Why is nuclear fusion so hard? • Hacker News Board

Re: Why is nuclear fusion so hard?

#61

Post by DennisP » Sun, Dec 09, 2018, 5:14 PM UTC

kamaal wrote: post #53

mikeash wrote: post #46

Earlier quoted context omitted.

That also shows why fusion power always centers around exotic isotopes like deuterium and tritium rather than plain hydrogen. The reaction paths needed to fuse plain hydrogen are so improbable that they don’t happen fast enough to be useful.

Is this an another way of saying that the only way to fuse to Hydrogen is to create a structure as large and with gravitational features that of a star? Then a star/solar energy is the only way you will ever have large scale fusion energy generation and harvesting.

No, it's saying that large-scale fusion using hydrogen fuel requires a star, because hydrogen fusion reactions are very improbable. That reaction only happens when four nuclei collide at once.

Fortunately we have other fusion fuels that are much more likely to fuse, including deuterium, which makes up 1/2500 of the hydrogen atoms in the ocean. Most projects target the easiest fuel, deuterium-tritium; while tritium isn't naturally available on Earth, it can be bred from lithium.

Re: Why is nuclear fusion so hard?

#62

Post by xyzzyz » Sun, Dec 09, 2018, 5:21 PM UTC

sanxiyn wrote: post #34

adrianmonk wrote: post #16

Earlier quoted context omitted.

If the objection is that nuclear can't be built quickly enough, then the renewable options need to be judged against that same standard. Suppose everyone in the world already agreed to go a hundred percent renewable in 10 years. Effectively replacing the entire world's energy infrastructure is going to take a huge amount of raw materials and production capacity. Is there going to be enough? It's possible the answer i…

> There will need to be a lot of battery storage, which means there needs to be a lot of lithium. No, if we actually want this, pumped-storage hydroelectricity is a proven solution that works. There is actually no uncertainty regarding energy storage: we can do it.

There is only limited number of sites where building pumped hydro is possible, and lots of them are already taken.

Re: Why is nuclear fusion so hard?

#63

Post by curtis » Sun, Dec 09, 2018, 5:23 PM UTC

mannykannot wrote: post #41

This is somewhat tangential to the specific issue here, but I was surprised by this: At the center of the Sun, fusion power is estimated by models to be about 276.5 watts/m3. Despite its intense temperature, the peak power generating density of the core overall is similar to an active compost heap, and is lower than the power density produced by the metabolism of an adult human.[1] So practical fusion power requires…

> Despite its intense temperature, the peak power generating density of the core overall is similar to an active compost heap...

I've seen this comparison before and it also seemed quite surprising to me. But in hindsight maybe it shouldn't: The sun is already billions of years old and it hasn't burned out yet.

Another way to think of it: A larger compost pile has to radiate more heat out of the same surface area as a smaller heap because of the Square–cube law [1]. Now imagine a compost heap more than 800,000 miles in diameter.

[1] https://en.wikipedia.org/wiki/Square%E2%80%93cube_law

Re: Why is nuclear fusion so hard?

#64

Post by Fjolsvith » Sun, Dec 09, 2018, 5:24 PM UTC

pitaj wrote: post #55

mdavis6890 wrote: post #52

Earlier quoted context omitted.

Your point about political risk is very valid, but your points about the waste and cost need to be reconsidered. A big part of the cost is artificial, not inherent: regulation. A person may or may not think the regulations are a good thing, but it's not an inherent cost. The waste is not much problem. The amount of waste to power the lives of a whole family for their entire lives would easily fit in a shoebox and be…

Don't forget to mention that old waste can even be consumed in modern reactors.

Perhaps one day we'll put the waste in cars, a la "Back To The Future ".

Re: Why is nuclear fusion so hard?

#65

Post by HocusLocus » Sun, Dec 09, 2018, 5:26 PM UTC

QUOTE

"I took this fusion class when I was at Georgia Tech and I will never forget it. We started studying and I go, "Man, this is really hard." Charged particles don't want to get near each other. Bare nuclei are both charged, positive charged, they want to avoid each other.

