Hacker News Reader

Comment by wcarss | Hacker News Reader

Serious question: how does an article like this get written without mentioning SPARC[1] once? They're hardly secretive, they're MIT's people, they're moving on a plan to get results in ~5 years, and they're talking about the same kind of things: tokamaks, better computer modeling, and easier to make higher temperature superconductors for stronger magnets.

I'm just a layperson, I only know pop science level stuff here, but they seem clearly like the best game in town -- so, seriously, can a person in the physics world shed some light on why they aren't at least a small part of every article about fusion, but ITER is?

1 - https://en.m.wikipedia.org/wiki/SPARC_(tokamak)

(And, for those with an interest who still haven't seen Zach Hartwig's absolutely excellent 2017 talk: https://youtu.be/L0KuAx1COEk -- watch it!)

View on HN

Because SPARC is already happening. This is planning for research in areas that the next reactor will need including a neutron source to test materials and a blanket test facility.

See Bob Mumgaard's slides from last year's annual meeting of Fusion Power Associates where he outlines the research for technologies needed for by all the approaches to fusion power.

http://www.firefusionpower.org/FPA19_Speed_Mumgaard_CFS.pdf

Quote from the last slide: "It would be a travesty for somebody to have a concept that works and have fusion stall because we, CHOSE not to prioritize the parts that help everybody."

This is where government funded research helps the emerging fusion power industry.

God I wish we could Manhattan Project this shit. I'm a layman too but I get the impression that the theory is basically there and we just need to figure out the engineering. That's a solvable problem with lots of money and lots of teams.

Are we going to wait until the sea level rises 20 feet and then say, geez, compared to the costs this imposed, fusion would have been easy?

Fission is a solved engineering and science problem though and we’re still blocking it to let the sea levels rise. Point being, nobody actually cares about climate change enough to make hard decisions so a Manhattan style project is highly unlikely.

Yes, I've been saying it over and over: raise the price of oil until gasoline et al are $20+/gallon. Then nuclear will be cheap.

Why are we letting a silly thing like scarcity, supply, and demand dictate whether we self-destruct?

I've been saying for a long time that the US should increase the tax on gas steadily every year. Even a nickel a year would be OK but more would be better.

People would be able to plan on the price increasing every year and make sound economic decisions on whether they wanted to move closer to work, etc.

Unfortunately, it became apparent about the same time that the US is almost incapable of doing something in an organized fashion. People would scream about the price of gas going up and it would be rolled back. Or there'd be subsidies that would negate the tax increases.

Fuel in the US is already much more expensive than two decades ago. I remember buying a gallon for around 90 cents in 1999, and at the same time I read a newspaper article about some immigrant-run stations on the Jersey shore that had slashed their prices to several cents below that, as they were content to make a tiny profit margin and play a long game. Now a gallon costs more than twice as much as then, but as far as I can tell the average American has not made significant lifestyle changes with regard to vehicle usage: they just suck it up.

$0.90 in 1999 is $1.41 today. Near me, gas is ~$2 (give or take). So, more expensive (~40-50%), but not double.

Granted, it may not be quite double, but average American salaries have not entirely kept pace with inflation in the last twenty years, so fuel expenses are appreciably higher than simply counting the inflation in.

Here it's closer to $2 / liter. (1/4 of a Gallon).

Sure, but your "here" isn't in the US.

Indeed, which is why I provided that datapoint.

0.90 is also an outlier for 1999. The average price was closer to $1.20 back then.

...and I remember paying almost $4/gallon, around 10 years ago (I think), with no expectation that the price would drop. Gas is crazy cheap rift now.

Most fossil-based electric generation is not oil; it's either natural gas (reasonably clean, though it does produce a lot of CO₂), or coal (terrible on many counts).

I'd say that killing coal would have a large effect. It would also put entire town populations out of work — a really hard problem, read about how UK tackled it in 1970s and 1980s, it was painful and ugly.

The death of UK coal was all to do with economics of importing coal and labour prices and relations, rather than climate change. But also that's why most green parties talk of a "just transition", subsidy to the workers who will lose their jobs.

It was also to do with cleaner energy generation and taking advantage of natural gas resources in North Sea. Dash to Gas was a major enabler of the switch away from coal in the 90s.

