Gravity Battery Concept | Hacker News Reader

Gravity Battery Concept

gravitybattery.info

67 pointsvrepsys72 comments

At larger scales, these are called pumped-storage hydroelectric plants. There are currently more than 100 GW of pumped-storage capacity in the world, with efficiency ranging from 70% to 87% [1]. They are, in fact, some of the largest batteries we've ever built.

They store electricity by pumping thousands of tons of water uphill when demand is low, and letting it fall back down past a bunch of turbines when demand is high. Water is much easier to handle than a solid block of steel, and it's much more scalable as well. You just need a hill and some water, possibly an already existing reservoir. Pumps can be turned on and off almost instantly to meet fluctuating demand. There's one about 10 minutes' drive from where I live. It's marvelous, and the two artificial lakes (one at the top, one at the bottom) also make nice parks for the public to enjoy.

Since pumped-storage plants seem to work so well, I wonder if there will be any need to install smaller versions in each home. It's probably going to be difficult to match the efficiency of much larger units. Maybe these will be more useful as backup batteries.

[1] https://en.wikipedia.org/wiki/Pumped-storage_hydroelectricit...

I suspect that the big problem with this concept (vs. the hydroelectric example) is going to be the amounts of energy involved. Let's suppose the shaft is 500m deep and a 100kg battery -- potential energy is simply F * z, so F = 980N and z is 500, which gives you 490kJ. A wH is 3.6 kJ (you'll need ~10-20 of these to run an iPad for an hour). You're going to need a really big weight and a really big shaft.

This shouldn't be surprising -- with a fairly small motor and good gearing you could raise a remarkably large weight a remarkably long distance with relatively little energy.

Also note that with a hydroelectric dam, the storage mechanism also happens to be the power generation mechanism (dam + turbine) so you aren't incurring huge additional construction and maintenance costs relative to just building the power source, so even if you replaced the weights with a giant tank of water, the entire storage mechanism is extra capital investment and maintenance on top of the solar panels.

If the fundamental problem you're trying to solve is baseline power, it's probably more efficient to bring power in from somewhere the sun is shining (8000 miles away, say):

"As of 1980, the longest cost-effective distance for Direct Current transmission was determined to be 7,000 km (4,300 mi). For Alternating Current it was 4,000 km (2,500 mi), though all transmission lines in use today are substantially shorter than this." (Wikipedia)

You only need a large shaft if you construct the weight out of some material other than the matter from a neutron star.

Yeah, the problem is simply practical - digging a shaft like that and keeping it from filling with water is expensive.

Having a shaft full of water would reduce the efficiency, but as long as the weight still sinks, it will still work.

Corrosion might be a greater concern.

A wH is 3.6 kJ (you'll need ~10-20 of these to run an iPad for an hour)

I'm pretty sure a whole iPad battery only contains ~7 Wh (1,900 mAh * 3.7v) and it can run for 10 hours.

Edit: OK I was looking at completely the wrong device for these numbers. Thanks miahi!

if an ipad has a 40Wh battery and can run for 10 hours then it consumes 4Wh per hour so runs at 4W or 4J/s. 100kg dropping 1m releases 1kJ so will supply an ipad for 1000/4 = 250s = 4min. so a 100kg weight needs to drop 15m to power an ipad for an hour.

i don't completely understand the parent comment - what are "these"? but you do not need multiple weights and shafts to run an ipad for an hour. a single 100kg weight and a 15m shaft is sufficient.

"these" are the Wh's

Also chargers and transformers are not perfectly efficient. Chances are you aren't transmitting power from this gizmo to your house at USB voltage (horribly inefficient) so it's probably being sent at a higher voltage (110V seems the obvious option) which means converting it back down using an power brick, so the Wh capacity of the battery understates the power requirement for running the iPad by the inefficiency of the power brick.

iPad 2 had a 25Wh battery, 3rd gen has a 43 Wh battery (11500 mAh * 3.7V). An iPhone has a ~5Wh battery.

The less energy you use to lift the weight, the more time it takes to move that weight; if it takes you 10 times as long to generate the energy as it does to use it, then you're going to need massive generation capacity to get anything appreciable happening in a reasonable time frame.