And my professor had a really great way of putting it. "It's like going to the mini golf." He says, "You know how in mini golf you've got the volcano, and the volcano's got the hole at the very top, and you've got to putt your ball in a way that it goes all the way up the side of the volcano, and 'phwep!' falls in the hole." He goes, "OK. That's like fusion. The ball is like a nucleus, and the volcano is the scattering effect.

So any time you want to have a nucleus go to another nucleus, it scatters; it rolls up the mountain and it rolls down the side, it rolls over here, over there... and only when you just perfectly get it on the right angle does it go in the volcano." Now, the problem with fusion, he goes, "You can't steer the ball, you have to have enough temperature so that it can make it all the way up the side of the volcano and fall in, and then you have to have enough balls because you can't steer them there at the mini golf park", that's density, "and then because they're flying all over the place, you've got to make sure that there's a fence around the mini golf park so that they don't get away." That's confinement.

He said, "Those are your three things: density, temperature, and confinement, to make fusion happen."

I said, "Dude, that's really hard!" So, I came up with another analogy, "So, I guess fission would be like the mini golf park except now the volcano was flush, the hole was about this big around, the balls are going slow, and every time the ball goes in the hole, two more balls come out."

He goes, "Yeah, that's pretty good."

ENDQUOTE

Watch the whole thing. It's good.

https://www.youtube.com/watch?v=lG1YjDdI_c8

Re: Why is nuclear fusion so hard?

#66

Post by hwillis » Sun, Dec 09, 2018, 5:46 PM UTC

mdavis6890 wrote: post #58

hwillis wrote: post #44

Earlier quoted context omitted.

You're both right- technically, nuclear reactors are on of the fastest-ramping sources available. A reactor in SCRAM can drop down to However, zero reactors are capable of doing so economically. The cost of a nuclear plant is in its construction, upkeep, and staff. Those costs are the same at 110% and 30% output, but revenue isnt. In order to turn a profit a nuclear plant needs to be running at 80%+, 24/7. Ultimately…

There is a flaw in your reasoning: By definition if a plant needs to be throttled then it's surplus power generation has negative (or zero) utility/value and therefore does not contribute to it's profitability whether it's running or not. Throttling the plant doesn't reduce it's profitability, since the power it would product at that point has no value - that's why you need to throttle it.

I'm not making an argument from semantics- whether or not you approve of the above explanations, they are the reason nuclear power is commonly said to be incapable of load-following. If you want a detailed and reputable investigation, the OECD has one[1].

The inability to load-follow is not a design flaw; any reactor built in the last ~40 years can ramp an order of magnitude faster than coal or gas-steam, and will run very happily at anywhere from 50-100% power, ramping up and down multiple times per day. Many can ramp at hundreds of megawatts per minutes. In fact all of them can ramp from 100%-10% output in about a minute, since that's legally required for reactor SCRAM.

That's as fast as an average gas turbine[2] and at the high end it's faster than the fastest gas turbine. Nuclear plants are the most agile nonrenewable plants on the grid but the fact is that a nuclear plant operating > Throttling the plant doesn't reduce it's profitability, since the power it would product at that point has no value - that's why you need to throttle it.

I get what you're saying, but it's wrong. The grid is not a free market and excess generation is not legal. Electricity markets are based around contracts and producing too much is just as heavily penalized as producing too little. If you generate so much excess that you bring grid frequency or voltage out of spec, people start going to jail.

To rephrase: If a nuclear plant throttles, its costs are constant and its revenue is lower. If it doesn't throttle, its costs are higher and its revenue is lower. Nuclear sees rising revenue/kWh by throttling, but not enough. Because of all that nuclear is not profitable in short term contracts. Generally they aren't used for ancillary services either, but that's more due to conventions- for instance they are used for frequency control in France.

[1]: https://www.oecd-nea.org/ndd/reports/2011/load-following-npp...

[2]: https://www.wartsila.com/energy/learning-center/technical-co...

Re: Why is nuclear fusion so hard?

