a country is a tiny percentage of it's area, and is basically a rounding error.

The limiting factor is cost of instalation and eqioment, not land. Additionally there are mnay deserts which are totally uninhabbited and uninhabbitable.

Most people live in cities where land is very expensive, but rural land is cheap. So, it makes sense to do large-scale solar projects that aren't near cities at ground level, and put solar panels on roofs in cities and towns where the power is being used on-site and land is expensive.

If people are putting installing solar panels at ground level in expensive urban settings, then yeah I think that's pretty weird. (It might be an artifact of land use planning rules in some towns where people put solar panels in places where they aren't allowed to put buildings or parking lots, for whatever reason.)

In my country (the Netherlands), there are a lot of huge distribution centers, large warehouses with flat roofs. But instead of putting solar on the roofs of those, rural land is used, and we don't have so much of that. It's a shame.

Hopefully more shared use becomes a thing. Stick the panels a bit higher, space them out a bit more. No reason why you can't have a mix of livestock and solar panels on one piece of land. Raises the price a bit for now, but everyone wins I guess?

This is a very flawed and short-sighted argument. Averages don't matter when you are talking about fat-tailed/power law risk distributions. Nobody would be opposed to PVs on roofs in their neighborhood b/c some construction workers fall to death every year - this risk is well calculable. But (almost) everyone would be opposed to a fission plant or nuclear waste facility next door - and rightfully so.

Without enormous direct and indirect subsidies, nuclear (fission) isnt commercially viable anywhere in the world. Heck, you still can't insure a fission plant.

Yes, in theory fission would have been the best option for carbon-free energy. No, in practice humanity never figured out how to safely and efficiently use this power source and now renewables are a way safer and cheaper bet. You won't find any objective economic analysis (that incorporates such indirect subsidies as the implicit state guarantee and realistic building and waste handling/storing costs) that can show otherwise.

> But (almost) everyone would be opposed to a fission plant or nuclear waste facility next door - and rightfully so.

Reading this kind of debate from France is a good laugh.

If you're interested you can check where our electricity comes from here: https://www.rte-france.com/eco2mix/la-production-delectricit... (between 60 and 70 percent of nuclear energy today).

Because France was desperate to have its own nuclear weapons.

It was not economic for them either (tho back in the day it displaced coal - try having a coal plant next door)

What is the point of bringing up economics when we are talking about life and death? How can you even measure if something is economic or not if you haven't defined what a human life is worth?

France's nuclear power program actually displaced oil. France had a lot of oil fired electric power plants in the 1960s when it was a cheap fuel. As a reaction to the oil price shocks of the 1970s, France committed to the "Messmer Plan" for nuclear electricity:

https://en.wikipedia.org/wiki/Nuclear_power_in_France#Messme...

France demonstrated its first fission bomb in 1960 and its first thermonuclear bomb in 1968:

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

The 1970s and 1980s build up of nuclear power in France had nothing to do with desperation to have its own nuclear weapons. France already had them before the Messmer Plan.

"France's nuclear power program actually displaced oil"

No. Oil is much cheaper

To maintain a nuclear arsenal a nuclear industry is needed.

It is the only economic justification for nuclear power

So you don't think something can be expensive because it is heavily regulated? Or do you just think it is not possible with less regulation?

Also see electricitymap.org

And then check out the cost overruns on Flamanville 3. What is now, 4x the initial cost promise? Or is it 6x?

New nuclear is dead in France just like it is in the US.

France exported and imported a vast amounts of electricity during it’s nuclear hay day and still needed huge subsidies. In effect they where little different than people living near a local nuclear reactor getting most of their power from it while extremely dependent on the wider electric grid to meet demand and lower overall prices. Which as odd as it sounds demonstrated civilian nuclear inability to scale as they still needed to reduce power plant utilization by ~10%.

I agree that costs are a major problem with nuclear. It seems to be a problem with all large scale construction projects in the past 20-30 years or so.

On the safety tail risks, IMO it's more of a psychological/perception problem than a problem of actual risk. We've had a number of serious accidents in the history of nuclear power, and none of them have led to anything close to the death toll of a single year of running coal plants.

The problem is actually that people are forced off of their homes, personal property and way if life should there be a big accident. For an average householder it's their life savings, friendships, community and way of life wiped out through no fault of theirs. Sure there'll be compensation and resettlement but who trusts governments to do this fairly and painlessly?

It hasn't helped that no civilian nuclear operator has demonstrated the ability to sustain safe operations with zero incidents and consistently prioritize safety over the course of decades. Every operator has a string of nuclear incidents of varying severity and i think more than anything, the Fukushima thing shook people up because while you could excuse away Russian and American accidents with cultural factors nobody perceives Japanese to be careless or irresponsible (and this was expressly the reasoning that led Angela Merkel to reverse her stance)

So yeah in abstract at a nation level, it's not a major risk but for the families possibly affected, it's a catastrophe - as opposed to increased chance of a few people getting cancer over their lifetime from Coal.

Therefore given the asymmetric risk and demonstrated inability of the nuclear industry in ensuring zero incidents, there's naturally grassroots opposition which translates into political pressure that no amount of "risk is so low" data-flashing can wave away.

> and this was expressly the reasoning that led Angela Merkel to reverse her stance

I know a lot of Germans who think that reversal was just opportunistic pandering to environmentalists to grab votes from the Green party.

Averages don't matter when you are talking about fat-tailed/power law risk distributions. Nobody would be opposed to PVs on roofs in their neighborhood b/c some construction workers fall to death every year - this risk is well calculable. But (almost) everyone would be opposed to a fission plant or nuclear waste facility next door - and rightfully so.

How is this even an argument.

Nuclear risks are also calculable. The regulatory hurdles nuclear has to go trough, even for testing reactors, to meet risk criteria (among others), are enormous. The vast majority of nuclear, as exposed by deaths/TWh generated, is extremely safe. Flying vs driving argument.

Take the worst nuclear catastrophe. Take the wildest overestimation in deaths. It's still less than most other sources, including renewables.

No, in practice humanity never figured out how to safely and efficiently use this power source

What? France. All their active residues probably fit in one or two football fields. 3th/4th Gen will consume them, making medical radioisotopes in the process. You can't contain megatons of CO2 this easily.

In the meantime, air pollution kills in the range of 6-8 million people a year. German support for keeping nuclear grows

https://twitter.com/hh_mikawe/status/1284770806090825728?s=2...

What makes you think nuclear power has a power law risk distribution? For that to be true, it'd have to be possible for an unboundedly large nuclear accident..

Just because something has more concentrated risk doesn't make it fat tail. Nuclear power's winning safety record already includes Chernobyl and Fukushima.

Renewables are only safer and cheaper bet once the battery issue is solved and can be made equally green, safe and cheap. The primary reason why nuclear isn't commercial viable today is that combining natural gas and renewables makes for a very cheap energy grid.

Nope.

There are other alternatives, and grid scale batteries are viable now. Just cheaper to burn gas/coal and fsck the future!

The obvious alternative is pumped hydro, that is well established. Another is demand management, unexplored because the greed heads in control do not believe that our society can work together for the common good.

Coming from Aotearoa, as I do, I know that is a fat lie.

I have seen Aoraki with my own eyes, and I understand what you seek to protect.

For the locals Aoraki is a literal ancestor.

So it is not my job, it is taken care of.

I suppose I was unclear. Please forgive me. We need to protect the idea of preservation in culture and society, so that future people also hold these ideals of caretaking and awareness of our ancestors, including the planet and life itself, to be true.

Another alternative is to just build so many renewable energy plants that demand can be served even when production is comparatively low. During peak production hours, you can just blow the energy back into the atmosphere (or run a steel mill with it). This is a question of political will, and nothing more.

Currently there does not exist a commercial viable pumped hydro installation in the world that buy overcapacity renewable energy and later sells it at a profit. There are a few experimental and proof of concept installations that demonstrate the possibility of hydro as a battery, and at least one operate as an additional reservoir to a hydro dam, but the commercial viability of buying cheap renewable energy to pump water in order to generate electricity later when the price is high is not yet here.

Demand management is also a great idea in concept but the commercial viability is questionable as long as people can just turn on the cheap gas burners. Commercially, turning off production and keep the industry closed when renewable are not producing is much more expensive compared to just paying what ever price society currently demands when burning fossil fuels.

Remove fossil fuels from being viable choice in the energy grid and the economic viability would change dramatically for every other energy source, including hydro pumps and nuclear. Demand management might even become a possible strategy to a point where it can have an significant impact on the energy grid. For now most grids operates by combining cheap renewable with cheap fossil fuels, with the environment taking the real cost when the fossil fuels burns.

There is no "battery issue". There are lots of energy storage mechanisms that work well in combination. [0]

What you said is a tautology and brought nothing new to the table. (non cheap) nuclear can't compete against cheap renewables. Everyone knows that but almost nobody knows why nuclear is so expensive today.

The reason why nuclear power is not commercially viable is that it doesn't benefit from economies of scale. If you build custom tailored humongous monolithic nuclear plants you're going to pay a huge amount of money. Just think about how expensive it would be to build one giant solar panel with a total area of 1 km² instead of a million 1m² panels. It's absolutely nonsensical yet it happens every single time a nuclear plant is being constructed. The few success stories like France simply standardized on a single design. General corruption and budget bloat probably did more to stop nuclear than all anti nuclear hipsters combined.

[0] https://cleantechnica.com/2020/02/09/correcting-anti-renewab...

Hydro's main problem is that it's just not available in most countries / areas.

Solar's main problem is that it's not reliable everywhere. Try setting up solar in Chengdu and you'll find that there's just not enough cloud free days in the year to make it worthwhile.

We need baseline power and renewables are not usually enough. If you're lucky enough to have a river or sunny climate then great, but that's not possible in some places.

Is it windy?

In Chengdu? No wind on most days, which is (in part) why the fog / smog blocks out the sun almost every day of the year.

> "solar/wind have trouble beating nuclear on metrics like death/watt because you need lots of infrastructure per watt."

That may be so. But solar and wind are clear winners when it comes to cost/watt, and that's what determines which gets built.

