[cogman10]

> or rather, it’s not good enough for grid-scale requirements

I disagree with this point. LFP batteries are cheap, high density, and have huge cycle life. The big drawback of LFPs is manufacturing is just starting to ramp up on them. That is, they aren’t available.

LFPs just came out of patent protection last year and you are already starting to see them everywhere. The biggest problem with LFPs today is demand is outstripping supply.

[ac29]

Assuming LFP means Lithium Iron Phosphate (LiFePO4)?

Not sure how to reconcile "they aren't available" with "starting to see them everywhere".

They're great batteries (especially compared to lead acid) and at least in small quantities, they have been widely available for years.

[cogman10]

> at least in small quantities

This is specifically what I'm talking about. Because of some weird patent agreements, manufacturing of LFPs have been confined to China and they've not really been producing a large enough number of them to fill the role of grid storage (at least outside of china). [1]

[1] https://table.media/china/en/opinion/lfp-advantage-for-china...

[Wingy]

I read it as “they aren’t available right now, but they will be very soon.”

[Turskarama]

I can't see Lithium batteries of ANY kind ultimately being used for grid scale storage, beyond the initial pilot batteries we have going up now.

Current worldwide lithium production is at 3% of what it needs to be to electrify every car, which is a use case that has strict weight requirements. Ramping up lithium production by a factor of 30 is a big deal, and that's before we use any of it for grid storage!

Grid batteries are static so weight is not a concern, using the chemistry whose main advantage is weight for this purpose is a waste of resources. Heavy battery chemistries have largely been ignored because traditionally batteries have always been for mobile purposes, so we can expect an even better learning curve from low energy density but cheap battery technologies such as iron-air.

[coryrc]

If we're using renewables, we need seasonal shifting, so cycle life doesn't matter at one cycle per year.

You could build nuclear to supply your winter power, but then you're overbuilt for summer and don't need any renewable. Or you could store heat directly in the ground like that Alberta pilot project, heat collectors on the roofs all summer drive the heat underground, pump it back out all winter.

Or our current plan, pretend to be "green" by spending money on solar while increasing coal usage and no feasible plan to replace space heating.

[bryanlarsen]

We don't need seasonal storage. The sun still shines in the winter, unless you're in the Arctic circle. We can use over building, production diversity, interconnection and short term storage instead. Or just use natgas peakers for the last 1% and call a 99% solution good enough.

If we had reasonably priced seasonal storage we'd use it, but we don't need it.

[coryrc]

In winter cloudy conditions solar PV produces 10-15% power. Assuming some hydro storage, that's 4x overbuild. Not cost effective.

Europe all gets winter at the same time. If you've got a cold snap for three weeks with low wind, the only plan is reliance on massive fossil fuel backup. The cost of keeping that capacity for only using a week a year isn't priced into solar either.

The CO2-intensity of electricity generation in France stood at around 57 CO2/kWh in 2020 (source: Statista). In Germany, the electricity mix at the same time had a CO2-intensity of 366g CO2/kWh, which was more than six times higher

But yeah Germany's approach is really working!

[rbanffy]

> In winter cloudy conditions solar PV produces 10-15% power

Depends on the latitude and these numbers seem to be for very high ones close to the polar circles.

> Not cost effective.

Are we pricing in the societal collapse due to climate change? I’d suspect a 4x overbuild would be quite cheap against that.

OTOH, keeping some natgas capacity for when a freakish cold snap with no wind hits seems kind of OK.

[burmanm]

> Depends on the latitude and these numbers seem to be for very high ones close to the polar circles.

No, those values are far from polar circle. I'm guessing closer to central Europe, since for example in Finland the PV produces 0% during the winter months.

10-15% would be insane to get here, but there simply isn't any energy in the sun (and closer to the polar circle you get - there's no sun at all during winter) and the panels are often covered in snow in any case. And I'm not even talking about cloudy days now, but "sunny" ones.

March/October are already approaching those 10-15% levels. Nov-Feb is closer to 0% in most of the Finland.

[ben_w]

Finland (the south of it anyway) has about a gigawatt of HVDC lines connecting it to other countries, about 200W/capita.

[DennisP]

GP is clearly talking about cost-effectiveness compared to other solutions, such as France's approach and seasonal storage.

[laurencerowe]

> The cost of keeping that capacity for only using a week a year isn't priced into solar either.

Keeping gas power plants around for backup power isn’t all that expensive since fuel accounts for two thirds of their cost of generation.

Offshore wind is far more expensive than onshore wind and solar but even so costs about a third of new nuclear power, with strike prices in the UK of £37/MWh vs £106/MWh for Hinckley Point C. Maybe keeping gas backup adds another £15/MWh to that but it still works out at half the cost of nuclear.

By building more France will probably get nuclear costs down some, but even so will struggle to be competitive with renewables and backup.

> But yeah Germany's approach is really working!

Germany’s approach of keeping coal plants around while closing existing nuclear is extremely dumb.

[ben_w]

> Or just use natgas peakers for the last 1% and call a 99% solution good enough.

Sadly, no. Given how long the CO2 stays in the air, anything less than 99.9% over all emissions from all nations — and that also includes cement and iron chemistry leading directly to CO2, cattle biochemistry leading directly to methane, etc. — then we're not pushing hard enough.

Natural gas is great as a way to buy time, but we can't let it be more than a rounding error in the ultimate ensemble… well, not unless there's corresponding CO2 capture.

[concordDance]

Got any numbers here?

[Retric]

They’re cheap enough for EV’s, but the big number here is you can get 10,000+ charge/discharge cycles. Predictions are near term ~2c/kWh grid storage looks realistic which completely changes the economics of the electricity grid. (As in whatever you pay for the electricity to charge the batteries you need to charge 2 more cents/kWh at discharge to break even.)

Pair with renewables and not only is nuclear and coal looking obsolete but even natural gas is uneconomical at current prices. Of course the economy reacts as you start to bring this stuff online which means wholesale natural gas prices could fall quite a bit etc. You also need enough wind/solar/hydro to actually charge the batteries, but that doesn’t seem to be an issue either.

[cogman10]

For LFPs at the moment we are looking at 100->150/kwh of storage. [1]

Cycle life is around 3000->10000 cycles depending on how deep the cycles are (more shallow cycles, longer battery life). [2]

Note: Cycle life is about how long it takes for a battery to lose 20% of it’s capacity. That doesn’t mean you can’t still use it.

[1] https://www.fastmarkets.com/insights/the-ev-battery-chemistr...

[2] https://www.powertechsystems.eu/home/tech-corner/lithium-iro...