I've done some web searches, but I don't see anything very current on how close we are to having a home energy storage flywheel system that's comparable in price and performance to a battery system.

Also, as a bonus, what is the current state of a domestic-scale flywheel system in terms of maximum energy storage, power output, and usable energy (maximum energy minus minimum energy -- assuming there's a minimum speed they must maintain, unless there's not)?

  • 1
    As others have said - there is a lot of work on flywheel storage, but I don't think anybody is planning to deploy them in domestic settings.
    – Flyto
    Jun 19, 2017 at 13:57
  • 1
    en.m.wikipedia.org/wiki/Flywheel_energy_storage High power, short term, low capacity seem to be the main characteristics. For domestic applicartions you typically care more about capacity and the time you can keep the energy than about power. Jul 2, 2017 at 8:25
  • 2
    Velkess is a name that was going around a few years ago. The company was looking to raise about $50,000 to develop a residential flywheel energy storage system. Eventually, they send the money back because the project didn't end up being viable.
    – LShaver
    Feb 27, 2019 at 21:04

11 Answers 11


Lets check the pros and cons on flywheel energy storage and whether those apply to domestic use (source):

Compared with other ways to store electricity, FES systems have long lifetimes (lasting decades with little or no maintenance;[2] full-cycle lifetimes quoted for flywheels range from in excess of 105, up to 107, cycles of use),[5] high specific energy (100–130 W·h/kg, or 360–500 kJ/kg),[5][6] and large maximum power output. The energy efficiency (ratio of energy out per energy in) of flywheels, also known as round-trip efficiency, can be as high as 90%. Typical capacities range from 3 kWh to 133 kWh.[2] Rapid charging of a system occurs in less than 15 minutes.[7] The high specific energies often cited with flywheels can be a little misleading as commercial systems built have much lower specific energy, for example 11 W·h/kg, or 40 kJ/kg.[8]

  • long lifetime -> that would be a nice thing but not of the utmost priority. With 100,000 cycles (minimum taken from Wikipedia) it significantly outperforms rechargable battiers such as Tesla's Powerwall 1 with 5,000 cycles within warranty. Other sources report a design life time of 30 years (though for utility scale storage systems not for residential use)
  • capacity starting from 3 kWh -> A few to a few tens of kWh would put it well in a useful range for domestic use.
  • Specific energy of 11 Wh/kg (typical commercial system, value taken from Wikipedia) -> With an usable capacity of 13,5 kWh at 120 kg system weight Powerwall 2 outperforms commercial flywheels tenfold.
  • price -> While I did not find any good numbers for small home-use flywheels but I somewhat doubt they beat the 5.5k US-$ for a Powerwall 2.
  • round-trip efficiency (charge/discharge) of up to 90% -> in the range of what Powerwall claims (92.5%, source)
  • large maximum power output -> Again, no numbers for smale-scale systems but a high power throughput is not a primary concern for domestic use (7 kW peak; 5 kW continuous of the Powerwall seem quite reasonable for that use case).
  • safety -> It's hard to compare the risk from batteries (fire) to catastrophic failures of flywheels (kill'em all projectiles from a disc turning at 50,000 rpm).

In conclusion: with all the effort put into development of rechargeable batteries and upscaling of a whole industry to massproduce and market them I would doubt that flywheels are ever going to fly (pun intended) any time soon in a residential setting. I would rather expect to see them as grid energy storage in utility-scale levels that would benefit from the advantage of a high power output and where a higher capacity warrants the complexity of such systems (moving parts, vacuum, mag-lev).

  • 1
    Re specific energy: For a stationary system, energy stored per unit volume probably is more relevant that energy stored per unit mass. Jun 25, 2017 at 18:23

Domestic flywheels are unlikely to happen for 3 reasons:

  1. They must be heavy to store significant energy. If you need a crane to install one at your house it’s never going to be super cheap, even with high volume manufacture.

  2. The risk of the spinning mass fracturing requires special safety precautions - commercial operators put them in the ground but that would be super expensive on a domestic scale. Note that this isn’t the same as failed bearings where it might grind to a noisy halt. We’re talking heavy projectile at great speed.

  3. They don’t store energy for long 20-50% loss in two hours according to Wikipedia. In a typical solar powered home set up, to boil the kettle just before dawn you’d need to have stored ~13times that energy at sunset. You’d be better off having paid the extra to get it from the grid.