If the energy source is not providing energy when i need it, i dont care that its cheap at some other time when its of no use to me.

Safety is not a real problema in rebewavles, they are much safe than mining/ fossil fuels of any kind.

Nuclear is the safest energy source in terms of deaths per terawatt hour [1]

[1] https://ourworldindata.org/safest-sources-of-energy#:~:text=....

South Fork was never a hydroelectric dam. It was an earthen dam meant to create a lake for the robber barons of the day.

And it was fine, until said robber barons decided not to maintain the dam.

> solar/wind have trouble beating nuclear on metrics like death/watt because you need lots of infrastructure per watt.

Really, this makes me think one thing, jobs. We can take steps to make jobs safer, but if solar/wind get even close to nuclear but employ a lot more people, then that's a huge gain overall.

Reducing personnel costs is a gain for a company's bottom line. Increasing personnel costs is good for society, as it means either more employed or higher wages (assuming it's not higher wages that somehow results in fewer people). Decreasing company/product costs with a new technology that also does so while employing significantly more people is a huge win for everyone (except those that refused to diversify from the old technology).

Edit: whoops s/jobs cheaper/jobs safer/

>Increasing personnel costs is good for society

This is your brain on capitalism. :p

More seriously: more jobs, more human effort, more accidents and injuries and death, less free time... this is objectively a bad thing. It's only from the lens of the current economic system that it becomes a positive thing, which speaks volumes in itself.

Well, part of the problem is that I said "we can take steps to make jobs cheaper" when I meant "we can take steps to make jobs safer".

We keep automating away jobs. People want/need work, mostly for money, but also because want to feel they are doing something and part of something. So I don't take it as a given that less free time and more human effort is objectively a bad thing. It really depends on the person and whether they feel a sense of accomplishment in their work.

As for death and accidents, that's somewhat addressed by my typo fix.

> It's only from the lens of the current economic system that it becomes a positive thing, which speaks volumes in itself.

The thing about the current economic system is that it's the current economic system. It can change, in small ways and big ways, but I'm not not sure it will (even if we're probably in the absolute best time to try out UBI we'll see in our lifetimes, and it will be a shame if/when it passes us by in that respect, even if it means life is much better overall).

Hydro also directly saves lives by reducing flooding and indirectly via clean water. On net world wide hydroelectric dams have saved more lives directly than they have cost from dam failures. Which is unique among all energy sources. For context: https://en.wikipedia.org/wiki/List_of_deadliest_floods

As to nuclear, it’s horribly expensive when you try to scale it. France was regularly exporting and importing vast amounts of electricity to other countries and their power plant utilization was still 10% below the US etc. That directly equated to significantly higher prices.

> olar/wind have trouble beating nuclear on metrics like death/watt because you need lots of infrastructure per watt.

There were some studies claiming nuclear has lower deaths/kWh, but as I recall they (1) used old numbers for wind that do not reflect current safety figures, (2) assumed solar was rooftop instead of utility-scale which is now dominating installations (because it is so much cheaper), and (3) ignored deaths from uranium mining, which is where most of the release of radioactive material is in the fuel cycle.

>For the contractors-on-roofs, the slip & falls add up even though they'll never get an HBO mini-series.

Are solar roofs significantly more dangerous to work on than a normal roof for some reason? Roofs are going up whether they have panels or not.

>This is just not true

A lot of smart people would beg to differ.

> Hydro is dirt cheap already

Said HN reader as the Three Gorges dam was about to burst in China, submerging millions of households.

> Hydro is dirt cheap already.

Apparently, hydro produces a significant amount of greenhouse gases within the first 100 years or so, resulting from decaying plant matter and so on.

There was a story on HN a few years ago, IIRC it listed hydro providing 4% of the world's electricity while being responsible for 1% greenhouse gases.

I couldn't find the original story, but I didn't look too hard. Here's the first other source I found:

https://academic.oup.com/bioscience/article/66/11/949/275427...

I would like to see some models on how solar power could work in northern cities, or honestly any city. The NYC metro area has over 23 million residents and there is essentially zero space for a solar farm. In the winter it's cloudy almost all day and there are only 9 hours of sunlight.

Rooftop solar is an option but the models I have seen show a theoretical max around 1,200 MW which is about 1/30th of the NYC metro area electricity requirements.

I just don't see how solar is even close to being viable for anything outside of small cities with access to massive swathes of empty land.

You could imagine building a globe-spanning power distribution network. Solar panels on one side of the planet could provide power at night on the other side of the planet.

I bet the power transportation cost and loss would be too significant.

If we could transmit power across vast distances like that then we wouldn't even need to bother with solar. We could build a thousand nuclear power plants in the middle of nowhere.

Unfortunately, power transmission suffers from line losses. Even a few hundred miles requires several hundred thousand volts to avoid losses. Maybe that's an easier problem to solve than fusion.

You'd JUST need to move current from AC to DC, and span the whole of earth with superconductor cables.

You know, easy stuff to build and maintain.

Use the local electricity to crack water into H2 and O2. Release the O2, use the local electricity to compress the H2 into liquid hydrogen, pump in across the country. The pipes themselves are a storage facility.

China has built a 1100 kV line for 12 GW with losses of only 1.5% per 1000 km.

source: http://en.people.cn/n3/2018/0622/c90000-9474097.html

That's impossible, unless you have an invention for extremely cheap superconductors that you've been hiding away.

Electricity transport losses makes that impossible.

That isn't even the major electrical cost of supporting those people. The fertilizer needed to feed them, metal refining/recycling, chemical reagent production, shipping fuel, all dwarf the usage by individuals in their homes.

Personally I don't think solar is ever feasible for industrial scale work, covering like 1/3 of your landmass is a ridiculously large project and needed to fuel industrial operations that currently use fossil fuels to fuel them including fertilizer production. Even if you can shrink that down a bit, how much of it is competing for farmland? Or replacing natural forests or plains or other wildlife housing.

Where did you get 1/3? You dont need to cover even 0.001% of landmass to supply all of US!

https://model.energy/

Generally, low latitude places want more solar + batteries, high latitude ones want more wind + hydrogen. When optimizing with these four, plus nuclear, with the costs at that site, nuclear typically optimizes to 0%.

Using their data for the UK (2011 weather) and the 2020 technology scenario (rather than the projections which conveniently make renewables and storage significantly cheaper and Nuclear costs constant) and requiring that the system be carbon neutral it only takes very small tweaks to their assumptions to make Nuclear a clear winner. They are assuming by default that countries will be able to use salt caverns for hydrogen storage, which seems unrealistic given the amount of it their scenarios require. Using the numbers they give for steel tank storage instead, it only takes a 5% cost reduction for nuclear (which seems self-evidently possible with economies of scale) to make the optimal solution a 100% nuclear grid.

Right now I don't think people are pragmatic enough for that to be politically viable in most western countries, but that may shift as the adverse effects of AGW start to be more acutely felt (and, hopefully, as more and stricter carbon taxes are implemented across the world).

Europe has like 100x the salt cavern capacity needed. Hydrogen can also be stored in deep aquifers or hard rock caverns.

Using the 2030 data is proper, since any nuclear reactor we begin to build today won't be available until about then (which renewable and storage systems can be built in just a couple of years.)

Some of their cost figures are already too high, btw. Their 2030 estimate for the cost of electrolysers was 600 euro/kW; it's already down to half that (or even less, in China).

No I guess NYC needs to keep that nuclear power plant in Lower Manhattan then.

I know you’re saying this as if it should be something scary that I should oppose.

But I am not even remotely concerned about nuclear safety. If noise and traffic were no concern I would live right next to a nuclear power plant.

No I say this because if someone says solar is impractical in NYC and but nuclear somewhere else is, the comparison does not make any sense.

Hydro can actually be worse than coal in terms of ghgs. All depends on what they flood. If it's forest, it's bad.

>>nuclear power is the only realistic way to solve the looming energy crisis of the 21st century while still maintaining the same standard of living for everyone

This is also not true if we dealt with the massive overpopulation of the earth instead of acting as its some sort of moral crisis not to stuff as many people as possible onto a planet with finite renewable resources.

How do we go about dealing with the massive overpopulation?

As per sibling comment to yours, society wide development seems to lead to lower birth rates. Lift the masses out of poverty.

One of the simplest ways is to offer economic incentives to people who have fewer (or no) children instead of the opposite (which is our current policy).

Soon enough we will have the opposite problem to worry about.

This appears to be a problem that is solving itself. Developed countries have birth rates below replacement and more countries are becoming developed (with the same trend happening as those countries develop).

That’s all well and good if you reach stability below a level that is sustainable. We aren’t going to be anywhere near that from an environmental point of view.

And whybdo you believe that? We will never surpass 12 billion by most estimates. Where is the sustainability point?

Do you think those 12B people are going to pollute the ocean with less plastic or agricultural runoff?

The last time there was a "population crisis" fertilizer was invented the problem vanished. Currently GMOs have shown they produce massive improvements in yields and we're still in the early phases of this research. So it isn't exactly naive to think that this won't be an issue (though I'm not saying you should completely dismiss it).

As for land, well Europe is the same size as America and has twice the population. And neither have anywhere near the density that China or India does. So there's quite a bit of evidence that we shouldn't be overly concerned about these issues.

And what looks like the best way to solve them is by helping other countries develop more rapidly (which is actually a bonus for tackling climate change too!) but this is a pretty unpopular opinion.

The sustainability point is where we can survive as a global population without destroying and depleting the resources of the planet. What point that is exactly is impossible to say because we have never done proper research (as far as I know), but what we can say for sure is that global population passed that point a long time ago. Mass extinction, overfished oceans, deforestation, saturation of the environment with chemical pollutants (not limited to carbon), soil degradation from nitrogen saturation (and the nitrogen runoff that has created massive dead zones in our oceans and other waterways) - the ways we have destroyed our planet in order to sustain our overpopulation go on and on. As evidenced by the downvotes to my original comment, the unfortunate reality is that many, if not most, refuse to accept reality out of religious-like delusion. Some worship at the altar of technology, some believe that its immoral to see a moderation of human population, some worship at actual religious altars and believe its blasphemy to even discuss the issue, and some blindly accept conventional wisdom that we need an ever-increasing population in order to sustain the economic ponzi scheme that modern society rests on (despite the fact that we have a massive supply of surplus labor).