At the end of 2016 there were 50,000 coal miners in the US.[1] Their average annual pay is $29K.[2] We could give each of them one million dollars, invested and paying out 3% annually, and the $50B it costs would be a drop in the bucket compared to the economic benefit of eliminating those carbon emissions.

In fact, it would even be cheaper than the healthcare costs of pollution from coal plants, which amounts to $187 billion in the U.S.[3] And in Appalachia, healthcare costs from coal production are $75B annually, so we'd come out ahead there too, which would be a double benefit for the miners since they'd be both richer and healthier.

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

[2] https://www.glassdoor.com/Salaries/coal-miner-salary-SRCH_KO...

[3] https://e360.yale.edu/digest/coal-costs-us-half-trillion-ann...

Yeah, but that's a logical, pragmatic solution that actually looks at the cost-benefit from a scientifically informed perspective. Most (or at least half) of America would hate it.

> Then nuclear will be cheap.

But it still won't be built, because Solar PV and wind are roughly two-thirds the cost and falling. (Including the cost of backup power supplies for periods of low sun or low wind.)

(Not to mention having much shorter construction times until first revenue, lower delay risk, lower investment risk, and lower operation and maintenance costs.)

Nuclear had its chance in the 1990s. Today, we have cheaper and better options.

Is there a way to make people care?

Personally, I think, for United States it does make perfect sense to redirect, say, half of the military budget onto solving real homeland security issues: climate change, sustainability, ecology, pandemics. There might be enough budget there to address climate change [1].

It'd be a risk to spend less on military, but considering that the United States has nuclear weapons and allies, is it a risk really?

[1] https://www.nature.com/articles/d41586-020-02460-9

The real risk is politicians not getting their kickbacks from defense suppliers, sadly. The military industrial complex that Eisenhower warned us about has a death-grip on the congressional budget, even against the wishes of Pentagon leaders.

Politics, NIMBY, risks of accidents, and radioactive materials pollution are unsolved problems. You could argue it’s sociology than engineering but a problem nonetheless.

I think nukes is the future but nukes belong in space; just build some fission on high orbit to perfect fusion, and bring the meat down over microwave or as bulk hydrocarbons or whatever that works.

Solar system has unlimited free meal supplies of raw materials on first-come-first-serve basis. It’s been said that extraterrestrially constructed spacecraft can be launched for free merely by gently pushing it out of dockyard, or at most shooting out of a maglev train, for at least half a century.

So build some gigantic fission satellites in space!

> You could argue it’s sociology than engineering but a problem nonetheless.

It can be both when the consent is manufactured! :)

Why do fission in orbit when you have unimpeded access to a huge natural fusion plant, and virtually unlimited real estate to harvest its energy?

People underestimate what 21% efficient means when it comes to solar panels... If you bring a solar panel to space and receive sunlight 24/7 at 44% higher intensity than the best place on earth you will generate an obscene amount of electricity.

A month or so ago I was listening to an NPR interview about insects. The interviewee said that the insect population in Germany has declined something like 75% in 25 years. He noted that people don't tend to notice big changes over the span of a couple decades, which is a good point, and helps explain why climate change will probably never be seriously tackled.

Edit- link to article: https://www.theguardian.com/environment/2017/oct/18/warning-...

It will be tackled through adaptation. Some coastal cities may die. Others may have to spend a ton on remediation. More people will migrate inland, and farming centers will shift.

It’s far too late to stop it with prevention, though there is still value in trying to cut CO2 when possible to blunt it a bit.

> Are we going to wait until the sea level rises 20 feet and then say, geez, compared to the costs this imposed, fusion would have been easy?

Well, yeah, the US spent the last 50 years (e.g. "WTF Happened In 1971?"[0]) widening the income gap and doing everything possible to keep their "undesirable" ethnic groups as poor as possible. When the bottom falls out, the only people left standing will be the people who would have been able to use the nice water fountain back when America was "Great", and anyone suffering will get individual blame for their lack of individual bootstraps.

e: In case I was unclear, I think this is a bad thing :)

[0]: https://wtfhappenedin1971.com/, but I would ask "WTF Happened in 1968?" instead.

