[standardUser]

"..cover about 20 minutes of world electricity consumption..."

That seems like a meaningless metric. No one is designing a global battery system to power the entire planet. What matters is how battery systems integrate with generation systems.

"People assume the reducing emissions would stop climate change..."

Why do you assume people assume this?

[adrianN]

Why do you think that this is a meaningless metric. It seems to me that we need at least enough storage to last for a night. Realistically we need enough storage to cover reduced production and increased demand through winter.

[kragen]

It's cheaper under most assumptions to build enough solar panels that the reduced production during winter is still enough to cover your increased demand, rather than to build seasonal utility energy stores. See my calculations at https://news.ycombinator.com/item?id=20664835 which show that 50x panel overprovisioning is cheaper than a week’s worth of battery backup. 26 week backup would then be a higher cost than overprovisioning panels by 1300x. The actual seasonal variability in panel output is normally less than a factor of 2, nowhere near 1300. Unless you're in Antarctica or something.

[adrianN]

I don't think we have enough room for overprovisioning panels that much. The average German for example uses 48000 kWh per year of primary energy[2]. Solar panels produce around 150 Watts per square meter peak. That means you need about 35 square meter years of solar panels per person to get to net 0. Let's say 15 square meter years because primary power consumption includes power plant inefficiencies and so on that you don't have with direct electricity.

42,339 MW of PV installations produced 39,401 GWh of energy in 2017 in Germany[1], so about 38 days of peak generation.

That means you need around 400 square meters per person at typical PV efficiencies to break even. That's already a lot. Where do you want to build 50 times that? There are 230 people per square kilometer in Germany, that's 4300 square meters per person.

[1] https://en.wikipedia.org/wiki/Solar_power_in_Germany#Statist...

[2] https://www.wolframalpha.com/input/?i=(total+energy+consumpt...

[kragen]

I don't understand what you mean about “square meter years”.

Germany is an especially difficult case, being very industrial, very densely populated, and fairly polar, though less so than, say, England.

Germany uses about 72 gigawatts of electricity and about 400 gigawatts of fossil fuels, according to https://en.m.wikipedia.org/wiki/Energy_in_Germany (but converted to SI units). It comprises some 82 million humans, so this works out to about 5 kW of fossil fuels per person. (Your "48000 kWh/year" comes out to 5.5 kW.) Generally 1 W thermal is worth about 0.4 W electric—less as transport fuel, more for industrial process heat, but generally about that. (I think this might be what you're saying about power plant inefficiencies?) We can take 0.5 W to be generous and we get 2–3 kW per person.

This would come out to 12–19 square meters of panels per person, using low-cost 160 W/square-meter panels, but those aren't average watts, but peak watts. Typical capacity factors (average÷peak) for PV installations in the US are 15–30%, so you'd normally need on the order of 60 to 120 square meters to get that much power on average, but the number Wikipedia is reporting for Germany is 10.6%. Maybe that's just a function of polar latitudes but it seems pretty extreme! Maybe something else is going on there, like production curtailment due to inadequate storage resources and demand response, or plants being offline due to equipment failure?

So I think that low capacity factor already incorporates some overprovisioning in that sense.

When I mentioned "50x", that was not because that's an overprovisioning ratio that is ever actually needed. I'm sorry that was unclear! It is far in excess of what is needed. I think 10x is pretty much the limit of what you need outside of the polar circle: instead of the 3.3 peak watts per average watt you'd need in Perú, you provision 33, and enough storage to get you through the night, and then you'll be fine even if every day has storm clouds blocking 90% of the light. In Germany that would be 400–600 square meters per person. Yes, that does mean covering 10–15% of the country in solar panels. The Germans would be well advised to put some of those panels in other, more equatorial countries, with lower population densities. Or work on demand response.

I mentioned the grossly excessive 50x number because even that ridiculous level of overprovisioning is still cheaper than a week-long battery storage system. (But not if you have to invade Egypt to install it, I suppose.) My point was that battery storage is far too expensive for anything where there is any alternative. In particular it is not a viable option to deal with seasonal variability. (Some other form of energy storage might be.)

Earth's population density is only 20% that of Germany, so getting the world's human population to be as energy-intensive as Germany without any demand response, you'd need to pave "only" 2–3% of it in solar panels. Fortunately or unfortunately, that includes the sea.

[adrianN]

1 square meter year of solar panels is one square meter producing peak output 24/7 for a year, or two square meters producing peak power for half a year. Like a man-month.

I was also surprised by that low capacity factor. Before doing the math I estimated maybe 100sqm per person. The US installation averages probably benefit a lot from the deserts and being much closer to the equator. Solar is pretty bad in Germany during winter.

I agree that we most likely don't need week long battery storage btw, but mostly because of wind energy that is also available at night and during winters. We can probably get away with just over night batteries and power-to-gas for long winters.

[kragen]

The part I was confused about was that you said "35 square meter years per person" but I guess you mean per person per year? Then wouldn't that just be 35 square meters per person?

Wind is awesome, but I don't think the total wind resource at tower-like heights is sufficient to wean the humans off fossil fuels; but I don't have hard numbers. Maybe power-to-gas is a solution.

Thank you very much for helping me explore this!

[cesarb]

> It seems to me that we need at least enough storage to last for a night.

That's only the case if all your generation is from photovoltaic solar. The wind doesn't stop blowing at night, and you need a long drought for hydro to stop producing power.

[adrianN]

Yeah good, then we need enough storage to last half a night and good enough transmission lines to get wind power from wherever it's windy to where the power is needed. That's still a lot more storage than 20 minutes.