One more time: we cannot destroy our planet, even if we wanted to. We can make it less comfortable to live in some places, though.

Also "we don't know where the point is, but we have passed it" is a very bad argument. "What can be asserted without evidence can also be dismissed without evidence".

https://doi.org/10.1016/S0140-6736(20)30677-2

Are you factoring in the costs and environmental impact of producing and maintaining storage for solar, the production and maintenance of solar cells and all the other externalized costs and impacts of so-called renewables? From what I understand, there is a great deal of hype around these technologies that doesn’t really stack up and that the renewables industry is rife with shady practices and bogus claims.

Personally, I think nuclear (fission and hopefully fusion) are the best options, though I can accept a minor niche role for tech like solar.

The problem shouldn't be thought about in this way, or maybe I'm misunderstanding your position. There is no single technology that will solve the issue of climate change (and in this case just limiting our conversation to the production of electricity). The landscapes in large geographical areas like the US are not homogeneous. Solar and wind work great in the Southwest, but not as well in the Northeast. Hydro works fantastically in the South, but not in Southwest (yes I'm aware there is hydropower in the SW, that's not the point). There's questions about base load, storage, peak and intermittent demand, transmission, and so on.

So I dislike these conversations because they seem to be framed as Solar/renewables VS nuclear. When really the conversation should be "should nuclear be part of the solution?" I do believe that the answer is yes (because above factors), but the phrasing matters. This is because it leads to the next obvious and more important question: "If yes, how much?" Clearly a fully nuclear grid is not a smart idea, just like a fully solar grid wouldn't be. But the framing matters. It isn't a "OR" debate, it is an "AND" debate.

When accounting for all externalities, hydro doesn't usually pencil out all that well. It's a sad reality, because through a certain limited lens hydro looks amazing.

There are many efforts underway removing old hydro dams to restore the environments and ecosystems that were totally devastated by them.

A notable recent example is the Elwha River Dam removal that finished about 4 years ago in Washington: https://therevelator.org/elwha-dam-removal/

The question about needing nuclear really comes down to how you model energy dependence. Is the US (or X country) homogeneous in ability to produce, store, and distribute energy or is it heterogeneous?

Of course, things get much more complicated once we start talking about lifetime emissions and external environmental impacts.

I am a fusion power fanboy but would hope other researches are getting some loving too... I mean what about seasonal energy storage? would be a great technology in current times where weather's getting wilder. we capture excessive heat and dump it back when it's cold etc...

With tokamaks, there is a concept of confinement time. It turns out that plasma is not stable in a standard donut configuration and it likes to twist and break the continuous column, sending all the plasma shooting at the walls. The confinement time is how long it takes for this to happen. After the many plasma experiments over the decades, scientists have a really good idea of how the confinement time scales with magnetic field strength and volume of the plasma. It turns out that you want a big machine. With the JET tokamak they were really close. Iter is designed to have the confinement time well within the requirements for net positive energy, with like a 10-fold margin of safety. IMO it should work, but it is taking an unacceptably long time to construct.

If I recall correctly, the ITER project was very badly managed until 2015, resulting in the project being about 6 years late. A new management took over then, and the project is now moving on pretty swiftly. But indeed it's taking way too long...

Containment energy is proportional to surface area, while power produced is proportional to volume. Thus, scaling up should make it easier to generate energy.

If as you say...

"nuclear power is the only realistic way to solve the looming energy crisis of the 21st century"

because...

"most people around the world think that nuclear fission is scary"

then we are fucked as a species.

Science does not respect people's feelings. If people have scary feelings about energy/climate change/wearing masks to prevent a disease that travels in the air, then those feelings need to be changed, or those people need to be sidelined.

What we must not do - what we cannot afford to do - its to derail the science to assuage the deniers.

Given the opposition from our left to nuclear and painting any nuclear proponent as a right-wing science denier trying to defund renewable energy I’m not sure what we can do. Michael Schellenberger has been pilloried by our left despite the decades he put into different approaches before arriving at his nuclear advocacy (and perhaps also deeming for a lot of the climate change doom and gloom as alarmists for the 12 year cliff being cited). But in my view climate change is already here and even if we cut all emissions to zero worldwide we’d be kind of screwed, just not as badly screwed.

Holistic yet science-driven conversation on climate change doesn’t seem possible anymore in public discourse and everyone is angry, which doesn’t make them sound exactly objective either.

Viewing this in terms of left-right politics rather than in terms of hard data is one of the primary reasons that nuclear fails again and again as it is tried in the West.

The problem isn't politics, the problem is construction logistics. Nobody knows how to build large projects effectively anymore, whether it's the Big Dig or something super complicated like miles of piping with precision welds and specialized concrete pours.

If you want to find out why nuclear hasn't worked, look into all the individual cases of construction from Vogtle to VC Summer to Hinkley to the UK's Sizewell C, to all of France's EPR efforts.

What you will find is that political fear and regulations are not the problem. It's just management. And when you go back to the US's failures in the late 70s and 80s, you see a similar story of management failure causing construction financial disasters.

Even South Korea's apparent successes in construction have been rocked by revelations that inspections were skipped and completed through corruption, nor competence.

The primary reason nuclear survives in discussion, IMHO, is as a political wedge issue. Actual political discussion has no connection to reality of the subject matter (as is the case with too much of politics). Shelleneberger is a prime example of the afactual debate when it comes to nuclear. He's looking for ways to convince people, and attract followers, not in rational and informed discussion, and a few minutes of fact checking typically makes short work of his screeds. The anti-nuclear political argumentation is just as bad, and as easily destroyed.

The solution is to elevate the debate, and look for actual ways to build nuclear if one thinks it will be a useful tool. That means abandoning the large reactor model and trying SMRs; however I have little hope of those being economical, unless the waste heat or primary has some specialized industrial uses that are uneconomical from electrically driven processes. But at least the will likely be constructable.

It’s not just about making big things. China has recently made some of the largest structures mankind has ever seen, yet their fusion program is decades behind US, Russia, Spain, UK, and France.

I’m a modular reactor advocate as the most rational approach in fission for the future for the reasons you stated - our collective inability to construct and manage complex monoliths seems to be a problem in physical and software construction regardless of culture or economics as humans. This causes reactionary regulations that bloat costs and thus efficacy. We’d need less stock market and banking market if markets didn’t explode all the time, for example, and lay waste to millions of people’s living, but they don’t take care of themselves exactly without bad externalized costs.

Here's a video where an energy professor goes through the finances of gas vs nuclear power (very slowly) - https://www.youtube.com/watch?v=cbeJIwF1pVY

In short, it takes a very long time for a nuclear power plant to become profitable. Investors would rather pick something like gas that starts being profitable earlier, so they have the profits available to reinvest into other enterprises.

I think this, rather than public perception, is why we don't do nuclear power (after all, our society does plenty of other things despite similar levels of outcry).

"is if a laboratory experiment cannot even produce desired outcomes, "

the lab experiments delivered the desired results. The data made it possible to project how a fusion reactor can work. The same happened with nuclear bombs. They did a lot of small scale experiments until they knew what was needed to produce a function nuclear fission and later fusion bomb.

I don't know about the details of why size matters, but in terms of producing energy: "When supplied with 300 MW of electrical power, ITER is expected to produce the equivalent of 500 MW of thermal power sustained for up to 1,000 seconds" [1]

That's thermal power, not electricity, so that's not enough yet to break even as a power plant, but it's a big step in the right direction.

[1] https://en.wikipedia.org/wiki/Fusion_energy_gain_factor

An excellent quantity that indicates progress on fusion and how far designs are from breakeven is the triple product. Here is an EXCELLENT review of fusion progress so far. (https://www.fusionenergybase.com/article/measuring-progress-...)

Is it really clear the fusion power, even if it's achieved, will actually be cheap enough to use? Sure, the fuel is almost free, but the fuel for solar, wind, and hydroelectric power is also free and that doesn't make them cheap. If we're not worried about cost, we could build enormous batteries using existing technology and use solar and wind to power everything. Fusion has to be cheaper than that, and judging by the cost of ITER, it doesn't seem obvious that it can achieve that.

Remember how nuclear fission was going to be "too cheap to meter"? Turned out it's expensive to build and operate the plants, regardless of how cheap the fuel is.

Well, fusion has the benefit of having much lower operating costs. Violent terminations, quench events, and tritium containment breaches are serious concerns, but a risk assessment of a fusion power plant would give a much cozier feeling than a risk assessment of a fission power plant.

The neutron activated materials are safe after a century and there is not much of it.

All of the cost in a fusion plant sits on manufacturing and maintenance. If you can build a GWe plant for $30Bn or less, then it’s a no brainer. The five trillion dollar question is how much investment fusion power research needs before we get to that point.

Fusion will likely have much MUCH higher operating cost. Fusion reactors are fearsomely complex things that will be too radioactive for hands on maintenance. They will be operating at much higher levels of radiation damage and thermal stress than fission reactors. That all adds up to a reliability and maintainability nightmare. The fuel being cheap doesn't mitigate this.

Also, the need to replace major reactor components many times over the life of the reactor, due to cumulative neutron damage, will itself cause very large operating costs.

> What makes scientists believe that the solution to nuclear fusion power is only a matter of scale?

The sun is nuclear fusion, so it is proven to work if your scale is big enough ;-)

>I desperately want nuclear fusion to work because nuclear power is the only realistic way to solve the looming energy crisis of the 21st century while still maintaining the same standard of living for everyone.

If that's all you care about we can do without nuclear fusion. We have a sun in our neighborhood so we can just harness its energy in various forms such as wind, fossil fuels, biomass or via PV. The real reason we need nuclear fusion is because it is necessary for interstellar travel. There's no sun in deep space. You'll have to bring your own fusion reactor with you.

> What makes scientists believe that the solution to nuclear fusion power is only a matter of scale?

I guess there are too many physicists who can lobby well and their income depends on it.

I don't expect any results after the many attempts (as opposed to fission, which had almost immediate results).

