Hacker News Reader

Comment by tahoeskibum | Hacker News Reader

I think that solar/wind + battery (e.g. Tesla Megapack) will make fusion irrelevant.

No one is giving you numbers for why this is wrong, so let me ballpark.

Typical solar irradiance is 1000W/m^2. Mean sunshine hours where I live is 3000hrs/y. Assuming 100% efficiency, 100% solar irradiance, perfect angle etc for those 3000hours, we get ~10^10J of energy. Those are bad assumptions, but very generous.

In a gram of deuterium, per e=mc^2, we get ~10^12J. There is 100x more energy in a single gram of deterium than a year of intense sunshine. Deterium is far from rare -- 1 part in ~10,000. Consider the vast amounts of deterium in the oceans and realize if fusion becomes a reality, nuclear, solar, oil, wind etc all become irrelevant overnight.

I recommend this https://dothemath.ucsd.edu/2012/01/nuclear-fusion/

for a layman's introduction of the energy scales involved here.

If energy density were proportional to cost, nuclear fission would already be dominating by a factor of 2,000,000x over fracked gas and coal. The reality is that complexities of systems needed in energy conversions end up coming into play. Getting net energy out of fusion was done in the 1950s (explosively, but still). Even given sustained controlled fusion, you still have to build a power plant around it and 'break even' on costs.

Right. The energy density argument is obviously wrong, and I marvel that people can still make it with a straight face.

The comment is mostly about energy scales -- Getting into the actual conversion rate, system efficiency, amortization etc is beyond the scope of a comment, especially when it's still an area of very active research.

This ignores the enormous cost involved in building and maintaining the machines to fuse that deuterium and capture the resultant energy.

You might end up being right, but I think it's actually very far from clear that fusion will ever win on cost.

> In a gram of deuterium, per e=mc^2, we get ~10^12J

Deuterium is not antimatter, where they entire mass of the substance is converted to energy. In nuclear reactions only a very small percentage of matter (~1%) is converted to energy.

So by your estimate (~1%), a gram of deuterium can provide power equivalent to a year of intense sunshine?

That statement doesn't make any sense without a unit for the amount of sunshine.

> year of intense sunshine

per square meter.

Both are fair small quantities.

From the Wikipedia, D+T fusion leads to 17.6MeV, while their total energy is around 4.5GeV, so 1% is overstating it a few times.

Also (too late to edit my previous comment), this assumes fusion converts all of the deuterium mass to energy, but actually it's only the change in nuclear binding energy between D + T and n + He-4, which is 17.6 MeV per reaction, or 2.8e-12 Joules (note minus sign). So this assessment above seems to be off by about a factor of 10^24, or 1 trillion trillion.

To get all the energy out of a deuteron, you need an anti-deuteron. This is dramatically less practical than thermonuclear fusion because we have no credible source of meaningful amounts of anti-matter fuel.

The linked ucsd page is pretty solid.

Miniaturized fusion (I know, a long way from this -- but so were computers) would revolutionize everything.

Including transit, shipping, airlines, and other industries where batteries won't quite cut it (not to mention deep-space exploration).

Unless it's ready to replace ~all of our fossil energy within the next 20-30 years (ideally less) it's not good enough to stop climate change.

Solar+wind alone is also quite unlikely to stop climate change in that time frame and enough nuclear capacity will take decades to build. By all means keep converting to to cleaner energy but also build up the sea defenses and improve both building construction (better HVAC to deal with higher average temparature) and city infrastructure (better storm sewers to deal with predicted stronger storms and better electricity grid to deal with higher peak loads from HVAC and electric vehicle charging).

As far as we know, it's not possible to miniaturize fusion. The mass of the shielding alone for a hypothetical miniature DT fusion reactor would prevent it from being used in anything smaller than a large ship. 'Advanced fuels' with reduced neutron radiation are orders of magnitude below DT in power density for given reactor performance levels, so miniaturizing them is even more in the realm of science fiction.

Still very useful for space travel, colonies, perhaps even cargo vessels.

DT fusion is bad for space travel because the reactors are too large. It has no advantage in space over fission.

I still think weather does not allow this triplet to cover all scenarios. There are times in the PNW where you won't see the sun for 2-3 weeks, long enough to deplete almost any grid storage solution

While not due to cloud coverage, this is why Texas runs super large transmission lines from West Texas wind farms to Houston oil refineries.

Can you elaborate on this point please? I don't understand the analogy.

Refining oil takes power... and if you use (cheap) clean power to convert the rough hydrocarbons into refined gas fuel or other higher end hydrocarbons... you're making it pollute less (viewed from the lifecycle) and cheaper. So in a way, you're storing excess solar power in the gas, via reducing the need to burn gas to make more gas.

Generation doesn't have to be near consumption. If weather is bad in the PNW power can be move to it from other parts of the country. Texas does this with massive transmission lines that where built crisscrossing the state in the late 90's and 00's. These transmission lines allowed for a massive expansion in wind and solar without large subsidies because the electricity could be moved to wear it was needed and sold on the open market (the state operates an real time auction market to balance to help balance supply and demand). Today the state has more wind and solar than coal. Something like 30-40% of the states electricity comes from renewables.

> Generation doesn't have to be near consumption*

* As long as the generation and consumption both take place under the same government and the transmission infrastructure is safe from disruption.

So many plans about theoretical continental grids in Europe (with the extra mile of them being powered by solar in Africa) ignore the political reality that no government is going to give up significant chunks of the energy security of their country to something ouside their control.