>God I wish we could Manhattan Project this shit.

Interesting point of reference: the Manhattan project cost around around $25 billion in today's dollars. The US government spent at least $6 trillion on the coronavirus response, which could buy 200 Manhattan Projects. Coronavirus will kill around a million people this year at most; air pollution kills around five million per year (and sea level rises could kill a lot more).

Aren't you implying, intentionally or not, that a "Manhattan Project" level of expenditure, $25 billion, is enough to make meaningful progress towards commercial fusion energy production? I don't think that is a given.

The implication is rather that we’ve grown much more inept when it comes to getting things done.

Because there isn't an existential war going on forcing people to listen to the people who want to be economically efficient.

Yeah it’s not like we’re in any kind of emergency that required reorganizing the whole society and economy, or anything like that.

Or because all the low hanging fruit is plucked.

That's roughly the total cost of ITER.

According to David Edgerton's The Shock of the Old, Brigadier-General Leslie Groves had previously overseen the construction of several munitions plants costing much more than the entire cost of the Manhattan Engineering District project.

1. Chapter 8, "Invention", p198 in my paperback edition.

Things sure were cheap back then. ITER alone is expected to cost at least 22 billion Euros.

In my opinion, the main reason for success of the Manhattan Project wasn't the budget, but getting people such as Feynman, Fermi, Teller, Oppenheimer, Bohr, Szilard and Wheeler to work on the project together with a sense of urgency.

25 billion USD buys a lot of necessary stuff, but one Feynman doing his best to crack a secret of nature is priceless.

Before the 10x programmers there were 10x physicists and the Manhattan project had a lot of those

And as opposed to the current advertisement websites with auxiliary functions, I believe they had a more physics centered role

That budget was essential. A hundred Feymans would have been useless without those thousands of engineers and workers who built and operated the plants Kringle Washington and Tennessee that produced fissionable enough fissionable material.

Interestingly enough, the most expensive program was The B-29 bomber program, which cost 50% more than the Manhattan project.

Then one B-29 had an emergency landing in Sovjet and the soviets reversed engineered it into the TU-4. Tu-94 which is still flying is a newer scaled up version of this.

It's amazing how much of the innovation the U.S did in the 1940s ended up in others hands(the bomb, B-29). The world would probably be very different if it hadn't happened.

The physicists knew an atomic bomb was possible the US just did it fist. I am sure secrets got out that helped the Soviets create one but I think they would've created one regardless.

History says that that's not true and the Soviets relied on extensive networks of spies to get tech secrets, be it the bomb or so many other military projects.

The US copied a lot from Germany too. https://www.businessinsider.com/6-things-us-stole-from-germa...

Jets and Rockets were built by the Germans first.

https://www.bbc.com/future/article/20160201-the-wwii-flying-...

The US medical system wastes enough each year ($950B) to start an LHC sized project (~$10B), fully funded, in each US state every year and have enough left over to fully fund three new Apollo scale set of missions ($150B). You can't recover all that, but still, it's hard to picture just how big some of these figures are until you compare to other large costs.

That number is shocking. I'd love to know details. Can you provide a source please?

Just the US value from https://en.wikipedia.org/wiki/List_of_countries_by_total_hea... vs #2 (Switzerland) is $US800 000 000 000 for all US residents in extra spending.

I'm confused by these data.

It's common knowledge that the US spends much more per capita on Healthcare, however the commonly offered explanation for this is that the private healthcare system in the US is less efficient than the public systems in Europe. But the graph in the provided link shows that US costs are split evenly between private spending and public spending.

This is confusing because the private system in the US covers twice as many people as the public system. Per the US Census Bureau: "in 2018, private health insurance coverage continued to be more prevalent than public coverage, covering 67.3 percent of the population and 34.4 percent of the population, respectively."[0] If the provided data are correct, this would indicate that US private spending is nearly twice as efficient as public spending.

What am I missing?

[0]https://www.census.gov/library/publications/2019/demo/p60-26...

The current laws restrict the government healthcare programs from negotiating on e.g. drug prices, so the manufacturers name whatever absurd prices they want and the US Gov just pays it. Just another corrupt government giveaway to the private sector.