In case anyone read this comment and wanted the real answer: empirically derived scaling laws from a series of machines increasing in size. The last/largest is JET.

http://fusionwiki.ciemat.es/wiki/Scaling_law

I think fusion is cool, but I also think it distracts from more important and promising approaches to clean power. No one has ever demonstrated a self-sustaining, net positive fusion reaction. It will probably take >30 years to make this technology cheap , safe and reliable and another >20 years to deploy it at scale. We don't have 50 years to wait on clean energy, we need it today. We should be focusing our attention on renewables, like solar, which are rapidly dropping in cost, and good 'ol nuclear power, which is already proven to be safe and reliable, instead of crossing our fingers and waiting for some magical tech to save us. We already have the tech, we just need to get serious about deploying it. I bet a concerted roll-out of nuclear + solar could cut US emissions due to electricity generation by at least 30% over the next 20 years.

EDIT: I see this is getting some downvotes. I want to reiterate that I support fusion research, I am just saying that I don't believe fusion is the quickest or most realistic approach to reducing emissions.

Don't be so alarmist. We don't need it today.

Everybody else but the ITER people are focusing on solar and wind. What more do you expect?

I agree we should look at multiple possible paths for renewable energy, but to say that they shouldn’t work on this, is a bit far fetched.

This assumes the researchers working on fusion can work on the other renewable energies, when that’s typically not the case with highly specialized fields. Further, estimating the time it takes for advancement with the described time magnitude is a frivolous pursuit, if you put any weight behind it. Within 5 years, sure let’s predict. But at 30 years, now we’re talking a bit too far into the future. It’s much harder to guesstimate what will occur from now to then.

Lastly, if we assume it will take 30 years to produce a working model, then that only assumes we’re working on it today. If we could, let’s say, divert all scientists away from fusion research onto other renewables, the 30 year estimate would be perpetually pushed until we start working on it once more. Let’s have people who are passionate about their field pursue as many paths to solving the energy crisis as possible. We could have a breakthrough sooner than 30 years, which I would place money on.

Your argument is sound, but the issue is also budget allocation: financial resources are not infinite, and we must choose how to allocate them.

That's a very good point, we need to allocate much more towards fusion R&D

There’s so much things to divert budget from to fuel energy research. Completely unrealistic, but think luxury cars, football player’s salaries, just to name a few.

Coal subsidies …

Military industrial complex? Billionaires? The things you mentioned are literally nothingburger.

It's very clearly not a nothingburger.

High paid professional athletes in the US today have the income of billionaires.

Professional athletes earn around $200 billion in league salary each decade (it excludes endorsements, which tilts the point that much further if included). It's far more than all S&P 500 CEOs combined. Just the three major sports leagues in the US are at around $14 billion per year in salary. That excludes coaches, which now have very high salaries as well. Stripping it down further, all NBA players combined (a mere ~450 players or so) earn more in salary every year than all the S&P 500 CEOs combined.

As of 2019 there were 122 players in Major League Baseball earning $10 million per year or more. The top 30 NBA players earn a combined $1 billion every year.

Pretty funny, the notion that ~$80+ billion in income taxes every ten years, is supposedly a nothingburger. It's a lot of tax revenue, even if it pales next to US Military spending.

Can you explain why and how this distracts from renewables? I don't get how you reached this conclusion.

I do not think solar is going to work. As you say, it might take up to 50 years to crack the case on nuclear fusion. I argue that it might take up to 50 years for governments and companies around the world to even begin cooperating in a fashion such that they make a dent in carbon-based energy sources at the global scale using solar.

The reason why solar is not going to work mostly is because there is no legitimate mass installation of batteries that can handle the demand of energy. And the reason why the batteries cannot handle the demand of energy is because governments are not serious about deploying more, and people are already comfortable with the quality of life they already have, so neither are going to step up to make this work. And to be fair, why should the average citizen impose restrictions on themselves to solve this crisis? It's not realistic. Governments are supposed to be the ones to make these initiatives. Analogy: I hate having to pay for taxes, but I like that the government gets it from everyone, so that it can implement common services that benefit a lot of people, which wouldn't be possible otherwise.

The only way is nuclear fission because, as you said, it is proven to be safe and reliable. However, the people have spoken: They do not want "dangerous" nuclear reactors around, even if those impressions were formed by reactors designed and built in the 1950s-1960s. I personally think this is a shut case. There is no convincing of the people to go to nuclear fission anymore.

It's a bit odd to me that you seem to deem fusion power realistic, while discounting other ways of generating and storing power that are basically available already.

Power from renewables can be stored for the long term as hydrogen. That's well understood technology, scaling up is an engineering problem. When the sun shines and the wind blows, solar and onshore wind are the cheapest ways to make electricity, so there's a huge economic incentive to solving the storage issue.

Thanks for pointing out, because now that I think of it, my view does come off as odd. What I think it reflects is that despite nuclear fusion being a technical, engineering problem, I believe getting companies/governments to abandon carbon for wind/solar to the point of carbon-neutral is even more insurmountable. Does that viewpoint make sense? For them to get off fossil fuels, it's going to need a ball-out-of-the-park solution. It's too hard to compete with the fracking revolution.

Why do you think economic incentives won't be enough? Solar and wind are already the cheapest energy sourve in terms of production. We just need large scale storage to become cheap and practical and we're sorted.

The UK government has committed to being carbon neutral by 2050. That's well before you're even prediciting the fusion becomes available.

Even if we deploy batteries, wind and solar in an amount needed to mitigate a climate change crisis, the amount of mining and industrial activity that would take would greatly poison our planet in other ways. In terms of "other stuff you need to do" per MW none of these come close to touching fission. Fusion, being much cleaner than fission would be a holy grail of clean cheap energy.

I hope it only takes 50yr to figure out and scale up fusion, that would be great.

What part of a wind turbine 'greatly poisons iur planet"?

Its an electric generator on a mast. How is it more poisonos than any other industrial object or consumer device?

>What part of a wind turbine 'greatly poisons iur planet"?

The batteries they're going to be paired with unless some other storage tech becomes better or windmills become so cheap that we can use less storage or inefficient storage. The windmills themselves are mostly fine since they're just fiberglass, metals and concrete.

Mining for the things that go into batteries is just as destructive as the mining done for coal. It just happens in Asia so nobody cares. If you destroy the suitability of land for habitation or certain crops (likely by affecting the watershed) then you create many of the same problems climate change would/will cause. If you make mining less destructive to the environment batteries become more expensive hampering renewable adoption. It's a catch 22.

Don't get me wrong, it's better than fossil fuels and better than anything involving a solar panel (which has all the mining related dirtiness on the energy production side as well as the storage side) but compared to the bang for your buck you get by knocking atoms around it's a hell of a lot of industrial activity.

> "Mining for the things that go into batteries is just as destructive as the mining done for coal."

Most of what goes in to batteries can be recycled. You can't recycle coal.

They dont have to be paired with batteries, there are several cheaper industrial storage systems like comoressed air and pumoed storage, or concrete blocks moved vertically

Net-positive was achieved in 2013. As far as I can remember, ITER is a test-bed for economical and scalable nuclear fusion.

Also, the corporate world (at least here in Texas, of all places) is perfectly capable of rolling out state-scale renewable power (wind).

> Net-positive was achieved in 2013.

Do you have a source for this? I was also under the impression that fusion has never yet produced a net of energy.

Probably referring to this highly controversial re-definition of Q: https://en.wikipedia.org/wiki/Fusion_energy_gain_factor#Scie...

If you define your efficiency metric in a clever way, anything is possible: https://io9.gizmodo.com/breakthrough-the-worlds-first-net-po...

Nope. JET achieved Q~0.7 in 1997[1],and that hasn't been beat since. In 1998, JT-60 would have gotten Q~1.2, had they put deuterium-tritium fuel [2] (but it wasn't built for that).

You might be thinking of some news from NIF. In 2013, they sorta claimed to have reached ignition, but that was only by moving the goal-posts.[3]

ITER is not by any means a test of an economical or scalable reactor -- it's still at the stage of a science experiment.

[1]https://en.wikipedia.org/wiki/Joint_European_Torus [2]https://en.wikipedia.org/wiki/JT-60 [3]https://en.wikipedia.org/wiki/National_Ignition_Facility#Bre...

I would argue that we are already seriously building more and more solar and wind power, along with research into better battery storage. Even if we solve the current climate crisis with solar and wind, we may still be able to do even better with fusion in 50 years. Fusion would be very energy dense, requiring a lot less resources and land than solar or wind, and has the potential for even cheaper and less environmentally-destructive energy than current renewable resources.

> "I bet a concerted roll-out of nuclear + solar could cut US emissions due to electricity generation by at least 30% over the next 20 years."

30% over 20 years in unambitious. The UK has cut grid carbon emissions by 60% in the past 10 years. (2010 average: 467g/kWh. 2019 average: 189g/kWh). Emissions continue to decline every year.

This was largely achieved by closing coal power plants, and building a lot of wind (both on and off-shore), solar, biomass, and new interconnections with continental Europe. No new nuclear was built in this time. Wind is now the second largest source of electricity in the UK, after natural gas. (But natural gas usage is also declining!)

Just switching over all new development to nuclear and ignoring the luddites would basically make us guilt free and green, but that won't happen until we're desperate.

agreed!

the CEO of JET (largest working fusion reactor, until ITER is finished) Ian Chapman said in a lecture that renewables (and nuclear fission) are key to holding off climate change until fusion is a more mature technology

why should the research be mutually exclusive though? there's resources and brainpower to pursue both.

It's not an either/or... we should be pursuing lots of different avenues for power generation because we don't know what will pan out and what won't. Fusion might still be a huge win even if it takes 50 more years to perfect.

butter and guns

At this point I've got a lot more hope in the MIT/Commonwealth Fusion Systems approach with REBCO magnets: https://www.youtube.com/watch?v=L0KuAx1COEk

At this point it looks like ITER is hampered by it's relatively old supeconductor technology (ultra low temp/moderate field strength traditional magnets vs high temp/high field REBCO magnets).