Now if you’re off the grid, that’s a different story. Storage Batteries have trouble with high power - short duration loads such as when your fridge compressor starts up. A very small flywheel could help here in conjunction with normal batteries. It being very small eliminates the problems of weight, safety and energy loss over time. But the right battery already does the job cheaper - a low capacity, high power battery like the one in your car (wired correctly; don’t try it!).

So for it to be cheaper it’d take a revolution in low cost, high speed bearings in tandem with a stalling in high drain battery development; and it’d still only be a very niche market.

  • 1
    Welcome to Sustainable Living! Nice answer. Could you please link to the Wikipedia page where it says that flywheels have 20-50% energy loss in 2 hours?
    – THelper
    Jun 6, 2019 at 13:06
  • Datasheet from a long term flywheel energy storage retailer shows their solution at ~86% efficient. The full details give a better view: a 32kWh storage what consumes 55W when idle and consumes 140W when charging/discharging at 8kW. For off-grid where you store the power for 20 hours at time the 55W draw will be pretty costly. But the 30 yr lifetime absolutely beats any electrochem battery.
    – deft_code
    Nov 9, 2020 at 23:23

We were looking into flywheel UPSes for my company and I read up on this a bit. Of note in recent developments:

  • Instead of traditional bearings, they're using magnetic bearings to mostly eliminate friction.

  • To be able to suspend the flywheel on magnets, that requires a lower weight/mass.

  • To counter the lower weight/mass, you have to replace that with increased velocity to maintain the same amount of momentum, which is your energy storage.

  • Supposedly these magnetically-suspended flywheels allow for a very high efficiency, as the reduced mass also means it doesn't take as much energy to spin it up. "Mass = Resistance to change in motion."

Living in a rather rural area with crappy power service, I would love to see a flywheel system just to clean up the input power and get a uniform waveform output coming into my house. I'm sitting in a room right now where the lights routinely flicker from dirty grid power.


My knee jerk reaction:

From what I know of mechanical systems, anything mechanical introduces losses due to friction in the form of heat, sound, and vibration. How would you even begin to use the mechanical energy stored in a flywheel? Simply storing the energy induces massive losses. Transferring the energy to its appliances introduces transfer problems such as losses, torquing rods, etc.

My reading after I wrote my knee-jerk reaction: Article showing state-of-art flywheel progress, with efficiency of 85%. Flywheels operate in vacuum. https://www.scientificamerican.com/article/new-flywheel-design/

Big $$ spent by Boeing on specialty-use flywheel: https://www.uaf.edu/files/acep/BoeingFlywheelOverview_06_20_2012.pdf

Afterthought: This seems to be something for specialty use. The promising flywheels seem to need uber-clean, highly maintained conditions. This is unlikely to be something for use in the home anytime soon. I expect that it is already in prototype experimentation in some factory.


TL:DR: They extend the life of expensive UPS batteries by absorbing the transients.

Example: Hosptial with an MRI Engery surge on start up with UPS graciously assisting. The MRI fires up 10 times a day = 3650 times a year. This hammering halves the rated life of UPS batteries, rated for say 100 deep discharge cycles.

In a month, you have one power outage and 90 transients- half of them comming from within the building itself. This can vary greatly by each place in the extreme. Some municipalities have thousands of transients.

So, this is where a flywheel which lasts decades can save big money: premature battery replacement. (athought: what about using a supercapacitor as a buffer)

source (page 10/13): http://www.power-thru.com/documents/The%20Truth%20About%20Batteries%20-%20POWERTHRU%20White%20Paper.pdf


I was only able to find one residential flywheel energy storage system for sale as of right now. It has a low (but manageable) power output and is priced competitively.

Unit Specs: Power: 3kW | Energy: 15kWh | Size: 40”x40”x40” | Weight: 750lbs Round Trip Efficiency: >=80% | Depth of Discharge: 100% Connectivity: 48VDC “Virtual Lead Acid” buss. Price: $6,000


They have an AC unit as well that is $1,500 more.

There are a few other companies that are getting close, Energiestro is one that comes to mind.

  • 2
    Unfortunately, per their customer letter, Velkess is no longer working on their flywheel.
    – LShaver
    Aug 2, 2022 at 15:17

TL;DR - Because of how flywheel energy storage scales it is unlikely that significant efforts will be made to develop the technology for home use.

This is similar to the case for windmills, where the power output increases as the square of the diameter, and the cube of the wind speed (which itself doubles roughly every 20m of elevation).

