Ideas on how to store energy | Hacker News Reader

Ideas on how to store energy

arstechnica.com

91 pointsrobin_reala62 comments

This is basically just a summary of the Wikipedia page:

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

... And really misses the important aspects, notably round trip efficiency, cost per kWh, and cost per KW capacity.

Yeah, Ars Technica used to have great science and technology coverage. For the past few years though, it has just been... sad. Articles either have gross, embarrassing inaccuracies, or are very shallow.

They do have decent coverage of the space and telecommunications industries though. So they aren't completely hopeless.

And she misses out on the easiest to implement: electric cars loading stations at company or city parkhouses. You can charge them, but mostly take the expensive peeks from them.

I would add gravity storage [1].

"The fundamental principle is based on the hydraulic lifting of a large rock mass. [...] Water is pumped beneath a movable rock piston, thereby lifting the rock mass. [Later] the water which is under high pressure from the rock mass, is routed to a turbine [...] and generates electricity using a generator."

"[For certain pistons] the storage capacity increases with the fourth power of the radius, r^4. The construction costs however only increase with the square of the radius, r². [...] Strictly speaking, the price per kilowatt hour of storage capacity decreases with 1/r². This is the outstanding competitive advantage of this storage concept."

===

[1] http://www.heindl-energy.com/gravity-storage/gravity-storage...

Interesting concept, but lifting rock "only" gives you a ~3x advantage over lifting water, which can be lifted much higher.

It can be lifted much higher if you have the right topography.

I also wonder about using something much denser, but still reasonably cheap (lead is what, $1/lb?)

I saw a talk of the inventor a couple of years back. The test site they where exploring they wanted to build a rock piston with a diameter of 250 and a height of 500-1000m (cannot remember exactly, but it was in this order). Since they can lift the piston to half its height that would be 250-500m altitude difference. Not sure what the limiting factor of the height of the piston is, so it might be even bigger.

For comparison the Three Gorges damn has an altitude difference of ~100m [1].

Also, it is much more compact then a traditional dam [2].

===

[1] https://en.wikipedia.org/wiki/Three_Gorges_Dam#Composition_a...

[2] For anybody wondering why ~3x: density of water 1 g/cm^3, density of rock ~ 3 g/cm^3

So...what is that in Megawatt-hours?

Lifted rock is 3X whatever the number for lifted water is. The KWH is proportional to how much was lifted and (approximately) how high it was lifted.

potential energy = HAM = Height * Acceleration(due to gravity) * Mass.

Height = 250m

Acceleration(on earth) = 9.8 m/s/s

Mass = pi * r^2 * h = pi * 125m * 125m * 500m * 3 g/cm^3 = 7.36e13

HAM = 250 * 9.8 * 7.36e13 = 1.8e17

Compared to the potential energy of lifted water mass in a dam, isn't the water also highly pressurized under that rock mass? Would that not need to be taken into account as well?

Water is very incompressible, so you don't really get anything in addition to the potential energy. If you decompress water from 300 bar (equivalent to 1 kilometer height of rock above it) to 1 bar (atmospheric pressure) it only expands by 1.5%.

I wonder what the cost would be to excavate a 250m diameter, 1km cylinder of rock versus building a large reservoir on top of a 3km mountain.

Granted, there are places in the world where it's going to be impossible to find a nearby mountain, but I'd imagine that it's similarly difficult to find places where you can dig out a stable 1km cylinder. Bedrock is not always a single hunk of rock, it can have seams that separate it into separate chunks of rock. And while in some places bedrock reaches the surface, in others it may be hundreds of meters below the surface.

you where faster than me :)

The seal is great: http://www.heindl-energy.com/gravity-storage/idea-function/s...

I just thought of a concept of using a water fall to fill a tank that gets lifted up and then when gravity makes the lift fall to the bottom, water is dumped by opening a gate. Something near the top blocks the water from refilling it while the lift moves back up. This is more efficient than moving a rock up and down, given the gate opening and water blocking is energy efficient. You can put this under hydro dams for extra energy after it passes the turbines. Why waste right?

I think I recall seeing on Hacker News a system they were building where they sent train cars filled with rock up a mountain. That seems like a good way to move a lot of mass. (I'm not sure if it is better than water...)

I read that as well... good spot being left out of this article. Here is the link: https://www.wired.com/2016/05/forget-elons-batteries-fix-gri...

I hadn't heard of / thought of the thermal storage idea (freezing ice at night - when it's more efficient both cost wise due to cheaper power, and thermodynamically due to lower outside temperature - and then using that to reduce the need for daytime cooling), and it's so obvious in retrospect that it seems like it should be everywhere. Given what a huge amount of electricity is used on air conditioning (~6% of all electricity generated in the US, apparently), this seems like it could be pretty big.