ITER and Commonwealth can (and in my opinion should) be seen as complimentary endeavors.

ITER has been designed with relatively conservative magnet technology and will very likely provide the physics results that need to be understood in order for fusion power to become a reality. This includes experimental tests of the physics of plasmas where the heating is dominated by high energy alpha particles rather than external heating. This is a regime that's not yet been studied in a laboratory and there is important research to be done there.

Commonwealth is pushing the envelope of high temperature superconductor magnet technology and is relatively high risk compared to ITER's magnets (and this is a good thing). Lots of ITER technology will be useful to Commonwealth even before ITER turns on. For example decisions about which low activation steels and the huge amount of physics work that's already gone into planning for ITER.

I think the most likely outcome is that both accomplish their goals and contribute to making commercially viable fusion energy a reality in the future.

Luckily, since ITER and SPARC are being built with the same aspect ratio, all the learning about plasma control, materials, cooling, tritium, remote handling, etc. is fully transferable.

The tritium extraction and processing is a whole separate multi-story building full of first-of-a-kind equipment which will be 1:1 transferable to any breeding fusion reactor.

The work that ITER and IFMIF will be doing on material lifetime and handling under heavy neutron bombardment - a really substantial engineering problem - will be fully transferable.

The work on first-wall material which has to handle the neutron flux, very high thermal loads, and not poison the plasma when traces of it come off is also fully transferable.

Basically everything that's really new about ITER except for the size will work the same way on an HTS based machine.

I'd say about 2/3 for the science done at ITER would need to be done for any D-T fusion device, another 1/6 applies to all similarly configured tokamaks (i.e. it's less less relevant for stellarators or spherical tokamaks), and 1/6 is ITER specific (high estimate TBH).

If things run according to schedule (obviously questionable) then in 2025/2026 ITER will have first plasma and CFS will be on schedule to start building SPARC. CFS is being very clever in doing all their magnet design work first - investors are funding it because even if they don't get either SPARC or ARC funded and built, they will at least have some very useful IP on large HTS magnets which is bound to be worth something.

18b doesn't seem like a huge amount of money these days with the economic stimulus funds being announced around covid-19.

For some reason I often think about things like this in terms of NHS budgets.

18bn is, very roughly, 2 years of healthcare for London.

so, how many healthcare's do we have budget for in a year?

Or like three pre-launch startups

I don't even want to know how may Aeron chairs you could buy for that money.

Why did you pick London as an example? It has about 0.1% of the world population.

It is hard for me to conceive of the steps required in these projects that add up to years - just in assembly.

Is it due to precision requirements being difficult to achieve? Supply chain delays for custom parts? Lots of experimental runs that require dis-assembly and tuning?

Making complicates shapes these sizes is very very hard. And the fact that it's a one-off project for now makes it so much more expensive and hard. From my understanding somewhere some factory would have to dedicate a whole production line just to this one device.

There is no supply chain for a lot of the parts in these projects. Even for a board that doesn't require a custom chip (You can buy some ridiculous high-spec chips off the shelf these days), the silicon will be expensive, the PCB will have to be made to a very high tolerance etc. To top it all off you may require engineers who are not only experts in their own field but also require some in depth knowledge of aspects of theoretical physics (i.e. If you are looking for new particles your usual rules of thumb may not work) which for the most part engineers don't go anywhere near in their normal education.

On top of all the custom hardware you have to be able to set up the infrastructure to use your shiny new hardware, places for staff to work and then land to put it all on (And this is on top of the cost of actually building the main experiment).

In short, everything involved in experimental science is expensive.

One part of it is that it's a research project. The other part is the bureaucracy of international cooperation and the pieces being built all over the world according to a huge centrally managed specification. It's not like SpaceX that starts with a small design and then builds incrementally larger versions to have something to show in the meantime. It's from 0 to finished product.

I'm currently working for IFMIF (sister project of ITER), and I know some people who work or worked for ITER, and yes, the bureaucracy is an absolute nightmare. Coordinating so many countries with so many different working cultures is really difficult.

On the other hand, we DO build incrementally larger versions. The machine is built and tested in stages, it would be impossible to build such complex machines in just one go.

> On the other hand, we DO build incrementally larger versions.

You're referring to JET? Does ITER benefit from production facilities that were used for it?

First of all, I'm just a computer guy who joined this field two years ago. I still don't know much, so please excuse my ignorance.

Regarding JET and ITER, according to what I heard from my coworkers, all the knowledge that was gained at JET is being used for ITER. I guess many of the productions facilities are being reused too, but I would need to ask them.

Regarding our project (IFMIF), we are building a linear accelerator to simulate the neutron flux inside a fusion reactor and study the behavior of different materials. The accelerator is composed of an injector, several acceleration stages, and a beam dump. First the injector and the beam dump were installed and tested, and then all the other stages are being installed and tested incrementally. But a linear accelerator is a very different machine from a tokamak.

I'm asking because afaik (and that's fairly superficial knowledge) is that JET is mostly a european project while ITER is built internationally and the production is intentionally assigned fairly (to spread experience to contributors) rather than with a focus on efficiency. So that implies to me that there would be many new manufacturing places all over the globe. In other words it's made from scratch, more or less.

I'm surprised that we're still chasing moon shots in strong force fusion, while weak fusion, which has actually been demonstrated already, gets little attention. Chemically assisted/low energy nuclear reactions have been proven to produce clean energy and useful transmutation. This process can be used to turn radioactive waste into inert material. For example, technetium-99 is a particularly pernicious by-product of nuclear fission. It has a half-life of ~200,000 years and is very mobile in the atmosphere, making it difficult and expensive to handle. In a CANR reactor, this nuclear waste is converted into the safe, and valuable, palladium-106, and the by-product is clean heat. Given the double benefit of clean fuel and nuclear waste recycling, I think this should be a bigger priority than strong-force fusion.

http://nautil.us/issue/86/energy/einsteins-lost-hypothesis

https://www.jstor.org/stable/24216601?seq=1

https://www.sciencedirect.com/science/article/pii/B978012815...

So you suggest that they should dump nearly 2 decades of work because there is something else which seems to be a potential energy source? The ITER project was founded in 2007. From there on hundreds of physicist, chemists, engineers and computer scientists started planning this RESEARCH project in order to better understand this kind of technology. Shouldn't we stick to what we have started (since this is (more than) cutting edge technology) and have a look at other technologies in parallel?

> So you suggest that they should dump nearly 2 decades of work because there is something else which seems to be a potential energy source?

OP said nothing of the sort. They said that they didn't understand the prioritization of funding given an alternative that they found personally much much more valuable.

I am pretty sure that this project is not prioritized in any way. It's a research project with funding from all over the world. It should be compared to the CERN project where scientists from all over the world try to understand science behind it better. One could argue that CERN is a waist of money since there are other fields of study in physics which seem to have more potential. We should be happy that states from all over the world are willing to pay for a research project, which does not have a direct "value" for the next couple decades. There will never be a private investor for this scope of project since there is no monetizable value behind it yet and maybe in the next 50 years.

No one claimed the opposite.

That's the very definition of a sunk cost fallacy

Came here to say just this. While I'm glad to see that research is finally culminating in investment and assembly we shouldn't be committing to this project just because people put time and effort into it.

Of cause we should put time and effort into it because it is not proven that it is wasted. There have always been people who said CERN is a waist of money but in the long run, mankind wouldn't be the same as today if we didn't fund it back in the days.

Not only that, before ITER we used to work on JET in 1991, which is in kind of a twin relationship with ITER. Abandoning ship now for a technology that isn't even proven to be viable in large scales would be insanity.

> ITER began in 1985 as a Reagan–Gorbachev[18][19] initiative[19][20] with the equal participation of the Soviet Union, the European Atomic Energy Community, the United States, and Japan through the 1988–1998 initial design phases. Preparations for the first Gorbachev-Reagan Summit showed that there were no tangible agreements in the works for the summit.

https://en.wikipedia.org/wiki/ITER#Background

You're surprised that they would continue to pursue something that we are now pretty sure will work (though not necessarily cost effectively) instead of some quackery?

If we stopped what we worked on every time some nice theory came around which could work "better" (for whatever definition of better) we would never get anything done.

Scientists/companies working in these areas are free to produce results that make ITER obsolete. After they've done that we can discuss about abandoning the - from our current understanding of science - most likely path to (more or less) unlimited energy. So far, they are not further along than let's say fusion experiments in 1960.

But is the consensus that ITER will be self-sustaining from an engineering perspective?

It’s predicted to be able to have a Q>1.

This lecture was enlightening: https://youtu.be/L0KuAx1COEk

You're confusing the strong and weak nuclear forces from particle physics with the fission and fusion nuclear reactions. The (residual) strong force underpins the energy release in both fission and fusion reactions.

You've also linked to the debunked pseudoscience of 'cold fusion', (dressed up as low energy nuclear reactions...) from Pons and Fleischman.

Plenty of work is being done in the field of handling radioactive waste from nuclear fission. My field (decommissioning and waste management) is entirely dedicated to it.

From the first link: " It was only in 2006, with the publication of a landmark paper in the European Journal of Physics C, that neutron-induced transmutations, as something distinct from cold fusion, began to emerge as a viable theory. The paper predicts that electrons on a metal surface coated with hydrogen, deuterium, or tritium atoms can behave collectively (as Einstein had predicted) when driven by an oscillating electromagnetic field at a particular frequency. This collective behavior can give them enough energy to combine with the hydrogen, deuterium, or tritium to make neutrons.

The paper goes on to say that the resulting neutrons travel very slowly, slow enough, in fact, to get gobbled up by a nearby atom before they can even leave the microscopic vicinity of their birthplace. The atom then becomes unstable and might burp out radioactive decay byproducts like a gamma ray or energetic electron. A separate paper by the same authors calculates that microscopic surfaces of electrodes, like those that tend to produce low-energy neutrons, are efficient absorbers of radioactive gamma rays. So radioactive decay can be transformed into a bath of innocuous heat. And of course heat energy can readily be converted into electricity."

Landmark paper is a strong term for something that has been repeatedly debunked and then ignored by the nuclear physics community (all its citations are cold fusion journals...)

