Hacker News Reader

Comment by crazygringo | Hacker News Reader

> So until battery systems are cheap enough for generators to stockpile electricity for hours at a time, renewables can’t constantly provide power like coal and gas.

I thought the solution wasn't expensive batteries locally, but rather storing pumped water in reservoirs, or other solutions like melting and re-solidifying salt, etc. -- huge-scale energy storage connected to the grid. Your energy is produced once from the sun/wind, and then second (at night) from hydroelectric turbines or steam.

Anyone know how that's going? Does it have to be done at the city/region grid level, or are there micro-versions that make sense to be integrated with a single wind turbine or field of them?

View on HN

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2811038

>Based on the 2014 German hourly feed-in and consumption data for electric power, this paper studies the storage and buffering needs resulting from the volatility of wind and solar energy, focusing on a “double-structure-cum-storage strategy”. While buffering wind and solar energy jointly requires less storage capacity than buffering them separately, joint buffering requires a storage capacity of over 6,000 pumped-storage plants, which is 183 times Germany’s current capacity.

>Buffering the overshooting production spikes associated with a market share of wind and solar of 50% would require an ideal, frictionless storage volume of 2.5 TWh or a storage capacity of 2.1 TWh in ordinary pumped-storage plants. This is about seven times the entire pumpedstorage capacity currently available in western Europe, including Norway and Switzerland; and 81% of the volume that the EU’s ESTORAGE project considers as “realisable” in western Europe.

Pumped storage is not a practical solution.

Pumped storage is not a total solution. But in the areas where it works, it works very well and cheaply. Do pumped storage everywhere you can, and fill in everywhere you can't with other options. By the time we get close to fully intermittent energy sources, pumped storage won't be nearly enough. But while we still have peaker natural gas plants around, every pumped storage site helps reduce emissions.

Most places need to expand the electricity grid so that electricity can be moved from where and when it is plentiful to places that can use it.

If you have a good east-west connection then energy can be sold into the time zones east and wet of the producer.

Good north-south connections mean that hot sunny locations nearer the equator can sell to consumers nearer the poles.

This is beginning to happen in Europe with the construction of big HVDC interconnects between Norway and the UK, Denmark, Germany and many smaller connections, but there is a still a lot of work to be done to connect Portugal, Spain, France, Germany, and countries further east and south. Europe

See https://ec.europa.eu/energy/infrastructure/transparency_plat... where you can toggle between the display of completed and on going projects.

With wind power, storage is hardly required to the extent that is needed with solar.

Wind farms, especially in the UK which is my point of reference, are almost always producing something. This can be statistically modeled, so you can then say: "Building <x> MW of wind capacity will meet our needs of <y> MW <z> % of the time."

You can then decide where to be on this tradeoff curve. So over-provisioning wind will reduce the need for storage or other generation types.

This is of course not possible at all for solar, which only produces power during sunlight hours, with a peak generation between the morning and evening demand peaks, limiting its usefulness.

What if you took the solar energy and used it to heat up a thermal differential between two areas, and then you tapped into the kinetic energy from that?

Too fancy, IMO. The easiest path is demand respond: you just need to control when you put on A/Cs, refrigerators, lighting intensity, industrial manufacturing, etc. By incentivizing responsive demand, the intermittency issue becomes far less important.

In the UK, there is "economy7" pricing, where consumer electricity is cheaper during the night.

In high solar output countries, I could see the opposite happening, that you start seeing tariffs where energy is cheapest around midday when solar output is high.

Pumped hydro storage is the gold standard for large scale storage, but it is only cost effective in certain geographic areas. Many of the prime areas for pumped hydro are already being utilized in developed countries. So, we can't really implement a ton more pumped hydro.

Melting and re-solidifying salt or water is done at smaller scale, but has not been used for large scale storage. Some grocery stores use systems like this to shift cooling loads a few hours, but nothing at grid scale.

To make a serious impact we need big scalable solutions. Batteries are in the lead here since the price keeps coming down. They also have the advantage of being able to discharge at both high power and low power.

"Battery systems" encompass all of the above methods of "energy storage connected to the grid". My personal favorite "so crazy it might work" idea is stacking concrete blocks on top of each other.

Li-ion batteries keep coming down in price, and they can be scaled to basically any size. Pumped hydro is generally considered the best if you have access to it. Compressed air storage works well if you've got a huge, airtight cavern (generally an old salt mine). But any cheap energy storage we can get is gonna be useful.

There's a lot of engineering challenges to still work out, but my favorite is currently flow batteries.

Another crazy one is simply pushing a big heavy train up a hill. Then when you need power just letting it go back down the hill.

https://www.wired.com/2016/05/forget-elons-batteries-fix-gri...

I one-up that with https://heindl-energy.com/technical-concept/ (city for scale?;)

That make me think. It seems we're losing a lot of kinetic energy with rain, just letting it fall and impact the ground without capturing the force. Your example was real though, so I suppose that's an advantage over crazy rain energy.

I imagine stack of ultracapacitors formed into solid walls or even building insulation. No space issues, no moving parts, extremely long lifespan, ultra quick charging and discharging. Cons: graphene is still expensive.

There is also the idea of gravity pistons which claim to have the advantage of a much wider allowance of siting options than pumped hydro.

https://heindl-energy.com/

Nice long-form discussion with the Omega Tau podcast:

http://omegataupodcast.net/299-gravity-storage/

Also this: (don't know if there is any association)

http://www.gravitypower.net/technology-gravity-power-energy-...

I believe China is commiting a huge amount of effort to building more pumped stored power, which should be coming online in the next few years [0]. If they keep up with building out a large amount of solar, this could work very well.

0. https://en.wikipedia.org/wiki/List_of_pumped-storage_hydroel...

I guess the main problem is that pumped hydro is not that energy dense - even if you assume you have a lot of space.

Your typical 18650 li-ion battery stores around 12.25Wh - that's enough energy to lift a metric tonne of material 4.5 meters high.

For practical capacities the cost of land becomes significant.

That depends on location. There is a lot of cheap land in the US but much of it is far away from population centers.

I went through the math and indeed - in Wyoming, where land is $660 per acre, assuming water and structures holding it are free, the break even point is at 17cm in potential energy.

That's actually amazingly cheap, because the absolute worst land in my country is almost three times as expensive.

I guess then it's the structures and labour here that potentially are the deciding factor.

What's overlooked is the demand side flexibility. With a bit of statistics, your electric car might know that it's ok to wait for a better time to charge, or even return power to the grid when connected and charged. Dishwashers, air conditioning, and washing machines have some flexibility to choose when to consume energy, and there are large industrial consumers such as aluminium smelters that could also participate.

There is hydrogen too. There was talk of Germany using the excess electrickery from their coal stations to create hydrogen, which can then be used to power fuel cells to create more electrickery.. Maybe hydrogen should become a higher priority?

On the "melting and re-solidifying salt" thing, I had no idea that was a thing. For others similarly ignorant, I found this great entry on thermal energy storage. (sure, I was aware of heat exchange and thermal energy generation, but as a medium of energy storage at massive scale, that was new, and fascinating) Anyway: https://en.wikipedia.org/wiki/Thermal_energy_storage

Maybe this is a good place to ask. During off-peak hours my local utility charges $.025/Kwh versus $.25/Kwh for peak. If I charged a 11Kwh Telsa powerwall, I could easily get through the day on $.025/kwh electricity, even on hot days, and the payback would be something like 11 years. Why wouldn't this be cheaper at grid scale and why isn't it doable today?

Reply on news.ycombinator.com