West Texas has a lot of empty desert or near-desert land, not a lot of people, and not a lot of industry. But it does have a lot of wind.

Houston is a big city. Its metro population is about 7 million. And Houston has a lot of industry, probably the most industrial of the major Texas cities.

Texas is big, and depending on what part of West Texas you're talking about, Houston is pretty far away from it. For example, it's a 350 mile (~550km) drive between Abilene and Houston, and it's a 600 mile (~950km) drive between Amarillo and Houston.

After lots of wind generation got built in West Texas, they had to add more power lines to bring the power to the more populated part of the state. And they did, and it worked.

Well, maybe. We can probably meet our current energy needs with solar/wind and some kind of energy storage and more long-distance transmission lines, but it won't be cheap or easy. (Maybe it'll be cheap in the long run, but in the short run it's a lot of infrastructure to build out.)

Fusion shouldn't be considered an essential prerequisite to address climate change, but it would be better to have it as an option than not. Being able to generate constant power reliably is very convenient.

Solar is the cheapest source of power in a wide range of markets. Storage will become economical long before fusion at this point.

Also storage isn't even a concern until solar/wind reaches far greater penetration.

Won't batteries run into the Lithium problem? I heard somewhere that the global Lithium reserves allow us to convert the entire car fleet of humanity to electric, but not anything beyond that, certainly not allowing large scale battery deployments that would make a dent.

There’s more lithium in earths crust than Lead yet we mine ~50x as much lead. The only limitation on mining lithium is how concentrated is making it more or less expensive to extract. Currently raw Lithium is extremely cheap, it’s highly refined battery grade lithium that’s somewhat expensive.

Rank 33: https://en.wikipedia.org/wiki/Abundance_of_elements_in_Earth...

Abundance in the crust is one thing, ability to mine it is another. That's why you need to look at reserves [0] instead of resources. And there are way more lead reserves (88 million tons in 2016 [1]) than lithium ones (17 million tons [2]).

[0]: https://newpacificmetals.com/mining-101/mineral-estimates-re...

[1]: https://en.wikipedia.org/wiki/Lead#On_Earth

[2]: https://en.wikipedia.org/wiki/Lithium#Reserves

Proven reserves are dependent on actual mining.

“Worldwide lithium resources identified by USGS started to increase in 2017 owing to continuing exploration. Identified resources in 2016, 2017, 2018, 2019 and 2020 were 41, 47, 54, 62 and 80 million tonnes, respectively.” https://en.wikipedia.org/wiki/Lithium#Reserves

There is little to suggest this trend will suddenly stop unless we simply stop looking. Further, a 5x increase in lithium has minimal impact on the long term EV or grid storage economics. That’s a lot of wiggle room for more expensive extraction.

Not sure if you've seen my edited version of the comment or replied to the earlier version. There is a distinction between resources and reserves.

That’s my point, resources are probable sources of economically extractable resources. It’s mostly a long term vs short term question, so limiting things to reserves vastly underestimates availability in 2040+.

Even resources is an underestimation as again it’s based on economic assumptions that are quite flexible. New technology for example can change what’s considered economically feasible.

Because a single molecule of lead is significantly heavier than a molecule of lithium, that’s technically an order of magnitude more lithium. 2.22E12 moles of lithium vs. 3.85E11 moles of lead.

I think at the moment cobalt is more of a constraining resource (though not all rechargeable battery chemistries require cobalt).

The 'best' ones currently do, which is the challenge. To wit; LiFePO4 (I don't know how to do subscript on HN, sorry) chemistries aren't exotic, and tend to be more 'stable'. But, you lose about 14% capacity per volume. [0]

[0] - https://en.wikipedia.org/wiki/Lithium_iron_phosphate_battery...

There are concerted misinformation campaigns out there against anything that upsets the existing, and lazy, powers that control fossil fuels.

One of my favorite debunkers of these terrible rumors and studies is Auke Hoestra. One of his more recent debunkings, in addition to being very wrong on the facts, was funded by a shell of Aston Martin, and shilled without criticism by several "journalists" in UK papers:

https://www.linkedin.com/pulse/astongate-fake-emission-figur...

Those in the renewable industry are too busy building to mount much of a defense. Perhaps once they've become rich and indolent they can start acting this way.

Also, as a side note, a typical electric car battery is enough to power the typical US house for three days. Three days of storage is probably more than we will need to get to 100% renewable energy. The very idea of converting cars to electric entails an amount of batteries that most skeptics will say is impossible. Yet we know it won't be that hard, as new reserves are developed all the time.

From what I know, wind/solar cannot power industrial needs, only residential. So we can get power at home to heat our house and watch netflix, but there will be no TVs or heaters to run off of the power.

Wut?

Industry uses three-phase power unlike residential (USA perspective anyway). You'd need one of these for your solar farm: https://en.wikipedia.org/wiki/Three-phase_micro-inverter

Industry often goes to great lengths to rectify 3-phase power. In some cases they'd much rather just have the DC anyway. But that's neither here nor there since utility-scale solar power feeds the AC grid and its alternating electrons are completely fungible and indistinguishable from other electrons.

Not enough output.

There are utility scale solar plants being planned/built that produce over a gigawatt of power. They rival the output of the largest power plants today. They can absolutely run industrial facilities as well as homes.

That's easily fixed by building more turbines and panels.

If only we had a way to combine the output from many power sources. Maybe we could call it a "grid".

Reply on news.ycombinator.com