The theory Widom and Larsen propose violates the laws of physics and has no experimental evidence.

Put simply, the weak reaction needs an enormous amount of energy to happen. The difference between a neutron and a proton is 1MeV, more than the mass of the electron. The electric field at the surface of the metal is of the order of eV and does not account for this. So where are these massive electrons magically getting their energy from?

Or if you're comfortable with theoretical physics, a better critique can be found here: https://coldfusionblog.net/category/widom-larsen/

> which has actually been demonstrated

For certain values of "actually be demonstrated". I don't think it's been reliably shown to be a real phenomenon, has it?

That is my impression as well. It needs a lot more peer-review and reproducable results in different labs around the world by different scientists.

In case you missed it, they are basically talking about "cold fusion".

From Here: https://drive.google.com/file/d/1p6jrlbUOuuOCHnhR9fcsvnQJl4d...

"Most of the power (~100 MW on ITER) released in SOL flows in an extremely narrow channel ~1 mm"

wow

Yeah, steady-state heat fluxes comparable to re-entry are expected.

Sorry I'm a layman at best on the subject, buts whats the difference between this project and, for example, W7-X?

W7-X is a stellarator [1], and ITER is a tokamak [2]. They are two different approaches to achieve controlled nuclear fusion.

[1]: https://en.wikipedia.org/wiki/Stellarator

[2]: https://en.wikipedia.org/wiki/Tokamak

W7-X is strictly for experimenting on plasmas, it's not supposed to break even on energy output, ever

To get fusion there is a couple of ways: There is gravitational confinement (that is how stars do it, but it is impractically large for humans), there is inertial confinement (get enough energy out before it explodes, but that only works for hydrogen bombs and NIF, not for a reactor) and magnetic confinement (that is what we intend to use for reactors).

Magnetic confinement works because a plasma consists of charged particles that gyrate around the magnetic field lines. So (to first order) they can not escape across field lines. But they can move along the field lines and hit the end of the device. The particles move fast, so simply making a linear device long enough is hard. So the next idea was to bend the magnetic field into a torus (the shape of a donut), because that way there is no end to the magnetic field lines. Unfortunately that configuration is unstable, the plasma donut will bend and twist until it hits a wall, stops being a plasma and falls to the ground. There is basically three option out of this problem:

1.) Tokamaks [1] such as ITER. Here we induce an additional current inside the plasma that goes around the hole of the donut. That produces a small additional field that stabilized the current. But driving that current can be hard (using what is called a plasma transformer work only for a limited times, but AFAIK that is not the first limit on a discharge that ITER will hit, wall heating limits single plasma "shots" to shorter times anyway). The upside is that the design is rather simple, which implies you don't need millions of core hours to design and the fields coils are reasonably easy to produce.

2.) Stellarators [2] such as W7-X. Here the field coils that produce the field lines in the donut are intentionally twisted to produce a more complicated magnetic field. The upside is that we do not need the current in the plasma and get better performance (for a device of similar size), but the design of the field coils is hard (impossible back in the 60ies and still really hard even with modern computers) and the production of the coil is not simple either. You can actually include "producability" as an optimization goal along with plasma performance in your design code, but even then you will have to build a large number of different coil designs.

3.) Active control. It takes some time for the plasma donut to bend and twist. Typically a few milliseconds. So if you stick a large number of sensors and computer controlled coils around the plasma you might be able to continuously keep the plasma confined, just like balancing a pencil on its tip. This was of course utterly unimaginable back in the 60ies, and even today there is mayor problems. Sensors aren't fast enough, optimal (or even good) control algorithms are unknown and rapidly ramping megaamperes in the control coils is hard if you don't want to rip them out of the device accidentally. And you probably only have ~ 10 failed attempts before the wall of the vacuum device is compromised and you need a new device. Consequently there is some small scale research on that (sorry I don't have a cool link handy), but nobody is trying that on large devices for now.

[1] https://en.wikipedia.org/wiki/Tokamak

[2] https://en.wikipedia.org/wiki/Stellarator

Very well explained, thank you!

I work at HSX and “produceability” is really a set of three or four constraints. Picking their relative weights in optimizers is still in the “art” category because we simply don’t have the computational resources needed to search all spaces with all simulated plasma parameters. A flux surface is optimized with the preferred model of the day, then coils are optimized to this surface. A few good candidates are chosen then the realized flux surfaces are calculated and plasma parameters simulated. It’s very far from the ideal plug-n-chug that it could be, but these models are also incomplete and benefit from people understanding them. We simply need people tending the machines if we want to make a better stellarator.

On that front: keep an ear out for an HTS stellarator in the next decade.

Not the guy that asked the question, but thank you for this - very informative :)

Some years ago, likely around 2010, I had the priviledge to visit Culham Centre for Fusion Energy (it was called UKAEA Culham back then). During my visit, we were shown around various experiments and after a long evening, we had the opportunity to attend a Q&A with the staff. The question everyone wanted to know the answer to was When will it be ready? The scientists and everyone else involved honestly thought 10 years would be more than enough time for the technology to mature and find its way into everyday lives.

Commercial fusion energry, always 5-10 years away.

About 15 years ago my physics professor told me fusion would always be 20 years away, so at least we're getting progressively shorter times away.

Reminds me a little of this xkcd on the James Webb delays: https://xkcd.com/2014/

People underestimate time. I'll do something very HN and compare with software. When did Git get really good? Finished, mainstream? It took a good ten years for it to be really established. And that's just for a simple hack of an idea, that needed this time to be fleshed out.

Well, they were not exactly wrong. Humanity as a collective would have been able to push the technology to a usable level in that timeframe, the knowledge for that was certainly there. But it would have required spending quite a lot more money on it. The incentives for that were simply not there, and that remains still true today.

With the current level of investment and committment, it's more like 40-50 years for the first real generator, and another 10-20 until the first commercial one.

There's a mythical man month issue in that even with unlimited budget and resources progress can only proceed so fast, even with the best people to learn from the experiments.

However there's also just the actual literal cost of RnD. There's no hard cost in budget either.

It's more like an apples to oranges comparison, but look at the progress that SpaceX has made in commercial space flight, and that's AFTER we've established that it's possible with all of the path-finding missions in the half century prior.

Fundamental research is an investment in the future. Aside from ensuring that the basic levels of Maslow’s hierarchy of needs (1) are met, funding research that expands the public infrastructure and public domain of knowledge of how the world works, and third designs for common industrial infrastructure, such as power generation.

Commercial fusion would be either much closer to reality or known to be infeasible with current technology if research were funded both adequately and predictably (so people could make careers and lives in that).

(1): https://en.wikipedia.org/wiki/Maslow%27s_hierarchy_of_needs

I have such mixed feelings about ITER. I hope we learn the things we want to from it, and that what we learn is good news. But ITER is very far from being a demonstration of a practical power plant -- it's a science experiment first and foremost. (Well, it has really been an experiment in international government cooperation...)

I'm concerned that fusion research could have an 'AI winter' if there are any problems with ITER. (Similar to what has happened to inertial fusion in the failure of NIF to achieve ignition.) On the bright side, I think Commonwealth's SPARC experiment has a chance of beating ITER to hit 'scientific' break-even (Q>1) -- although ITER should top out around Q=10-20, where SPARC is aiming for Q~4. (Q~20 is needed for a power plant.)

The issue of 'disruptions' (rapid unscheduled disassembly of the plasma) has not been solved. ITER's construction was premised on the idea that we need it to be solved, so therefore it will get solved. The situation is very similar to that of self-driving cars -- avoiding 90% of disruptions seems pretty doable, but a percent or so happen without warning.(https://fusion4freedom.com/pdfs/Disruption-Risk-poster-Wurde...)

I hope that people don't get the wrong idea about fusion from ITER. Fusion reactors don't scale down well, but they also don't have to be quite as large, slow, and expensive as ITER. Tokamaks only use about 10% of the available magnetic field pressure, which means about 100x less power density than is theoretically possible, for a given magnetic field strength. Also, ITER is limited by its superconductors to about 5T. REBCO superconductors could potentially triple that, which would increase power density 81 times. So, there is a ton of headroom to improve performance -- dealing with the outflux of power becomes the major issue, actually.

Ultimately, fusion is a long way from market still. It's hard to innovate rapidly with devices that cost billions and have life-cycles of decades. Private enterprises are pushing down on those numbers, though -- that's where I'm pinning my hopes. Given that renewables are approaching grid parity, it looks like the goalposts will start receding before fusion even achieves net power production.

Dumb question here, but how is the actual power transmitted to the grid? I get we are trying to have some sustained reaction with plasma inside a magnetic field... but that just makes heat, not electricity.

Other reactor types have water to heat up and make steam but where does this happen in a Fusion reactor? It seems as though there is no obvious "place" for there to be water to turn into steam.

Again, sorry for dumb question.

It's not a dumb question.

The answer is kinda dumb though, most working fusion reactors don't have a way to extract heat yet. But the plan is to use the same turbine (heat water into steam) technology from fission reactors. So you have cutting edge plasma physics in one part of the reactor and old school victorian-era steam turbines in the other.

So it all boils down to boiling water?

Yes, it's just a mindbogglingly complex and advanced steam engine :)

We really only have a few ways to generate electricity:

- PV solar generates electricity more or less directly

- Wind turbines use mechanical energy from the wind to spin a turbine

- Hydro converts potential energy to kinetic energy to spin a turbine

- Nearly every other form of generation in existence heats water to create steam and uses that steam to spin a turbine.

To be fair, todays steam engine are marvel of efficiency and reliability and water-cooling is a perfected science.

Almost nothing is has easy to manage and powerful as steam.

Finally! I have been following this for over 20 years and thought it would never happen.

And you were probably right!

