[new_realist]

Electrical grids don't store energy. If you wanted to store just 10% of the U.S. electrical generation from summer to winter using lithium-ion batteries, it would cost roughly $42 trillion dollars. Yes, trillion. Storing 10% of the U.S. daily needs for less time (say, one day) is a better story for batteries: only $115 billion.

That's how un-scalable long-term battery storage is. Hydrogen is more so: you can store hydrogen in vast quantities in salt caverns, and move it around. Pumped hydro is best, but sites are limited. Compressed air is another solid contender.

[mgfist]

You don't need to store that much energy. It's much more cost-efficient to over-provision energy sources. In a few decades we will see solar over-provisioned to 2-3x normal consumption and some battery and that will provide 100% uptime.

[u320]

The most cost-effective way is a combination of over-provisioning, batteries and hydrogen storage. The exact mix is determined by consumption patterns and local wind/solar conditions. Generally though it's extremely expensive to rely on over-provisioning+batteries alone in places with unreliable weather patterns.

Falling battery prices helps of course but we are very far from a point where we can eliminate the need for hydrogen.

[oblio]

You're overprovisioning at the wrong time with solar, though. What good is 2-3x the power output during the day when you're at 0 during the night? You'd still need 12h of storage or more realistically a mixed energy production setup.

[ncmncm]

There will be absolutely massive demand for H2, for myriad industrial processes, not limited to ammonia and hydrocarbon synthesis. So, whatever excess power you can generate will find a ready market with anybody equipped to bank H2. And, banking industrial quantities of H2 is cheap, whether you liquify and store in underground tanks, or pump into caverns under moderate pressure.

[u320]

Maybe in Arizona. In places where clouds is a thing you are looking at weeks of storage. And some of that stored energy would need to be collected months ahead of time. This is simply not doable with batteries, hydrogen storage is absolutely essential.

[oblio]

I think we're in agreement, maybe my phrasing was unclear :-)

My opinion is that we need a mix of energy sources and we need a mix of energy storage solutions.

[Tade0]

You don't need that much storage. You'd really need at most 12h of peak power, but probably closer to 4h.

A 10k km HVDC line is enough to draw power from places 90 degrees apart in terms of latitude. Similarly, between the east and west coast of the US there's around 3h of a time difference.

Chile is planning on building an even longer cable, so it's entirely in the realm of possibility:

https://www.pv-magazine.com/2021/11/15/chile-wants-to-export...

[cure]

> Electrical grids don't store energy. If you wanted to store just 10% of the U.S. electrical generation from summer to winter using lithium-ion batteries, it would cost roughly $42 trillion dollars. Yes, trillion.

Do you mean 10% of the generation on a typical summer day? Or 10% of the generation of an entire summer?

The problem with this kind of calculation is that battery technology (and cost) is very much a moving target.

LFP would probably be the better choice today. It's cheaper, safer, and can handle far more load cycles. That comes at a cost of a lower energy density, but that hardly matters for utility scale batteries. Tesla's megapacks use LFP already - https://cleantechnica.com/2021/05/11/tesla-transitions-to-lf....

Like you say, there are many other energy storage options like pumped hydro, compressed air, etc. My personal favorite is the train full of concrete that goes up and down a hill (https://interestingengineering.com/concrete-gravity-trains-m...).

[Hypx_]

10% of a whole summer.

[sdenton4]

Another contender I noticed the other day: build transport cables eight time zones long...

[Weryj]

How's the gravity storage going, it looked really promising and simple.

[dragontamer]

If you count pumped hydro, gravity storage is the #1 storage of the USA.

Turns out that pumping water up-hill and running generators when the water flows backwards is a very efficient energy storage mechanism.

[thinkcontext]

Aside from pumped hydro I'm only aware of tiny, one-off pilot projects.

[esturk]

Storage like all the cars we're planning to build anyway? Sorry, but there's something arbitrary about "summer to winter" that you're quoting to fit your narrative here. First of all, Li-ion battery wasn't made to storage energy that long. Second, there are flow batteries for that which will be far better than Hydrogen.