The gov't portion includes basically everyone over 65 via Medicare, who have a disproportionate share of healthcare needs.

It's reviewed occasionally, here's a few links

This was the original article I saw when it was $700B: https://www.theatlantic.com/health/archive/2012/09/how-the-u...

More recent one, I thought it was $950B but this puts it at $760-935B. The potential savings are lower but still in the low hundreds of billions.

https://jamanetwork.com/journals/jama/fullarticle/2752664

I originally started thinking about "how do we understand large numbers" when I saw the article as I think I had roughly the same emotional reaction as I would have done if the number was $70B. Most "big number" comparisons go to things like "dollar bills up to the moon" or "swimming pool full of X" which only gets across "this is a big number". I found comparing it to other enormous projects was hard even as if you take out the US contribution to the LHC you don't move the needle, so you need to then cover all countries costs, then the full project budget for all years and you're still left with almost the entire figure left.

A lot of the Manhattan Project's spending was to scale the processes to industrial level in order to produce what was necessary to build bombs.

The situation is quite different with nuclear fusion.

To be fair, it did scale for fusion bombs.

Its amazing how cutting emission by regulation is completely left out of the list of options, somehow spending infinite money is seen as easier than getting congress to act...

Anyway, I let Wellerstein reply to this:

"""The Manhattan Project was an unusual, somewhat dubious enterprise that had massive, world-affecting consequences. Ignoring that not only misunderstands the Manhattan Project, it misunderstands what happens when you pour essentially unlimited resources into a given field — which actually is the primary goal of those who use this metaphor.

The problem is, the Manhattan Project “worked,” if by “worked” you mean, “produced atomic bombs for use in the Pacific theatre during World War II.” It almost didn’t work — there are plenty of reasons to believe that the war would have been over fairly soon with or without the bombs (the main historical question is not whether it would soon end, but on what terms and at what costs)."""

http://blog.nuclearsecrecy.com/2012/04/02/do-we-want-another...

It's widely believed that the use of the bombs in WWII as a practical demonstration of the technology is what lead to there not being a WWIII and generally a period of (granted, uneasy) world stability throughout the cold war. Thanks to the bomb's use at the end of an almost-finished conflict, everyone was too afraid of the destruction to turn the cold war hot. Without it, the cold war would have turned hot and a lot more nukes would have been used. Who knows what the world would look like today without those two bombs.

Love too incinerate 150,000 civilians instead of some random atoll with media floating in a boat a few miles away

This is just one of many lacklustre attempts at justifying a hideous war crime. From Eisenhower:

"I was against it on two counts. First, the Japanese were ready to surrender, and it wasn't necessary to hit them with that awful thing. Second, I hated to see our country be the first to use such a weapon."

I think the USSR was typically very keen to avoid a war for obvious reasons: they had a much smaller economy, exhausted population, significantly smaller military and massively smaller industrial base than the West.

America was more keen, but the logistics of actually invading the USSR would make any war essentially unwinnable.

Maybe if the nukes didn't exist, the USSR would have annexed western Berlin, but I don't really know how they would have managed more than that, even if they had wanted to.

I also don't think nuclear weapons would have prevented a war if the USSR was more aggressive or more close to the USA in military power.

See wikipedia's list of common misconceptions for a demonstration of the explanatory power of things that are widely believed. Nuclear peace is a theory that find limited support for it's more conservative claims: as the single basis for the "long peace"* it is plainly lacking.

*So called because intra-state conflict outside of NATO doesn't count.

Cutting emissions by regulation is a continually losing battle.

You can't tell people to live a lower-quality life and expect to win elections, and you'll have the combined lobbyists of a lot of different interests against you.

It's actually more tractable to invent fusion than to try and win that battle long-term -- at least it's theoretically possible to make fusion work. And heck, if we could put some $ in pockets while doing it, that's something that could pass!

The established solution is you don't tell them right out, you just make it worse for most while ensuring there are a few success stories to distract with. See for instance: war on drugs, war on terror et. al.

Yeah, but those horrors made people rich. Cutting emissions makes nobody rich.

Developing fusion, carbon recapture schemes, we need to be thinking in terms of boondoggles that get the job done.

The Clean Air Act is a clear refutation of your assertion. Benefits there exceeded costs by a factor of about 40.