#67

Post by DennisP » Sun, Dec 09, 2018, 5:47 PM UTC

Tade0 wrote: post #37

adrianmonk wrote: post #16

Earlier quoted context omitted.

If the objection is that nuclear can't be built quickly enough, then the renewable options need to be judged against that same standard. Suppose everyone in the world already agreed to go a hundred percent renewable in 10 years. Effectively replacing the entire world's energy infrastructure is going to take a huge amount of raw materials and production capacity. Is there going to be enough? It's possible the answer i…

Are there enough lithium mines? No need to - there are 230 billion tonnes of the stuff in seawater. Just one percent of that is enough to create 25,000TWh worth of li-ion batteries - that's roughly the amount of electricity produced by the world annually. Also other chemistries like zinc-air are gaining traction and serve as a good, cheaper alternative.

Can you point to a commercial operation recovering lithium from seawater? Demand for lithium is high and expected to go much higher, so a seawater extraction plant should at least be in development if it's feasible to do that.

If it's purely a future technology, just like zinc-air batteries, then we should be allowed to look at future nuclear technologies as well, some of which are very much in development.

Re: Why is nuclear fusion so hard?

#68

Post by zipwitch » Sun, Dec 09, 2018, 5:51 PM UTC

kingkongjaffa wrote: post #50

kjar wrote: post #3

For most of my life fusion has been teased but not delivered. fission has been pimped for decades yet cannot get private funding, or build reactors in a decade, or provide solutions for the waste. Yet still the industries persist with their propaganda while renewables beat them in cost and time to operation. If I managed the funds I'd cut nuclear and redirect to wind and solar immediately.

Wow this is an insane view point, I actually thought you were going to suggest the opposite... Frankly the potential capability of nuclear far outweighs anything else. If the gen public were more on board with it, nuclear could be rolled out and be the main source of power in the world in 30-40 years, removing the need for fossil fuels for power generation. Even building them in remote locations and taking the transm…

My big problem with nuclear (fission) is that while on paper it looks like a great solution, in practice our civilizational apparatus does not appear to be up to the task of long-term responsible implementation. Fission power looks like a long game of hot potato, where everyone hopes they won't be the ones on the hook when the eventual equipment failure, natural disaster, or time to decommission eventually arrives. (And, as an observer outside the industry, building a plant while taking all those into account in advance seems to be something we are either unable or unwilling to do.)

Re: Why is nuclear fusion so hard?

#69

Post by wbl » Sun, Dec 09, 2018, 5:57 PM UTC

DennisP wrote: post #61

kamaal wrote: post #53

Earlier quoted context omitted.

Is this an another way of saying that the only way to fuse to Hydrogen is to create a structure as large and with gravitational features that of a star? Then a star/solar energy is the only way you will ever have large scale fusion energy generation and harvesting.

No, it's saying that large-scale fusion using hydrogen fuel requires a star, because hydrogen fusion reactions are very improbable. That reaction only happens when four nuclei collide at once. Fortunately we have other fusion fuels that are much more likely to fuse, including deuterium, which makes up 1/2500 of the hydrogen atoms in the ocean. Most projects target the easiest fuel, deuterium-tritium; while tritium is…

Nope! The p-p chain only has pairwise interactions. The issue is the p+p->d step is very disfavored, thanks to the weak step. D+T is all strong interaction.

Re: Why is nuclear fusion so hard?

#70

Post by ballenarosada » Sun, Dec 09, 2018, 5:57 PM UTC

Creatccount wrote: post #43

Would zero gravity be helpful in order to make it easier to get fusion working?

Good question. Under the influence of a magnetic field, charged particles do all sorts of counterintuitive things due to the Lorentz force, which acts at right angles to the particle velocity and the field.

One example is a sideways precession due to the interaction between gravitational acceleration g and the Lorentz force. Fortunately this effect is basically negligible, because gravity is so much weaker than the electromagnetic forces operating on the particles.

The difficulties tend to come from the geometry of the magnetic field itself. For example, in a torus shape, there is an unavoidable drift in the vertical direction due to the combination of centrifugal force in the big circumference, and the force implied by the gradient of the magnetic field.