You're going to need a really big weight and a really big shaft.

Which means you're going to need a motor with a lot of torque to lift that weight, which means a lot of input current from your power source, right? Or, if your input is small, then you need a lot of time.

There is no point in having the ability to store energy in the battery faster than your solar panels can generate it, or to discharge energy from the battery faster than your house is using it. So it seems that you would be interested in the power rating of the motor, such that for a small system a 10 HP motor would be fine.

The torque and speed of the process are less relevant, as you'll just gear the motor to match the load (and possibly to keep the motor at peak efficiency).

Pump mercury [or a safer dense liquid]?

If you still want a weight system then have a vessel as a weight and fill/empty it at the top/bottom of the drop and use a counterweight to return the vessel. Then you only need a small pump instead.

Cool idea, but what's actually a safer dense liquid? Maybe a very clever colloidal suspension.

Yes! Mercury! Why didn't I think of that?

It makes my felt lovely though ... have I introduced you to my friend the hare?

Or gears.

How do gears help? The potential energy capacity of the system is the weight (mass x gravity) multiplied by the altitude. Gears only impact the power input/output. Use gears on the way in and you can "charge" the battery with very little power (but it takes a long time). Use gears on the way out and you can discharge the battery very slowly, but get very little power out of it.

Which means time.

There is a proof of concept using gravel instead of water as well.

http://www.energycache.com/ http://www.youtube.com/watch?v=G3nz_kU604s&feature=youtu.be

No-one is denying the concept would work -- it's just an enormously expensive way to save remarkably little energy.

A Macbook Pro battery -- 63.5 Wh -- can store as much energy as a 46kg mass suspended in a 500m shaft operating with 100% efficiency. Now, which do you think costs more to build and maintain?

The only drawback I see with pumped-storage hydroelectric plants is that there could loss of stored energy due to evaporation.

I don't know how much loss does it amount to on a sunny day.

it could also rain in your pond and give you free energy.

This is in fact the entire principle of hydro-electric power (incidentally, a form of solar power since the water-cycle is driven by heat from the sun).

Another drawback is that building a reservoir on top of a mountain can be a little trickier than originally thought...

http://en.wikipedia.org/wiki/Taum_Sauk_Hydroelectric_Power_S...

Many pumped-storage plants just dam an existing valley to create the upper reservoir. Easier than trying to build a gigantic swimming pool on top of a mountain, and probably safer as well.

Of course, in order to do this, you need some very rough terrain, so it's not suitable for flat countries.

Apparently, there's a "Renaissance for Pumped Storage in Europe"[1]

[1] http://www.hydroworld.com/articles/print/volume-19/issue-3/a...

I'm really sick of hearing about this idea. At least this one doesn't make the same mistake of suggesting it should be human powered, but the energy density of gravity is ridiculously low compared to practically any other technology we have. It works for hydroelectric plants because they have HUGE reservoirs to supply them.

Assume one of these gravity batteries uses a 100m deep shaft with a counterweight the mass of a Cadillac Escalade. The energy stored is 100m * 2700kg * 9.8m/s^2 = 2.6 MJ.

The first deep-cycle battery I found through google (retail price $260) is 90Ah * 12V * (assume 80% discharge cycle) = 3.1 MJ. It just doesn't add up!

The problem is the poor density: power stored is just mass * gravity * height. So if we want to store 1kWh, using a hole 500m deep (about the max for elevator cables, which seems analogous) this would need a 750kg weight.

Storing the same amount of energy in a lead-acid battery would only take 21kg, a LiFePO battery only ~10kg. And those don't require digging out a 500m hole, or the supporting equipment to winch a car up and down a skyscraper.

Has anyone worked out how much energy this could actually theoretically store given the number of weights as n, the mass of each weight as m, and the height of each well as h?

My physics is a little rusty, and I'm sure someone will come up with an answer before I figure it out.

