[manfredo]

Even achieving just one hour of storage globally amounts to 2.5 TWh of storage. By comparison the entire world produces ~300 GWh worth of lithium ion battery annually. That leaves geographically limited options like pumped hydroelectricity, and solutions not yet deployed at any significant scale like hydrogen fuel cells, synthetic methane, thermal batteries, flywheels, etc.

Realistically we should saturate daytime energy demand with solar, and if there aren't any scalable storage options by then switch gears and proceed with hydroelectric where it's viable and nuclear where it's not.

[Qwertious]

>By comparison the entire world produces ~300 GWh worth of lithium ion battery annually.

And this will increase a hundredfold to make EV production possible.

That means that if 10% of production goes to stationary storage then within 10 years, we'll have 10 full global hours of storage.

If there's serious demand then the supply will scale up to create it.

Also, old EV batteries will provide plenty of extra stationary storage. Not to mention batteries still in EVs, in a pinch.

Realistically we won't throw insane amounts of storage at the problem. We'll make demand more flexible so it does work when electricity is cheap and eases off when it becomes more expensive.

For instance, something like heating: why store the electricity for heating? Wouldn't it make more sense for a house to have some form of heavily-insulated thermal mass that it can massively heat when electricity is dirt cheap, then tap into at midnight without drawing power? Storing heat is cheap, you just need a giant block of concrete with solid insulation. You don't need fancy nanoscale tech like with lithium-ion.

Even something like a kettle: the hot water taps you see at companies that are pre-heated. Have a home-version. Insulate the shit out of that and do 90% of the boiling with peak electricity.

And that's not even touching industrial power usage.

Trying to ape past systems that were based on flat electricity prices just seems like a failure of imagination. Of course it would be expensive, but why the heck would you even want to?

[Symbiote]

The heater you describe is called a storage heater.

https://en.wikipedia.org/wiki/Storage_heater

[cmcnab]

I had a house for a time that had just what you describe: a concrete block thermal mass set in the foundation. There were ducts running through it as part of the forced-air system. During the days, especially sunny days, the block soaked up the excess heat and then radiated it back out during the night. It worked quite well; indoor temperatures stayed remarkably consistent. A company called Adirondack Alternate Energy were the architects. https://aaepassivesolar.com/low-energy.html

[nicoburns]

> Even achieving just one hour of storage globally amounts to 2.5 TWh of storage. By comparison the entire world produces ~300 GWh worth of lithium ion battery annually

... so if we could increase battery production by just 10x, then we could create an hours worth of storage every year. That seems... very doable.

[manfredo]

And then we'd have to continue that production for two and a half decades to get to 1 day of storage. And we'd also have to drastically increase our battery recycling capacity to match (remember most lithium ion batteries last 1000-2000 cycles).

[jeffbee]

Nobody needs 1 full day of storage.

[ch4s3]

One could imagine a series of cloudy windless days in the northern latitudes during the winter. Perhaps a large enough gird solves that problem? I have no clue.

[pfdietz]

One would not use batteries for the "rare, but prolonged" storage use case. You'd want something with lower capital cost, even if it were much less efficient. For example: hydrogen burned in turbines.

[Manfredo_1]

Hydrogen storage remains in the prototyping phase. We have no significant amount of hydrogen grid storage. Like thermal batteries or synthetic methane, hydrogen represents a potential storage solution but not one that we know will scale and be effectively deployed at the scope required.

If we actually deploy 50 GWh of hydrogen storage, and demonstrate that it can cheaply and reliability be built at scale then your point would be valid. But until then, hydrogen represents a theoretical solution not an actual solution.

[rhodozelia]

Who pays for the shadow generation system that we keep perfectly maintained and ready to generate 100% of system demand on the 5 days stretch of cloudy windless days? This cost has to be added to the cost of building a 100% solar/wind system.

Nobody is arguing the solar and wind power isn’t cheap, but the cost of power on those cloudy windless weeks is going to be real high to make having all that standby generation around. It’s the cost to achieve the same reliability and 99% carbon free that is expensive.

Money is imaginary and global warming isn’t so let’s just print some bonds or move some numbers around in some database and build it all! - an electrical power engineer

[acdha]

That’s a problem if you’re an island isolated from everyone else, and you don’t have geothermal, hydroelectric, or nuclear options. The better question is how much capacity you’d need on a national grid to be able to handle large regional sags in production without endangering people.

As we recently saw with Texas’ catastrophic fossil fuel production failures the big problem is not the source but poor management and not being able to get help from the neighbors.

[lutorm]

But the key is that if you're averaging globally, the solar power probably doesn't change much. You'll need a way to transport the energy instead, obviously.

[nicoburns]

If we had smart metering, then we could simply restrict consumption in these (presumably rare) circumstances.

[Manfredo_1]

We'd actually need 3 weeks of storage to migrate to a fully renewable grid: https://pv-magazine-usa.com/2018/03/01/12-hours-energy-stora...

[Qwertious]

From your own source:

"The solar heavy network wouldn’t need energy storage with an HVDC network."

So no, we wouldn't need that. HVDC would be far cheaper.

[Manfredo_1]

A solar heavy network would still need 12 hours of storage to accommodate nighttime energy use. More actually, because of greater seasonal fluctuations further from the equator.