Thermal is also useful at the utility level, if you're trying to make concentrated solar work (though it can only really work okay, ultimately). The usual medium is molten salt.

The flywheels are interesting to me, especially for their use in vehicles. I wonder what the tests for catastrophic failure looked like though, especially since that Volvo engine spun theirs at 60,000rpm max. That is a lot of energy in a small package.

Interestingly, most of the energy in a moving automobile is actually stored as angular momentum in the drivetrain components, rather than what you might intuitively expect as linear kinetic energy in the mass of the frame, for example.

That is extremely weird considering the frame etc has so much more mass.

Do you have some math on that?

I would expect this to be true only below a certain speed given a certain vehicle mass.

It's obviously incorrect at 100 mile an hour with the engine idling and in neutral.

You seem to be ignoring the wheels. Lots of angular momentum there.

I just know the math indirectly, from observing that stopping distances scale linearly with rolling component mass but by the square root of the total mass of the vehicle. Relevant to modifying vehicles - if you put larger, heavier rims on your car, you must also upgrade the braking system or you're rendering it unsafe to drive by possibly doubling the stopping distances.

The number I've read is saving 1kg of rotating mass is equal to saving 4kg of chassis weight. I can't believe rotating components make up 1/5th of a car's mass though.

That's got to be averaged out over all rotation mass in the vehicle though right?

The shaft to a CV joint is going to have a helluva lot less angular momentum than a wheel.

It was a rule of thumb from a drag racing website so it's probably highly biased towards wheels. I think they did mention lightening the flywheel on the engine too, though - although that's best thought of as energy storage, rather than energy loss, since you reclaim the energy in the flywheel at the next gear change.

The failure mode of bulk energy storage (battery/fuel/etc) will almost surely be worse; and at least with flywheels once it pops it's done.

Williams F1 did the most recent relevant work on flywheel energy storage for F1 KERS Kinetic Energy Recovery Systems.

They dropped it from the original F1 application in preference to the batteries that the other F1 teams were using. A few Porache 911 GT3 cars were modified to use the flywheel. These track only cars didn't have a passenger seat, instead the flywheel was there, spinning round at tens of thousands of revolutions per minute, just next to your 'privates'.

More recently Le Mans cars have had the flywheel although nowadays the business is owned by GKN. Williams sold the business after it was not found to be what they needed for F1. Lots of London busses have the flywheel, it has found its application in stopping busses and getting them back up to speed quickly.

How much power does the flywheel give out? The peak power is about 3/5 of a Nissan Leaf. It is designed to be used in series with the main engine or in parallel. In this way the vehicle can be designed to have a much smaller engine with the flywheel taking care of all acceleration. Multiple flywheels can be used.

The applications the flywheel is sold into are specialist where things get serviced properly. I don't think it will ever be in mainstream automotive because we are giving up on ICE cars and if you have got batteries on the car anyway then you might as well use them for the regen.

Now where I think Williams went wrong was in not making a flywheel that happened to be made from some lithium-ion batteries. So that metric tonne of batteries lugged around on a Tesla, imagine how much potential energy you could store if you flew that around at 36k RPM. You could spin the thing up before leaving the house and not have to use the battery for your first few miles.

I would like to see a purely mechanical version of flywheel, so just gears and a clutch to transfer energy to the flywheel when braking and accelerating away.

Reminds me on gyrobus

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

"A gyrobus is an electric bus that uses flywheel energy storage, not overhead wires like a trolleybus."

Reminds you? It is completely explicit!

> Lots of London busses have the flywheel, it has found its application in stopping busses and getting them back up to speed quickly.

Interesting, but there's a serious challenge making a high-current electrical connection to a fast-moving part like a battery spinning at 36K RPM.

Eh, lots of energy in a small package is kind of the point. You know, like a tank of gasoline :) A catastrophic failure of that is also quite bad, but we mitigate it and live with it due to the advantages provided.

First we need excess energy to have a storage problem. Ignoring that, I think some market based mechanisms on the consumption side could gobble that capacity up.

If energy is a big enough input to some process like metallurgy, recycling, computation, etc in times of excess you could drop the price to the base mwh rate minus the cost of a theoretical amortized storage solution + transmission inefficiencies and give incentive to burn that capacity off.

Since we've already invented the economy, we can pass around excess money to use for on demand/scaling consumption instead of building big energy network storage and transportation and eating the cost.

In theory subsidizing scalable supplies with the fixed supplies.

We already need to deal with this problem every day as generation sources do not follow daily demand curve.

Money is not a very efficient way of storing energy.

I really like this idea: http://www.heindl-energy.com/media-and-newsletter/videos.htm...