From the Engineering Toolbox, the equations governing flywheel kinetic energy are:

E_f = 1/2 × I × ω²
I   = k × m × r²


E_f = flywheel kinetic energy
I   = moment of intertia
ω   = angular velocity (measured in radians/second, proportional to RPM)
k   = inertial constant (a value from 0 to 1 which depends on the flywheel shape)
m   = mass of the flywheel
r   = radius of the flywheel

If we were to assume all flywheels have the same shape, roll all the constants together in to some value K, and combine the two equations, we'd get something like this:

(flywheel kinetic energy) = (K) × (RPM)² × (mass) × (radius)²

Thus to maximize the energy storage of a flywheel we would focus on making it larger (increasing the radius) and faster, as the total energy will increase proportionally to the square of these factors. Note from @Ghanima's answer we know that efficiencies are already greater than 90%, so there isn't much potential there.

  • double the radius, quadruple the energy
  • double the speed, quadruple the energy

There's a fun calculator you can use to see this in practice.

For domestic applications, making things larger is not an option -- it has to fit in a home. Thus to advance the tech for home use, designers would have to increase the speed.

Increases in speed are limited by technology costs and electricity supply -- faster motors cost more, and require a higher current input than may be available in a typical home. Given that energy potential increases for this application are already limited by size constraints, it simply doesn't make sense to put more money into the motor and the electricity supply.

  • 1
    It's not the RADIUS but the angular momentum that matters. That means you can increase the RADIUS, or increase the MASS. Most commercial flywheel systems use a drum-shape to allow the flywheel to be extruded vertically instead of increasing the radius. A flywheel storage system is also almost identical in many respects to a power generator - it's often built around a magnetic coil so the motor and the output generator are the same piece of hardware, with simple switching reversing the circuit to switch between spin-up and discharge.
    – Steve K
    Mar 28, 2019 at 1:12
  • Why would you want to fit it IN a home? For many reasons it's best to bury it, or store away from home. Also, this would be great for rural/off-grid cabins. People bury their septic and propane tanks routinely. I would love the $6,000 15KWh solution from Velkess, mentioned in another answer (if only it was brought to market).
    – Ekus
    Oct 28, 2022 at 19:44

Flywheels for residential use should come as a kit, not only because of their weight, but also because of their size. Transporting a pre-manufactured rod that heavy is simply not an option for residential customers, poring any cheap but heavy material into a flywheel cavity may just do the trick, and surely almost everyone can dig a hole of 4m diameter in the backyard to accommodate for a flywheel kit. There have been initiatives on Kickstarter, with no success so far, but times have changed. Flywheel batteries are probably the most compact energy storage systems that can be designed with the lowest environmental impact and highest durability.


Not quite domestic, but the technology keeps maturing.

It's better suited for leveling short-lived and massive power needs rather than storing energy for days (note the 7%/hr loss below).

Each Kinext unit includes a flywheel with a mass of 5,000 kg and a large diameter (around 2.6 meters). The companies said it is able to spin relatively slowly, at 1,800 rpm, but it can reach a maximum speed of around 950 km per hour.

Source: https://www.pv-magazine.com/2022/10/04/abb-to-stabilize-dutch-grid-with-9-mwh-battery-flywheel-storage-facility/

Diameter quoted from: https://new.abb.com/news/detail/93092/regenerative-drives-and-motors-unlock-the-power-of-flywheel-energy-storage-to-stabilize-europes-grids

More data from: https://www.s4-energy.com/includes/PDF/Flyer_KINEXT.pdf

Technology Flywheel
Capacity 30 kWh
Stand-By Loss 7 %/hr
Power Capacity Modular, 0,1 – 1 MW
Cycle Efficiency 92 %
Response Time < 20 ms
Lifetime 20 years/1.000.000+ cycles
Footprint 4 x 4 meters

2.6m diameter, 5 ton flywheel with footprint of 4x4m, bolted to concrete floor

In addition to the Kinext solution described below, ABB is advertising flywheel solutions for ships, tugs etc. which need little energy most of the day (and can charge the flywheel) but need massive bursts of energy for minutes at a time (and today they carry 5x the number of diesel generators just for those bursts):


Domestic flywheels. Not going to happen. Flywheels make noise, and the ones spinning at highest rate (the ones storing maximum amount of energy) are dangerous, if the flywheel shatters it could send high-speed fragments everywhere.

Flywheels are heavy: even lead-acid batteries of a given size would weigh less.