Fixing market failures increases economic wellbeing.

Fusion is not going to solve the sea level rise problem.

You can see this by comparing it to fission.

Fission fuels is not expensive on a kWh base, so cheaper fuel is not really a big advantage.

Fusion plants produce radioactive waste too (perhaps less?), so you don't save yourself from that headache. Which isn't actually that big of a problem either, even if the costs are probably underestimated for fission plants.

Now, the real killer for fission is the cost of the plants. They are just so incredibly complex and expensive that they are not competitive.

Will a fusion plant be cheaper? Probably not. It will most likely, with current tech, be much more expensive. So it won't be able to compete with fission. And fission can't compete (in most parts of the world) with renewables, that are still falling exponentially in price.

So fusion may still be interesting from a scientific stand point, and perhaps things will change in the far future.

An argument I often see is that renewables need storage, and that's true. But so does a fission or a fusion plant, unless you overbuild and accept a bad capacity factor. Some amount of overbuild combined with storage looks like the most economical solution currently - the specifics depend very much on where you are in the world.

Fusion does not create any long-lived radioactive waste.

No, but fast neutrons are created in some processes which induce long lived radiation in reactor materials though.

100 years is not “long lived” in the context of radioactive waste disposal.

http://www.acamedia.info/sciences/J_G/fusion.html

Fusion will likely produce a larger volume of waste contaminated enough to prevent easy disposal. That could very well make the waste more expensive to deal with.

Wouldn't the major differentiator be the safety standards?

You can't have a runaway reaction, because you need energy input to the magnet field to get more energy out. Instead of creating the right environment to get a chain reaction started if your control rods end up in the wrong place. Drop the magnet field and it all collapses.

Even if you lose control of it with the magnet field on the quantity is limited, tokamaks can only sustain bursts before going unstable. Wendelstein like designs would also be limited by the quantity contained.

If everything goes completely wrong you simply end up with a dirty bomb from the neutron bombarded core material and an initial burst of radiation. Sure not amazing but not a meltdown releasing vast quantities of heavy radioactive decay materials which can leach into the environment. Remember, fission is messy, it's not a chemical reaction where A+B = C. You get a spread of materials and energies, as seen by the Z number here. [0]

Essentially, you can at any point do an unsafe abort which might damage something or if everything aligns it's no more than a dirty bomb, but you will never lose control. Your Swiss cheese model needs to have far fewer and simpler layers compared to fission, vastly bringing at least those costs down.

I'm still betting on renewables for the near future, but my guess is at least harnessing fusion will have a place for projects with extremely specific goals in the 50-100 year timescale.

[0]: https://upload.wikimedia.org/wikipedia/commons/thumb/6/68/Th...

Fusion will also be pretty much mandatory for deep space exploration and space colony ships, if we ever get that far

No, it won't. It is inferior to fission, solar, and beamed power.

Fusion will still require a containment building because of tritium and because of the pressure from volatilized cryogenic coolants. The building will also require very strict control of tritium escape in normal operation. Remember, a 1 DW(e) DT reactor will burn enough tritium in one year to contaminate 2 months total flow of the Mississippi river above the legal limit for drinking water.

Fusion will also require highly reliable equipement, just like fission. Not because of safety, but because the fusion reactor will be complicated and very difficult to repair. The reactor itself, even the magnets, will be irradiated and activated beyond the point where hands on maintenance could be performed.

We can make fission plants that are also immune to run away reactions. Even with old designs, you have to do something really stupid like have nightshift take out all your control rods to get a run away reaction.

While the fusion reaction itself isn't very dangerous, the magnetic coils in a fusion reactor can potentially quench, causing them to explode, which because of their position will throw radioactive debris from the irradiated reactor vessel all over the place, as well as your tritium breeding blanket which will be highly flammable, toxic, and a bit radioactive. Worse, you can't make the magnets passively-safe.

Then there are the standard issues like tritium release. And you have nuclear proliferation concerns as a fusion reactor is great at making plutonium by just doping the tritium breeding blanket with some natural uranium. In fact, the first "fusion" reactors will probably do this anyways as they need to produce more tritium to get more reactors online and because this dramatically increases power output - so you get all the fun of dealing with fission products too.