EDIT:

If my math is right (using this [1] as reference),

  E = m * g * h (J)

gives the energy E in joules. 1 watt hour is 3600 joules, so:

  E = m * g * h / 3600 (Wh)

So, if this system were made up of 4 x 200kg weights suspended over a 50m well, it would hold

  E = 4 * 200 * 9.81 * 50 / 3600 = 109 Wh

109 Wh. That's hardly enough to run a few high-efficiency light-bulbs for an hour. I don't mean to be a naysayer, but this doesn't seem very efficient at small scales.

[1]: http://physics.stackexchange.com/questions/39281/needed-ener...

It's just mass * gravity * height.

So, for instance, 1000 kg with a working height differential of 1000 meters can store a theoretical maximum of 2.72 kilowatt hours (9.8 million joules).

Yeah, that doesn't seem very efficient. I see why this works at larger scales, such as water reservoirs and such, but there's no way this would be worth the cost to install in one's backyard. Especially when the price of equivalent battery-backup systems are only a few thousand dollars.

Which is comparable to 10 kg of lithium-ion batteries - not terrible efficient.

You would have to factor the efficiency motors that raise the weights, and the generator that lowers them to get a more accurate estimate.

That's why I used a kilometer. There's little need for an accurate estimate, capacity for such a system is very expensive, either in dealing with huge masses or huge distances.

By way of comparison, a thousand dollars of lead acid batteries would have more capacity (and despite the issues with lead, such batteries are quite recyclable...).

Similar concept is already in use with hydroelectric power generation. It's called pumped-storage hydroelectricity: https://en.wikipedia.org/wiki/Pumped-storage_hydroelectricit...

Indeed, I was about to point this out. I visited the Drankensberg one[1] in South Africa a few years back. It's amazing how big the complex was for such a simple concept (if memory serves me correctly the turbines were 50 or so stories underground).

I'm very interested to see if graphene supercapacitors may eventually have a role to play here. Using gravity seems a bit "primitive".

[1]: http://en.wikipedia.org/wiki/Drakensberg_Pumped_Storage_Sche...

And it seems to work even better submerged: http://phys.org/news/2013-05-storage-power-seabed.html

This is a terrible idea. It would cost a fortune to build a "battery" that stores no more energy than a simple lead-acid deep cycle battery.

About the only advantage is that with simple maintenance, this system should last indefinitely, while lead-acid batteries have a limited lifespan.

(Oh, and didn't web-sites that are nothing but one giant image go out of fashion around the turn of the century?)

In Johannesburg we have a whole lot of deep mine shafts that are no longer being used, and one of the main ideas for reusing them is to turn them into hydroelectric power stations. During the night when the electricity is cheaper, the water will be pumped up to the top, and during peak hours the water will be allowed to fall back down and provide power if necessary. It's a similar concept but on a much larger scale.

Are they really that big? Stored-hydro stations are efficient because of lake-sized reservoirs, and it takes uncountable miles of tunnels to have the same volume as a medium-sized lake.

I (like many people) have had this idea an energy storage mechanism. Unfortunately gravity batteries have an extremely low energy density for weight [1] or volume and so are only really a valid choice for something that operates at the scale of a hydroelectric dam.

Heat batteries make more sense for energy storage at a household level, whether it's heating or cooling. They're smaller, can require almost no maintenance and have a much higher energy density.

[1] https://en.wikipedia.org/wiki/Energy_density#Energy_densitie...

LightSail Energy system uses heat and compressed air to store energy:

http://www.lightsail.com/

Using potential energy to store electricity is nothing new. Many hydroelectric damns pump water to a high reservoir when there is excess electricity generated to store it, and let it fall to a low reservoir when more is needed.

EDIT: See wikipedia link explaining it, with examples. http://en.wikipedia.org/wiki/Hydroelectric_energy_storage

Better would be to use your whole house as the weight. And store the energy in angular momentum.. That's limitless energy storage :D

Wheeeee.....!!

This is just a large physical battery. The only way to win is if storing the energy in a weight system is more efficient or has more capacity than storing it in a chemical battery. It's certainly not as scalable as a chemical battery and requires a lot of infrastructure to pull off.

Any reasonable estimates of (investment, maintenance) costs per joule of capacity?