What does HN crowd think of companies like hb11 or lppfusion which try to construct much smaller but still powerful (lpp plans for 5MW, relatively small, reactor. They also want to use a different fuel that produces very little neutrons, thus very low radiocativity. And moreover they do not need thermal part as electricity is generated by directly collecting alpha particles (and some Xrays IIRC). It _feels_ to me like ITER is inheritance from old times - trying to build the reactor the way we always did (i.e. thermal; and also using the "easiest" to ignite fuel - but that produces a lot of radiocativity compared to hB11 fuel). As someone pointed out here in comments, there are already better magnets than those used in ITER even now. Don't get me wrong, it would be sooo cool if ITER succeeded and started fusion energy generation, but it is just taking too long.

  https://www.hb11.energy/our-story https://lppfusion.com/technology/fusion-energy-generator/

Someone below mentioned the following link, in which they do a sharp comparison of fusion technologies. I am by no means an expert, but after watching that link, it appears to me that such solutions are probably unproven science: it took over 50 years to get anywhere near a Q-factor of 1 (energy in = energy out), with hundreds of technologies screened, tested and scrapped.

    https://www.youtube.com/watch?v=L0KuAx1COEk

I worked on Gene - a gyrokinetics app used to study fusion reactors - and I fondly recall helping a colleague debug one of their simulations in which they set a boundary condition incorrectly; that simulation ran for two weeks before the mistake was found. We estimated this mistake to have costed the tax-payer multiple millions of dollars _in electricity costs_ (just from the core's that were used, their wattage, and the time duration; which is a fraction of the cost of acquiring and operating a super-computer).

All our fusion models that correctly predict all the results produced by all the fusion reactors that have been built to date predict that ITER will work.

I have no idea whether these models predict that these other reactor designs will also work. What I do know is that one needs _many_ simulations to study that, and the costs of doing that feels absurdly out-of-reach for any startup. One does not only need to "verify" a design, but come up with it, optimize it, etc. as well as gaining access to the supercomputing resources or buying and maintaining their own supercomputer. That puts the initial investment already in the millions.

OTOH, a startup that checks a design that has been created, optimized, and verified in academia and whose goal is "only" to build it, would require a smaller investment, but no idea how big this investment ought to be.

Fusion scientist here. Hydrogen-boron is 500x harder than deuterium-tritium to ignite, roughly speaking. NIF couldn't even do that. I don't believe that these two groups have hit on the solution. Yes, it would be nice to have aneutronic fusion, but right now we're struggling to get deuterium-tritium to work. One step at a time.

FTA : "Iter is a collaboration between China, the European Union, India, Japan, South Korea, Russia and the US. All members share in the cost of construction."

Who contributes how much ? Because at face value this sounds like real international (that is, beyond little war games) cooperation. Sort of "mankind" project.

>Europe [The member countries of the EU, not the entire continent] is responsible for the largest portion of construction costs (45.6 percent); the remainder is shared equally by China, India, Japan, Korea, Russia and the US (9.1 percent each). The Members contribute very little monetary contribution to the project: instead, nine-tenths of contributions will be delivered to the ITER Organization in the form of completed components, systems or buildings. In this way, the scientific and industrial fabric in each Member is prepared for the step after ITER—the conception and realization of the type of prototype fusion reactor that will demonstrate industrial-scale fusion electricity within this half of the century. For all Members, the potential benefits of participation are significant: by contributing a portion of the project's costs, Members benefit from 100 percent of the scientific results and all generated intellectual property.

https://www.iter.org/proj/Countries

ITER won't be commercially viable. After ITER comes DEMO, which is itself a decades-long project. DEMO won't demonstrate electricity generation until 2048: https://en.wikipedia.org/wiki/DEMOnstration_Power_Station#Ti...

In my mind, that means we need to see deployment of Gen 4 fission reactors (https://en.wikipedia.org/wiki/Generation_IV_reactor) commercially, to bridge the gaps between a decline in coal-fired plants and the potential of renewables.

They take too long to build

There probably has to be a machine before DEMO to solidify the engineering and materials.

Also, after ITER, there will not be enough tritium to do another large machine (the tritium comes from heavy water reactors, but those have lost in the market and will be shutting down in the next couple of decades.) So a machine to just make tritium may be needed.

Looking at stories like this, you need to realize how far out of the running this technology is.

The power density of ITER (gross fusion power of the reactor divided by the volume of the reactor, not just the volume of the plasma) is 50 kW/m^3. This is horribly low, about 1/400th the power density of a PWR reactor vessel.

The power/$ is also horribly bad. The cost is going to have to come down by two orders of magnitude to start being competitive.

Fusion is an example of sunk cost thinking. The only reason we're working on it is because we had been. A clean sheet energy strategy would put very little resources into fusion.

You don’t speak for the people working on it, so your claim of sunk cost fallacy is null.

I am allowed to form my own opinion from the evidence we have.

The evidence is that fusion is a huge boondoggle that is extremely unlikely to deliver anything useful.

Asking people working in a field if it's promising is silly, since they are a biased sample. If they didn't think it was promising, they likely wouldn't be working in the field.

Indeed you are allowed to come to your own conclusions. I object to this statement.

>Fusion is an example of sunk cost thinking. The only reason we're working on it is because we had been.

This is putting forward the opinions of others. The premise is the people who stay in the field lacked some form of awareness of your arguments. They are aware and they disagree. You can’t assume your argument is accepted in the argument.

No, I'm sure the people in the field are, at some level, well aware of the arguments. They're not new, after all -- look at Lidsky from 1983. But the human power of rationalization is very powerful when your career depends on it. They get a comforting assurance that the argument was wrong, and they don't look too closely to see if the refutation was actually correct. What did Upton Sinclair say? "It is difficult to get a man to understand something when his salary depends upon his not understanding it."

If the arguments are old, how is that a point in favor of people in the field being unaware of them? There has been 50 years of people arguing against the use of resources in pursuit of this field. You still think people in a field of physics have somehow sheltered themselves from these arguments? You have had a conversation with a physicist, yes? They will shred someone making dubious claims.

Further, if you think plasma physicists would go hungry if fusion research ceased, you’re simply wrong. Nobody researches fusion for the paycheck. An industry paycheck for the equivalent quality of work is double.

> how is that a point in favor of people in the field being unaware of them?

Perhaps they are movtivated readers and deliberately don't read the argument for comprehension. Kind of like what you just did there.

You write with a healthy amount of bias. Are there some stakes you should disclose?

No stakes that I'm aware of.

Solar was 2 orders of magnitude out of the running around 1975 -- and now it's at grid parity. So maybe fusion is 45 years away? :P

You're right that the cost is going to have to come down, and the power density up, before fusion can compete. High-field superconductors are one obvious route, and using plasma configurations that make better use of the magnetic fields are another. Either one of those approaches could deliver about two orders of magnitude in power density -- and they could be combined.

>The only reason we're working on it is because we had been.

I think there's been an element of that in the way the federal program has been run, but claiming it's the only reason to pursue fusion is not defensible.

That approach doesn't work, because reactor power density becomes limited by wall loading (neutron and/or thermal), regardless of how good the plasma physics is.

It's the square cube law. At a given wall loading limit the volumetric power density is inversely proportional to the linear dimensions. And a DT fusion reactor must be meters across, due to the fixed cross section of the neutrons with wall materials. In contrast, fission fuel rods are 1 cm in diameter and are closely spaced, so the available surface area for heat transfer is much higher (neutrons in fission are also much less of a problem, since they carry a much smaller fraction of the energy output and are of much lower energy, on average.)

(I think the only hope for DT fusion is something like LINUS, where the entire first wall is thick flowing liquid lithium.)

> I think there's been an element of that in the way the federal program has been run, but claiming it's the only reason to pursue fusion is not defensible.

What is the other reason? I cannot find any other plausible justification. This is particularly true now that fission is out of the running. Fusion used to be justified by "it won't be too much more expensive than fission, but it's safer and won't run out of uranium." That argument is now pointless.

>the entire first wall is thick flowing liquid lithium

Yeah, I agree on the liquid walls. I think the optimum might be to have a thin Li first surface flow that is relatively cool, then a shell of SiC, and behind it a PbLi breeding blanket that can be at higher temperature. There's a trade-off because the first-surface can't be made too hot or the evaporation will pollute the plasma, but getting high thermodynamic efficiency means using higher coolant temperature. The thickness of the liquid first-surface plays into the lifetime of the SiC shell -- more shielding in front of it means longer life (so lower maintenance costs & higher reactor up-time, and less radwaste), but lower thermal conversion efficiency on average.

There are problems of course -- corrosion by Li, splashing of droplets into the plasma, MHD drag & pump power requirements, incompatibility of lithium with many forms of sensors & actuators for plasma control.... (OTOH, most sensors and actuators can't tolerate radiation anyway, so fancy control techniques just don't stand a chance in a reactor anyway. We need boring, stable plasma configurations that just sit there and work.)

>What is the other reason?

Some people think they can make a fortune :D

What downsides are there to fusion power, assuming it's commercially viable in a couple of decades? What negative externalities are there that could make it prohibitive? Fossil fuel has pollution etc, solar takes up lots of space and is apparently also ugly, wind energy is ugly and makes noise (according to NIMBYs).

What argument could a NIMBY person have against fusion power, besides the big building in their back yard?

Can't be miniaturized, thus expensive. Not mobile.

Any large power plant is pretty much a missile target if war breaks out, I guess.

Radioactive leaks/accidents -- they will be orders of magnitude less of a problem than fission, in the worst case, but people are innumerate, so they'll still object.

Can someone explain to me the advantages of fusion over fission? Except for the radioactive wastes and the enthusiasm of journalists I don't get it

It's mainly the waste.

Nuclear fission is a stable source of energy that does not produce carbon emissions in production. It's pretty much the answer to the climate crisis except that it produces dangerous waste that is horrendous to store and manage.

Fusion mainly avoids this :)

Fusion has been touted as avoiding waste and fuel availability problems of fission.

But the main problem of fission is that it's too damned expensive, with very large capital cost. Fusion makes this main problem worse, not better.

Also, the accident scenarios are not as bad -- there's less total radioactivity in the device, and there's no meltdown risk.

If you had a fission reactor with a power density as low as a fusion reactor, it would be incredibly safe also. All that thermal inertia, accidents would go in extreme slow motion.

You might ask why fission reactors aren't designed that way.