I would be willing to bet complying with the safety standards for fusion will in fact be more expensive than for fission. Yes the general population doesn't have the same irrational fear of fusion that it has of fission, but once people start seriously proposing to put these in people's backyards that will likely change.

Fission is too expensive because everyone tried to get economies of scale at big complex custom-built one-off plants. Those who see hope for fission look to get economies of scale with a modular approach, mass-producing simple, small, reactors, shipping them from the factory to the site on trucks, and attaching them to the grid.

I am not qualified to assess whether that hope is well-founded, but there are real differences in the approach.

Big nuclear reactors are cheaper on the $/MW metric than small ones. Big steam power plants are also cheaper on the $/MW than small ones.

It's not that clear how much economy of scale is there with a modular design and if it's enough to displace the diseconomy of making smaller plants. But the one thing that gets cheaper with a small size is safety mechanisms, and those are a big thing.

Fission was a fundamentally easier problem than fusion. The equivalent of ignition was achieved in 1942 after one year of effort.

As a technical answer: this is about the federal program, and SPARC isn't part of it, while ITER is.

More practically, Commonwealth & other private fusion efforts have been advocating for (A) federal funds to be allocated to solving the thorny nuclear materials science issues common to (essentially) all fusion schemes (B) a NASA-COTS-inspired cost-share approach to building the eventual pilot plant. (Presumably the industry players would like to get federal support for building their pilot plants (like ARC), much the way SpaceX got support for Falcon 9 / Crew Dragon development.)

This plan reflects (A) for sure, and a step in the direction of (B). I was involved in the early stages of the community input to the plan. I think there's openness to increased partnership with industry like this.

The article image is of ARC, the predecessor to SPARC.

Well, SPARC is a mini version of ARC. Smaller and without a blanket.

I am pro-nuclear, and I think it would be a mistake to assume this article is also pro-nuclear. Heck, the very first six words ("U.S. fusion scientists, notorious for squabbling") prime a reader to see those hardworking people as bickering nerds working on fleeting, individual, theoretical interests.

I sincerely doubt we will see any uranium-based energy technology ascend to dominance/ubiquity while over half of Earth's known uranium deposits are in the Afghanistan/Kazakhstan/Ukraine region alone. Nobody would want to buy crude (priced exclusively in USD in most markets) if nuclear were widespread. We might have to lay off a few Stuxnet malware developers then too. Takin' their jobs :p

[0]: https://en.wikipedia.org/wiki/List_of_countries_by_uranium_p... [1]: https://en.wikipedia.org/wiki/List_of_countries_by_uranium_r...

Fusion is not a uranium based energy technology.

You are talking about fission, not fusion.

Nuclear plants are using radioactive sources like uranium to generate energy. Fusion tech would use hydrogen and fuse it to helium.

Yep, and it is not just feasible but easy to build fission plants right now and stop using greenhouse-gas emitters entirely, but we won't, because Reasons. Our collective focus is being distracted from that working, clean, but politically-disadvantageous tech to this non-functional speculative future-tech that the author simultaneously insults.

People love to post comments like this on every thread where nuclear comes up, but it really distorts the reality of the situation.

Investment in nuclear has essentially evaporated. This is because fission plants take billions of dollars and around a decade to stand up. If you go through the numbers the ROI looks pretty ugly. Meanwhile you can look at the trend over time in levelized energy costs and storage costs, and can see that we're very near the thresholds for simply doing it all with renewables plus storage. All together this makes it clear to any investor doing the most basic of diligence that fission is a risky bet. Even if you waved a magic wand and eliminated any form of political opposition, that doesn't change this less than favorable cost picture.

> Even if you waved a magic wand and eliminated any form of political opposition, that doesn't change this less than favorable cost picture.

The cost is the manifestation of the political opposition.

No, it is not. It's intrinsic to the technology. You see the exact same unpleasant numbers in entirely state funded fission, including everything China is currently building. CCP certainly isn't paying the same costs out of "politics."

I want to be clear: I'm generally pro nuclear. I'm just exhausted by the smug "scared idiots and politics ruins fission" when the situation is considerably more complex.

Reply on news.ycombinator.com