All of the Americas experience night time simultaneously for at least 8 hours a day. Even if we ran HVDC lines to the Sahara, there's still a period of time where most sunlight is shining on the pacific ocean.

[thiht]

That sounds extremely expensive and not very green.

[pydry]

>Even achieving just one hour of storage globally amounts to 2.5 TWh of storage. By comparison the entire world produces ~300 GWh worth of lithium ion battery

What's the point of this comparison?

Lithium ion batteries are probably the least cost effective means of dealing with intermittency. It's also rare that the entire world is without wind and sun simultaneously.

In terms of cost:

Demand shaping < overproduction < pumped storage < < lithium ion batteries

[Manfredo_1]

"Demand shaping" is a nice euphemism for energy shortages. And if we demand shaping we're just externalizing the cost to consumers that need to buy their own energy storage or change their energy usage patterns to accommodate the unreliable supply.

Overproduction helps but doesn't eliminate intermittency. And pumped hydroelectricity is geographically dependent. The irony is that most places with extensive hydroelectric storage potential don't need wind and solar in the first place because they get their energy from hydroelectric generation.

[pydry]

>Demand shaping" is a nice euphemism for energy shortages.

It's a euphemism for storage heaters, storage air-conditioning, aluminium smelters that dial usage up and down and smart car chargers.

Lithium ion batteries are useful too, of course, but they cost more.

This is a problem where market based solutions shine. The only reason that fact isn't getting rammed down our throats by lobbyists is that the people who got religion about markets tended to be oil/gas people, who have since been thrashing the "renewables are unreliable" drum.

>Overproduction helps but doesn't eliminate intermittency.

Why should the goal be to eliminate it when we can adapt to it and thrive?

Personally, I'm more excited for applications of periodic free/-ve priced electricity than I am worried about shortages.

[crummy]

Isn't demand shaping things like discounts during certain periods? My electricity provider lets me set a 'free hour of power' every day, as long as that hour is off peak.

[Manfredo_1]

Yes, those incentives exist to try and shape demand. But in practice, people rarely take advantage of them. And some things really can't be shaped. The pumps that deliver your water cannot have their demand shaped, unless you're willing to go without running water for some hours of the day.

[pydry]

>in practice, people rarely take advantage of them

Overproduction is still not that common. These days wind and solar mostly just provide power that would have otherwise been produced by natural gas even when operating at peak capacity.

It is getting off the ground though. The UK has an energy tarriff popular with electric car owners for this reason. They can occasionally get paid to charge their cars. This type of thing will only become more common.

>And some things really can't be shaped.

Obviously not. Nonetheless pretending that all renewable intermittency has to be made up for with expensive lithium ion batteries is backwards thinking.

[panarky]

If we paid the true cost of peak power, it might be worth pumping water with off-peak power and storing it locally.

[Manfredo_1]

Then our transition to solar + wind needs to include the cost of installing a septic tank and water reservoir in every household. And a thermal battery for heating. And an electric battery for lighting. And all the other things we'll need to do to accommodate an unreliable energy grid.

[asdfasgasdgasdg]

There is always a shortage of electricity. Someone could always use more if it were free to do so. Economics is the study of the allocation of resources in the face of scarcity -- that is, all allocation of resources except perhaps breathable air. There's no need for a euphemism here because limitations on the consumption of energy are ever-present -- demand shaping is simply about making the signal stronger.

[Manfredo_1]

There are rarely shortages of electricity in the US. There were some in California during Covid, and the state had to do rolling blackouts. But no, there are rarely shortages of electricity.

Yes, someone could use more of it than we could supply. But they don't. The existing supply is sufficient to meet demand. And when demand changes, we are capable of increasing supply.

[Qwertious]

>to accommodate the unreliable supply.

You're conflating "expensive" with "unreliable". Even with infinite batteries, buying stored energy will always be more expensive than direct solar/wind.

[mgolawala]

Why would we use lithium ion batteries?

I would imagine the approach to store the energy would be to use the energy from solar panels to do work that can be used to produce electricity later.

For example, you could use solar energy to pump water back uphill to flow down through a hydro electric dam later.

Even if it isn't the most efficient, in the long run it would likely provide the best scalability and least long term environmental impact. Once you have the facility in place, the same water could be pumped uphill to flow back down a million times over with the only overhead replacing water lost through evaporation and maintaining the facility.

Am I missing something that makes such an approach unfeasible?

[kragen]

My comment at https://news.ycombinator.com/item?id=26597661 links to a number of overviews of the issue, including MacKay's chapter in which he covers pumped storage in about as much depth as you can possibly hope in a 20-page chapter aimed at a general audience. Go read it!

[manfredo]

Hydroelectric storage is geographically dependent. You need the right topography and access to water. Likewise, hydroelectric storage takes a long time to build.

[intrepidhero]

Let me introduce you to flow batteries. Lithium is a terrible choice for grid scale storage, except maybe as a secondary use of idle EVs.

https://en.m.wikipedia.org/wiki/Flow_battery

[psadri]

We could use day time electricity to generate fuels to burn at night. As long as it’s net-zero wrt to CO2 emissions, it will be fine.