Interesting question. NuScale (https://en.wikipedia.org/wiki/NuScale_Power) reactor appears to have a very low power density around 1.5 MW/m^3, by my estimate (2.7m dia, 20m tall, 60MWe -> 180MWth). They still rely on having coolant (water) present in the reactor to avoid a meltdown, although it should circulate automatically by boiling & condensing.

The trick is that something like 10% of a fission reactor's thermal output continues for many hours due to decay heat, whereas with a fusion reactor that can be orders of magnitude less, with appropriate material choices.

OK but some models of gen iv of nuclear power plants can pretty much do the same. It sounds pretty dumb to me that journalists are super enthusiast about fusion but much less about gen iv (which is way more mature)

"some models of gen iv of nuclear power plants can pretty much do the same"

I'd be interested to know which fission plants make fuel recycling and waste disposal trivial, I could make a lot of money!

In reality, even the gen 4 plants that have interesting approaches to reprocessing still produce a lot of harmful radioactive waste. Not all the fuel can be recycled and the process itself is imperfect and messy.

To have a sense of scale, it would be super cool if someone could make a photo montage with some aerial pictures from https://www.iter.org/album/Media/4%20-%20Aerial and some other big known structures.

Maybe one of the Tesla Giga factory on top of the same picture (correctly scaled of course) ?

I thought Lawrence Livermore laboratory does some cutting edge experiments on fusion and I read articles about how they are hoping to get some breakthrough. Is this new thing supposed to actually help take the tech forward or achieve something we havent achieved so far?

From the about page[1]:

> […] prove the feasibility of fusion as a large-scale and carbon-free source of energy […]

[1] https://www.iter.org/proj/inafewlines

“We hope to see first plasma in five years. That will only be a short plasma - lasting a few milliseconds” this is a quote from original article.I was wondering why `few milliseconds` can make such a huge impact?What does this `few milliseconds` plasma mean?

You gotta start somewhere! The plasma discharges will eventually last several minutes, but they will 'exercise' the machine very carefully/gradually over the course of a few years to be sure that nothing breaks.

In MCF machines plasmas are made by heating gas through RF (gyrotrons) or momentum (NBI). Both gyrotrons and NBI are very expensive, complicated, high power machines with big subsystems that go into them. Bringing on all of the heaters and diagnostics take time. Wall conditioning takes time. The first plasma is before most of the necessary systems are up. It’s a demonstration that the machine is built.

Those magnets are huge.

I wonder if, though possibly solving the problem of clean energy, it will also create other problems, say, something like "geomagnetic field pollution".

Is there a good single compendium of ongoing fusion projects?

Yes, check out https://www.fusionenergybase.com/ which I've put together. List of organizations are here, https://www.fusionenergybase.com/organizations and individual projects are here, https://www.fusionenergybase.com/projects

Is Iter aiming for breakeven or net plus energy?

One objective is a Q value of greater than 10 for a moment. And a steady Q value of 5.

Source: https://en.wikipedia.org/wiki/ITER#Objectives

Maintain a fusion pulse for up to 8 minutes What does this mean?

I am not a physicist, but AFAIK the problem is to maintain the plasma in a steady state. This could mean that they will try to keep the plasma stable for 8 minutes until it collapses.

The characteristic frequencies in the plasma are much higher. Kilohertz and up. The steady state will be reached within seconds, so to the plasma 8 minutes is basically "infinity". But at that point the all the power that is leaking out of the plasma (50MW of heating times a Q of 10 implies that the plasma will output half a GW) will have heated the wall to the limit (and that is with active cooling of the walls). One could beef up the cooling to run at actual steady state (for an hour, or a day or a year), but it would not tell you much extra, so they didn't bother.

Tokamaks need a changing current in the central solenoid. So while 8 minutes is infinity for a stellarator, there are important performance benchmarks at 8 seconds, 8 minutes, 8 hours, etc. You trade pulse length for ohmic heating.

My understating of the issue is that one need to keep the plasma is the adequate state for 1. the reaction to occur (i reckon we can do that for a fraction of a microsecond or something) 2. make it self-sustaining (we're not here yet) 3. Obviously without producing an H-bomb. Neighbours might get unhappy.

> 3. Obviously without producing an H-bomb. Neighbours might get unhappy.

If you do it wrong enough the neighbors won't even know.

Tokamak-style reactors like ITER rely on a rapidly-changing magnetic field to induce an electric current in the plasma, for stability of the plasma. The magnet that provides this rapidly-changing field (the 'central solenoid') can't keep increasing in field strength indefinitely -- so there's a time limit to the life of the plasma. There are other ways to supplement the current in the plasma, which can help extend the life of the plasma indefinitely.

But this is beyond ridiculous, so many years of R&D for 8 minutes of sustained acticivty activity? That sound like a joke

How long was the first powered flight?

> But this is beyond ridiculous, so many years of R&D for 8 minutes of sustained acticivty activity? That sound like a joke

...You know the duration of the plasma pulse being 8 minutes doesn't mean they do it once and then the whole project is over and they need to start dismantling it, right? After a pulse, the plasma is discharged, you have a few minutes of rampdown and cooling, and then you can do it again. IIRC in the first experimental phase they expect to do ~30,000 pulses.

That’s like saying “spending $30,000 for a license plate holder sounds like a joke”. This machine is being built for the information needed to make real reactors. Empirically derived scaling laws, confinement physics, and divertor physics all greatly benefit from having their regimes pushed past where they have been. It’s a science machine.

No, AFAIK. That is DEMO, ITERs successor.

ITER is aiming for net-positive energy in terms of heat coming off the plasma (but the energy will just be dumped); DEMO is aiming to actually generate electrical power.

ITER is nothing more than a money sink to underfund innovative research in fusion power.

Big Oil ️ ITER.

So many fusion sceptics in the comments that it sounds like a congregation of anti-vaxxers. Funding for fusion has been ridiculously low for decades and it's exactly because of sceptics with myopathy like the ones present here.

Countries could've been pumping trillions into fusion and other clean energy, together, for years, but we'd rather vote for politicians willing to bomb the oil out of a poorer nation than think more than a decade or more ahead. Now that climate change is finally becoming more of a reality, the myopathic sceptics are turning into temporal sceptics "but we won't have enough time". People like them create the issues we have and every time a solution is proposed, they'll cast doubt.

I'm glad things are moving along at all. I say stop oil, petrol, gas subsidies in every country, and get out of countries you all shouldn't be sticking your noses in. Invest in something that will actually let your grandchildren inhabit a planet with less friction and a much smaller threat to their way of life (climate change, overfishing, plastic pollution, etc.). Stop being so selfish.

The equation of fusion skeptics to anti-vaxxers is an unacceptable slur. We have excellent and incontrovertible evidence of the efficacy and value of vaccines. We have no evidence that fusion can be made practical, and there are well understood arguments for why it will not be.

Fusion is more like the Emperor's New Energy Source, an exercise in groupthink and sunk costing.

Fusion + AI = no more human labor

And always perpetually 20 years away.

Harnessing the perpetual motion required for it to always be 20 years away, is the key to fusion being able to generate infinite energy at a low cost. It's the most challenging aspect.

8 light minutes actually.

> The facility could see plasma generated in the machine - a notional start to operations - shortly after the assembly phase ends in 2025.

So you're saying it's... 5 years away?

The generation of plasma is approximately 5 years away.

Around 2035 the first experiments with tritium will me made, at least that's the expectation.

And even then ITER isn't designed to be a usable reactor, it's just a prototype. They intend to build a real one afterwards, and even that one will be a demonstration reactor.

With all this talk of the colossal ITER project merely being a warm-up for the even bigger DEMO one to follow which, in its turn will be but a pre-cursor to the real thing, I can't help being reminded of Douglas Adams's 'Deep Thought'

“There is no problem,” said Deep Thought with magnificent ringing tones. “I am simply the second greatest computer in the Universe of Space and Time.”

“But the second?” insisted Lunkwill. “Why do you keep saying the second? You’re surely not thinking of the Multicorticoid Perspicutron Titan Muller are you? Or the Pondermatic? Or the . . .”

Contemptuous lights flashed across the computer’s console.

“I spare not a single unit of thought on these cybernetic simpletons!” he boomed. “I speak of none but the computer that is to come after me!”

Fook was losing patience. He pushed his notebook aside and muttered, “I think this is getting needlessly messianic.”

“You know nothing of future time,” pronounced Deep Thought, “and yet in my teeming circuitry I can navigate the infinite delta streams of future probability and see that there must one day come a computer whose merest operational parameters I am not worthy to calculate, but which it will be my fate eventually to design.”

P != NP will be proven before we have a breakeven fusion.

The elephant in the room: fusion is no better wrt radioactive waste than fission. ITER is projected to end up with 30000 tons of radioactive waste.

https://thebulletin.org/2018/02/iter-is-a-showcase-for-the-d...

"Surrounding the ITER tokamak, a monstrous concrete cylinder 3.5 meters thick, 30 meters in diameter and 30 meters tall called the bioshield will prevent X-rays, gamma rays and stray neutrons from reaching the outside world. "

That article is quite outdated, JET has 2020 D-T campaign.

More importantly, fusion is a thousand times better than fission with respect to waste.

No one is claiming there is no waste at all, but tritium has a half-life of 12.5 years so most of the waste will be safe within a few decades. The rest will largely be neutron embrittled steels/materials (classified as low level nuclear waste) that is easily stored.

Compare and contrast that to nuclear fission which generates high level waste, spent nuclear fuel rods and radioactive liquid which must be painstakingly petrified and processed before safe storage. Even then this waste will still be dangerous for over 100,000 years.

So the waste output/power output is 1000 times better? Any sources pls?

And this "Neutron embrittlement" does not create any dangerous long-lived isotopes? Like, we just wait few years and then normally landfill the concrete or how should I imagine the process?

Fusion creates much less long term radioactivity.

However, the mass of material contaminated with radioactivity is very high, since the radioactivity that is created is not concentrated in fuel elements, as it is in a fission reactor.

Fission creates neutron embrittled materials and steels as well. And yes we tend to store them until they can be safely disposed.

Again, the main difference is the fuel. Tritium is safe in a few decades, spent uranium is safe in a few hundred thousand years.

Reply on news.ycombinator.com