Plus, the self-discharge is horrible. About the only case where you might want to use a flywheel is a device that is discharged right away when the power goes off. Even a day of wait for discharge to happen would self-discharge the flywheel too much. So you'd be limited to UPS devices, except devices that would weigh three times as much as a typical lead-acid UPS and be at least three times larger if not more -- plus the UPS would need to be sized to be round as opposed to typical lead-acid UPSes that are rectangular cuboids. Typical flywheel capacities start from 3 kWh -- typical UPS capacities end at 1 kWh, so that would be one helluva UPS: I'd imagine the UPS weight would start from about 250 kg, making it more of a fixed device than a portable device that current UPSes typically are.

And besides, UPSes will probably someday move to lithium-ion. That gives 3x additional power-to-weight benefit, so after this transition a flywheel would be probably 10x heavier.

What flywheels will be used in the future could be synchronous compensators in electrical grids to add inertia. Yesterday, electrical grids had all coal, oil, natural gas, nuclear and hydropower. Al of these had rotating turbines that could store about 5 seconds of energy at a typical energy consumption rate since the large turbines had lots of inertia. Not only that, but the turbines were synchronous, so if more power is drawn than is generated, every single turbine in the electrical system would slow down at the same time, decreasing the frequency in the whole network.

That's useful since voltage in the network can vary at different points due to voltage drop, but frequency is the same everywhere. If more power is drawn than is generated, it's possible for the grid operators to notice it from the frequency change, avoiding a catastrophical grid collapse, by adding more fast capacity quickly or removing some load. There's few seconds to do such a change before the grid collapses.

Today, energy generation doesn't have as much inertia. Solar has zero inertia: no energy stored in the cell or the inverter (well maybe there could be an electrolytic capacitor or two, but the storage of these is so small that they can't really be counted as inertia). You might think that wind power could have inertia, but actually not given our current software: wind power uses a clever arrangement where energy is generated between a rotating field and the turbine rotation (https://en.wikipedia.org/wiki/Doubly-fed_electric_machine). This allows to adjust the output frequency to be whatever needed (50 Hz) for any turbine rotation rate, which varies in different wind conditions. Because of this non-synchronous style of generation, wind power software simply tries to generate energy at 50 Hz and isn't contributing to the inertia in the grid. Maybe a future software update could use wind turbines to simulate inertia.

Similarly, batteries don't have built-in inertia. They have energy storage, but without the right kind of software, they don't contribute to the rotational inertia in the grid, i.e. they are generating power at whatever frequency is needed (50 Hz), rather than forcing grid frequency to slow down as energy is drawn. A correct software maybe could turn a battery to rotational inertia.

So to allow lots of solar, wind and batteries, we need more rotational inertia in the grid. A synchronous compensator does that. It's just a motor-generator with a flywheel rotating at synchronous speed, storing energy in its rotation. If more energy is drawn from the grid than is produced, it will slow down among all other synchronous generators, strengthening the frequency slowdown signal (drop in frequency) that can be used to detect the condition of too much power drawn and too little power generated.


The loss through storage due to friction by using a flywheel seems like it wouldn't matter much. In Florida there is plenty of sun. Who cares if some energy is lost. Anything collected is free. We as homeowners don't need a high tech composite spinning at high speeds. We should put a heavy steel one in the ground only turning at a safe slow speed. It would be a one time cost compared to batteries, that have to be replaced often. I'm sure the battery makers don't want to see flywheels on the market. It would be bad for their business. Flywheel is a new technology. The battery makers will make it seem like it's unsafe, but it's the perfect solution to energy storage. Municipalities won't go for it either, without a lot of money for licensing and even insurance. They're afraid of their own shadow. God forbid there's a spinning heavy disk in the ground outside my window. You'll never see them domestically. No company could ever guarantee that they'd be safe. The energy companies also don't want individuals to be self-sufficient. They want us to buy energy from them. They have a good thing going for themselves and they don't want anyone rocking the boat. Imagine thousands of homeowners not needing to pay an electric bill anymore. I visualize a one ton flywheel about 15 feet in diameter turning at 100 - 200 rpm. It would probably have to be in a cement enclosure, and in Florida a sump pump to keep it dry.

  • 2
    A 1,000kg, 5m, 200RPM flywheel would store 685,567J of energy if it was shaped like a disc. That's 0.19kWh of energy — enough to boil the water for about seven (7) cups of tea or run a typical airconditioner for about 10 minutes. I think you might be over-estimating how much energy these things can store.
    – Tim
    Aug 16, 2018 at 3:49
  • 3
    This does not answer the How close are we? question and has more opinions than facts.
    – user2451
    Aug 16, 2018 at 9:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.