I'd be curious to see the immediate/longterm ROI for something like this. It seems like there would be an incredible amount of resources spent on initially building a powerhouse like the one in the picture.

There's one in England that's mainly for tea kettles after major television events, like Eastenders. They're one of the power storage options that has the quickest ability to change their output.

I've always imagined a frictionless spinning top instead of a long shaft (using permanent magnets, perhaps with some copper coils for stabilisation). You don't need a big geometry, and you can go faster and faster, up to relativistic speeds if required, without any inefficiency. (of course, it must be in a vacuum tube.)

EDIT: thanks everyone, now I know that that's what a flywheel is. Had heard the name, never found out what one was.

You mean a flywheel: http://en.wikipedia.org/wiki/Flywheel_energy_storage

They're in use for a number of applications, namely in datacenter UPS systems.

The main downside is a catastrophic failure mode. Lots of mass, spinning at high speed. Just apply imagination.

Their energy density is pretty good relative to batteries, which is to say terribad compared to hydrocarbons. Good ones are pretty expensive, and they require periodic servicing for bearings, gaskets (for vacuum sealed systems), etc. Generally limited to niche applications for now, but there are various pilot projects to look at them for train system energy recovery and grid storage.

yeah, my thought process had some pretty complicated failure systems. The thing is, with magnetic suspension, you can "catch" the thing using it's own energy. With a good and fast enough controller, you could catch the thing where it is and flood the tube with air to slow it down. since all you're doing is dissipating energy, you can use it to soften the crash.

In terms of energy density, you can just *keep increasing the speed can't you?

>In terms of energy density, you can just increasing the speed can't you? //

In which case you only need a tiny mass to start with ...

The speed limitation is presumably going to come in with the rate at which you can alter the magnetic field to still accelerate the mass. Also you'll get drag as it impinges on local fields (Earth's magnetic field) and there'll presumably be eddy currents and local electrical fields to cope with too which will become more significant at higher flux rates.

:( you're probably right, the universe always finds a way to ruin all the good ideas.

You can't increase RPM very far. The flywheel will tear apart. If you research the topic you'll find they have to use fancy composites to get to 100,000 RPM. A metal wheel would blow up by 50,000.

Train and car regenerative breaking uses flywheels. In trains they don't spin fast, but they're heavy. That's another way of increasing storage density, use a heavier wheel at a moderate speed.

Neutron star matter would be perfect.

of course, I had forgotten about centrifugal force. what a moron, sorry!

I'm not sure I'd want a quasar in my back yard, but thanks :)

For safety, they can be burried, can't they?

They've got em. High speed flywheels on magnetic bearings in vacuum. They use them in (buses)[ http://en.wikipedia.org/wiki/Gyrobus] and you can get one for your (house)[http://theenergycollective.com/tyhamilton/189396/enbridge-ma...]

Seriously, give me a link to a store and I'll buy one. Temporal's design is remarkably similar to what I had sketched out in my head.

Then, I can hook up whatever weird renewable stuff I can find/build and spread the usage out.

There was a Kickstarter to do a flywheel system like this at a household level:

http://www.kickstarter.com/projects/1340066560/velkess-energ...

This is called flywheel energy storage, if you're not familiar

http://en.wikipedia.org/wiki/Flywheel_energy_storage

also used in KERS systems in F1 cars to store braking energy.

http://en.wikipedia.org/wiki/Kinetic_energy_recovery_system

There are energy storage systems based on flywheels.

A pretty illustration of a terrible idea. Gravity just isn't that strong of a force. Electrostatic forces in chemical bonds on the other hand...

GravityLight[1] applied this concept at small scale: to power a white led in e.g. rural areas without power, a bag of dirt is suspended from a dynamo.

[1] http://www.indiegogo.com/projects/gravitylight-lighting-for-...

>(H) Generator & CPU > The generator axis gets spinned

I think you mean spun. Aside from that, very cool!

I have another idea: just throttle the hydro and nat gas plants?

I refuted a few of his points and got banned from commenting ;)

The ghost of Storm King rises again.

Reply on news.ycombinator.com