[tjansen]

It's true, but when does it make sense to include plants that do not deliver reliable power?

Wind power makes a lot of sense as long as you are still using fossil fuels. Every watt generated by wind power means that you can reduce fossil fuel, and thus lower your CO2 emissions. But once you got rid of fossil fuels and you have a reliable source of power without CO2 emissions, you can get rid of the unreliable ones.

[adrianN]

Wind and solar are pretty cheap and storage keeps getting cheaper, that's why they make sense.

[jules]

Batteries are so expensive that it is unclear whether they will ever solve the large scale storage problem: yes they're getting cheaper, but they have to continue to get cheaper for a long time before they're suitable, and it's unclear whether fundamental limits will be hit before that. If battery technology improves to the extent that it becomes viable for large scale storage, then wind and solar can become our main source of energy. Until then, nuclear is the only proven solution. Betting on batteries now amounts to gambling with the planet.

[jhgb]

Batteries are not supposed to solve the large scale storage problem. They're best at solving the small scale storage problem. Recently they solved the problem of small scale storage on wheels.

> If battery technology improves to the extent that it becomes viable for large scale storage, then wind and solar can become our main source of energy.

Batteries are not the only way of storing electricity.

[craftinator]

First thought I had here was hydroelectric storage.

[jhgb]

Especially after that prospective sites research by the Australian National University (https://www.anu.edu.au/news/all-news/anu-finds-530000-potent...), that seems like a reasonable choice.

[pydry]

Battery backed solar/wind is cheaper than nuclear these days.

It's not been that way for long though. Economic grid scale batteries are here but still relatively new.

It makes sense to continue running old nuclear plants but not to build new ones. Much too expensive.

[sudosysgen]

Is it? Are you taking into account battery degradation from 1 cycle every day? The vast majority of battery chemistries won't last more than 3-4 years under those circumstances, and those that would are either much more expensive or experimental.

As of now storing 10kWh at 1kW costs around 1000$ from the cells alone. If you're changing them every 3 years then you have to spend 10 000$/kW over 30 years whereas nuclear is the same price per kW for a 30 year period.

If you don't take that into account then sure.

[jungturk]

Aren't lifetimes closer to 10+ years due to better battery management (managed operating temperature and charge/discharge)?

Tesla suggest such with its megapack https://en.wikipedia.org/wiki/Tesla_Megapack

[sudosysgen]

Only if you don't do daily ~80% discharges.

You can avoid that right now because the grid has baseload. But if it doesn't you can avoid the wear cycles.

10 years is about what you'd expect if you only discharge ~30% of capacity daily, which is how it is operating right now.

[pydry]

This suggests it's even cheaper than gas now:

https://www.spglobal.com/platts/en/market-insights/latest-ne...

And gas is a lot cheaper than nuclear.

Theres a new battery backed solar plant in california that can service the early evening peak times with cheaper electricity than coal.

Battery prices have been plummeting consistently for the last three years.

[denton-scratch]

> battery chemistries

Ah, you're focused on chemical batteries.

Hauling a lot of water up a mountain at times of low demand, and releasing it through a turbine at times of high demand, is a type of battery; it seems to me a reasonable approach to smoothing supply and demand for wind/solar.

I agree that it's going to be a long time before grid-scale chemical batteries can help much with demand-smoothing.

[sudosysgen]

I'm focused on chemical batteries because that's what's being talked about when people say battery.

Pumped hydro is promising but it's going to be as expensive as regular hydro.

[Gibbon1]

Probably.

I'll point out using existing natural gas peaking plants to make up for temporary shortfalls of solar and wind power is also a viable stop gap.

[ZeroGravitas]

The latest US plan aims for a 95% carbon free grid by 2035. They could have aimed for 100%, but it's cheaper to start electrifying more things at that point, as 95% carbon free electricity powering a heat pump is better than burning gas for heat. They therefore get the eqivalent of 105% carbon reduction for the same cost, more than they'd get by focusing on the final 5% of carbon on the grid.

[adrianN]

As far as I know, few people suggest (Lithium-)batteries for long term storage. Electrolysis, optionally followed by turning the Hydrogen into Methane, seems like a much more scalable solution. That works at scale today, it's just too expensive to make sense at this point. Then there are other types of batteries that might become much cheaper in the future, perhaps redox-flow batteries or something like that.