A gravity storage in form of a piston carved from rock. It's lifted with water that's pumped underneath. The seal is a rolling "sleeve": http://www.heindl-energy.com/gravity-storage/idea-function/s...

I am a huge fan of the Donald Sadoway team at Ambri. http://www.ambri.com/technology/ Liquid Metal Batteries

I'd really like Ambri to succeed. It's not clear if the use of low cost materials for the storage chemistry will produce low system costs. The inexpensive chemistry appears to have led to expensive problems with seals/separators. Similar materials problems have kept sodium-sulfur battery costs high, even though the active storage materials are cheap, abundant, and offer good energy density.

I always thought it was interesting that more of these systems didn't exploit phase changes and the cold side of the Carnot cycle. For example, instead of compressing air, you can liquify it, and then when you want to recover the energy you use it as a heat sink for a heat engine (for example, solar thermal) - in theory, this is just as efficient as adding energy to the 'hot' side of the cycle, and you can expand it out through an engine as well.

I'm curious if there could be any advantages/disadvantages to using a different, heavier liquid for pumped storage other than water. My physics knowledge is quite poor, but intuitively, wouldn't using something like mercury or iodine create more pressure going down, thus generating more energy by volume, thus requiring less volume, thus making it more practical to use a tower? Surely someone's looked into this.

I think that would work, but the challenge would be in finding a denser liquid than water in quantities that would make it useful - and cost effective.

And safe. Imagine the disaster if that mercury tower started leaking.

Water is freely available. In the examples mentioned, there's typically a natural reservoir (a lake) or man-made reservoir (a retention pond) that's used. You might get a slight boost from using sea water (density of 1027 vs 1000), but I feel like the salinity might wear your equipment faster.

Also, heavy metals like mercury and iodine very are expensive compared to water. And once you start talking about using metals for storage, it might be worth exploiting state changes instead.[0]

[0]: http://www.ambri.com/

> using something like mercury

you really don't want to do that.

Is there any prospect of making superconducting magnetic energy storage more cost effective? https://en.wikipedia.org/wiki/Superconducting_magnetic_energ...

The main point of these systems is to store energy for months without much loss. Considering the cooling costs of the mechanism you mentioned, it only seems viable for more rapid usage situations ie. the middle ground between fast discharging batteries and slow discharging hydro gravity storage.

The beautiful thing about gravitational storage is that it is reversable. So the same pump that used energy to pump the water into a tank 20 stories up can also spin down as water pours through the turbines, acting as a hydroelectric generator as the water is released. Having a single reversable mechanism makes these systems pretty easy to maintain and the only places energy is lost is through friction during conversion and through leaks. They are pretty hard to beat.

The idea of deep ocean pods is similar in execution but instead of working solely against and with gravitys energy potential, the water pressure is used as well...its just more compact..but the orbs will need to be well built to withstand insane pressures.

It's missing the supercapacitor.

I think you mean Superconducting Magnetic Energy Storage (SMES). Supercapacitors are a different thing.

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

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

Rather unlikely that parent confused the two. Supercaps are nice if you don't mind trading their ~<1/5 energy density (mass, volume) of batteries for their excellent cycle count. Clearly not grid scale since batteries barely are.

1/5 takes quite a stretch, or comparing to especially low-density batteries. Maybe in a couple decades they'll be there, but my understanding is that high-end supercaps are currently 25x less energy dense than a lithium ion battery.

~9Wh/kg for the best supercaps, compared to the old (and still standard) lead-acid maxing out around 42Wh/kg. Li-ion maxes out at amazing 265Wh/kg but they are always? small and scaled up by putting many in parallel... and if you damage one it's emergency fire preparedness time.

But if you have just about any concern for density, you're not using lead-acid. Not when that's another 5x worse than other batteries.

I'm willing to bet that most lead-acid users would happily switch to cheaper supercaps if they existed.

Agreed. When supercap J/$ hits lead-acid levels the J/kg difference wont matter and most people using lead-acid will happily trade 1/4th the energy density for near infinite recharges.

Aluminum has also been considered as an energy storage material, specifically one that is easily transported: http://evworld.com/article.cfm?storyid=765

practical engineering youtube channels talked about swell pressure under buildings, I find this kind of "osmotic" work interesting. Extremely interesting.

The https://en.wikipedia.org/wiki/Trompe is an interesting device for the compressed air category.

What about storing it in fat cells? Bio chemistry storage. Is this possible/advantageous?

Not sure about fat cells, but IIRC there has been some work in using algae to generate hydrocarbons from the air. I can't really see "fat in fat cells" being better than alternatives that skip some steps, though.

How about a system that loads springs as mechanical energy, do that have those?

paging dani fong to weigh in on this....

Reply on news.ycombinator.com