[p_l]

According to people I talked with, who did analysis for "Green hydrogen" as storage method, assuming Poland - we would need something along the line of 150% peak production, locally, before it started moving the needle at all - and I'm not sure of this wasn't in combination with nuclear (though limited by the idiotic free market on electricity).

All of that assumes that the demand doesn't go up... Which is not compatible with things like climate goals

[chess_buster]

Hydrogen, molten salt to drive ex-coal plants, redox-flow, …

[tjansen]

We are decades away from having enough storage to make wind and power a reliable power source. There is not even technology that would scale up enough to store a country's power for weeks or at least a few days.

China just has announced ambitious plans to install storage for 100 GWh by 2030. China's electric power generation capacity is 2200 GW (in 2020). That's not even enough to provide electricity for 5 minutes....

[ben_w]

> There is not even technology that would scale up enough to store a country's power for weeks or at least a few days.

Don’t mistake a manufacturing limit for a tech scaling limit. While it may take decades to get there, batteries could do that; in the mean time, intercontinental HVDC connections could substitute for some of that storage (not all the storage all at once unless mining increases, but certainly plausible over the scale of a decade or so and we would need that timescale to build the renewables themselves anyway)[0], and the batteries are in addition to existing pumped hydro, and even in the current “low wind” scenario the UK is still getting 3.8 GW (~11%) from wind[1][2] rather than getting nothing.

[0] https://news.ycombinator.com/item?id=28474201

[1] https://gridwatch.templar.co.uk/

[2] https://gridwatch.co.uk/demand/percent

[throwaway894345]

Do we have the lithium/etc reserves to meet the storage needs for the entire planet? Nuclear is proven and if we claim to believe that climate change is an existential threat I don’t know why we would pin all our hopes on solar and wind and some to-be-discovered storage solution. To be clear, I’m not against solar and wind—on the contrary, I want a diverse clean energy portfolio. But wasting time emitting while we pray for a storage solution for wind/solar seems utterly foolish.

[oezi]

Sorry, but Nuclear is just proven to fail. Even if we would reverse course on Nuclear today it would be 20 or 30 years until the plants would be build. By that time solar and wind will another magnitude cheaper.

The way forward is wind and solar. Everything else shouldn't be focused on.

[throwaway894345]

> Nuclear is just proven to fail ... the way forward is wind and solar

Nuclear is the only proven clean technology for base load generation. The only hiccup is political (i.e., people decided they don't like nuclear), and while it's a big political problem, the whole climate crisis is an enormous political problem. Yes, there's the waste to be disposed of, but we already have to manage some waste and once you have to safely manage a little nuclear waste it's a marginal increase in cost to manage a whole lot of nuclear waste.

Further, innovations in nuclear are making it cheaper, safer, and faster to build. Moreover, as another commenter pointed out, if we were willing to ease some of our restrictions on nuclear such that our nuclear plants didn't need to be a thousand times safer than our coal plants (but merely, say, twice as safe), then nuclear could be even less expensive and facilities built more rapidly.

Yes, wind and solar will play a major role in the future, but we incur tremendous risk by ignoring nuclear.

[foxfluff]

What's driving the price decrease in wind?

[ben_w]

> Do we have the lithium/etc reserves to meet the storage needs for the entire planet?

Yes. There are basically so many different chemistries (and non-chemical storage methods) that the important question is “which type should we prefer” rather than “can we even do it”.

[vagrantJin]

Yeah, I don't think it's an either or science decision any more than its likely a business-cum-political situation.

[nradov]

While intercontinental HVDC interconnects are technically feasible, no major world power would ever depend on those for essential power supplies. It's just too risky if foreign countries can cut off your electricity during a war or other crisis. Energy independence is strategically critical in a way that transcends economics.

[adrianN]

Then why do countries rely on foreign oil, gas, and coal all the time?

[ephbit]

Because they at least can store months worth of gas/oil/coal on their own territory.

[tjansen]

I guess a manufacturing limit is bad enough. The global battery production is expected to reach 2063 GWh/year by 2028 [0]. That wouldn't be enough to store China's electricity consumption for a single hour. The production would need an increase by several orders of magnitude. Are there enough raw materials for this? How much waste would there be, given the limited lifespan of those batteries?

What about a no-wind scenario? I don't know what wind in the UK is like, but in Germany this happens quite often. In November 2015 wind output dropped to 0.2 GW (0.5% of its 40GW power rating) [1]. Hydro doesn't help in such a scenario (<4% in Germany), nor will bio mass (<10%).

[0] https://energycentral.com/c/ec/world-battery-production

[1] https://de.wikipedia.org/wiki/Dunkelflaute (German)

[ben_w]

> The production would need an increase by several orders of magnitude.

Yes, but that doesn’t itself seem like an implausible economic shift given how large the existing fossil fuel sector is.

Challenging, sure — perhaps it is politically impossible, I wouldn’t know as I’m not at all politically astute — but physically it seems fine.

> Are there enough raw materials for this?

That part at least is fine. Earth is big, and while lithium is in the category “rare Earths”, it isn’t all that rare compared to what we need, and even if it was lithium isn’t even the only option for storage.

One of the things suggested in your [2] was long-distance HVDC to different weather zones, and Scandinavian (hydro? I’m unclear) storage. In principle we could also do antipodal HVDC (different time zone for day/night, different hemisphere for summer/winter), though on a previous thread I was encouraged to do the maths and realised the EU collectively would use a 1m^2 cross section conductor for current HVDC designs (if you wanted 100% substitution rather than it being merely part of the solution), and this will take quite a long time to mine at current rates.

> How much waste would there be, given the limited lifespan of those batteries?

No idea, but the current alternatives are “set lots of it on fire” (fossil fuels) and “bury a tiny quantity of extraordinarily dangerous stuff in scary artwork for geological timescales” (nuclear), and all it has to do is beat those.

IIRC the end-of-life batteries can be processed back into their raw material more easily than can the rocks we start with for fresh batteries.

[jhgb]

> while lithium is in the category “rare Earths”

No, it's not. Where did you get that from? Surely not from elementary school chemistry lessons, where you're taught that lithium is an alkali metal.

[adrianN]

We're also pretty far away from the kinds of renewable penetration where you actually need a lot of storage, so we have plenty of time left to build more batteries and electrolyzers.

[nradov]

Electrolysis is extremely inefficient. It's unlikely to be a practical means of grid scale energy storage any time soon.

[jhgb]

It's not just a means of energy storage; it's a method for producing a vital chemical feedstock. If your main alternative is processing natural gas, building more electrolysers is a no-brainer. You'll have to do it no matter what the efficiency, since we just don't have a better way.

[ephbit]

80 per cent is what you call extremely inefficient ... what percentage would be "efficient" then in your opinion?

> Accounting for the accepted use of the higher heat value (because inefficiency via heat can be redirected back into the system to create the steam required by the catalyst), average working efficiencies for PEM electrolysis are around 80% ... [https://en.wikipedia.org/wiki/Electrolysis_of_water#Industri...]

[adrianN]

Yeah round trip efficiencies are very bad, but at scale it's cheaper than batteries as far as I know.

[pfdietz]

The two storage modes are complementary. Batteries would be good for diurnal storage, hydrogen for longer term and rare event backup.

[ephbit]

Something makes me think that storing weeks worth of electricity isn't going to happen in the near future (< 30 years).

Storing hydrogen isn't that easy/cheap either.

So I'd guess we're going to see storage of energy in the form of liquid/liquefiable hydrocarbons (synthesized from hydrogen) like methanol or propane.

[dv_dt]

By recent performance, if we build nuclear we are decades away.

[tjansen]

Even if that's true, at least we would be betting on proven technology. What makes you think that unproven technology for storage can be built faster?

[dv_dt]

By the number of abandoned nuclear projects in the west, it’s not proven.

[korantu]

In this case 'proven' means was ever deployed at scale and worked successfully.

Nuclear has track record of decarbonizing entire industrial economy in just 10 years.

We dont have storage solution with such track record.

[autoliteInline]

>There is not even technology that would scale up enough to store a country's power for weeks or at least a few days.

My bet is that the Japanese will build some huge newfangled storage facility. There'll be a big earthquake. The storage will meltdown/burn/whatever somehow. It'll cause a great big semi-permanent problem. Everyone will declare victory and shout 'at least it wasn't nuclear'.