Today is...
Saturday, October 21, 2017
Welcome to, the global online
community of automation professionals.
Featured Video...
Featured Video
A demonstration of EtherCAT control of linear motors using the CTC EtherCAT master.
Our Advertisers
Help keep our servers running...
Patronize our advertisers!
Visit our Post Archive
Motor Generator - Brushless DC Motor
Thoughts on motor types for high speed flywheel energy storage.
By Tim Warren on 26 May, 2005 - 3:46 pm


I am thinking about desgning a high speed flywheel UPS. There are flywheel UPS's about on the market but these are all industrial units which can supply hundreds of kilowatts for only a few seconds.

I wish to make a flywheel UPS which can supply a smaller current (say 3KW max) but for a much longer period of time by storing a lot of energy.

I was wondering how to go about experimenting. Obviously I could use a standard DC motor as it will work as a generator when not on mains power however with the high speeds the brushes will not last long. So then I thought of AC induction motors. These can be controlled with an inverter but I'm not sure whether they will operate as a generator when at high speeds. I have seen induction motors used as generators buty they need to be run at their rated speed ie. 3000rpm for 50Hz because when you move away from this speed the motor stops generating.

So I am wondering about brushless DC motors as the best alternative? But can these be used as motor generators too or just as a motor? Does anyone have any experience in this?

Thanks for you help and ideas!


By Michael Griffin on 26 May, 2005 - 11:07 pm

As far as I am aware, a flywheel energy storage system would use a brushless motor / generator with permanent magets on the rotor. The flywheels would be driven by the brushless motor to store energy, and it would be used as a
generator to extract energy. The output would have to be rectified and run through an inverter to turn it into useful AC, as the frequency straight off the flywheel is not controlled.

However I would like to point out that flywheels can be very dangerous. Storing a large amount of energy as rotating motion means that a mechanical failure in the flywheel would be a catastrophic failure, with possible fatal results to anyone nearby unless the system was adequately shielded.

By Tomy Zacharia on 27 May, 2005 - 1:49 pm

Dear Tim,
I admit that I have no experience with implementing flywheels, but there is a reason why flywheel UPS systems are used with 100's of kilowatt systems. The flywheel tends to be very compact for the power output it generates. Imagine the size of the battery bank of a 250 KVA UPS. The flywheel solutions in this size are generally coupled to Diesel Generators. Thus the generator is expected to start after 10-15 secs of power failure.

The flywheel manufacturers have data sheets on their sites.

Coming back to the original question, does your application have problems with a 5 KVA online UPS?

Tomy Zacharia

OK, some of you may notice the date on this posting. I posted my origional question and forgot to take a book mark of the page. I then went on holiday and promptly forgot that I'd asked a question. I found it again completly by accident when searching google and I saw my name in the extract text. Lucky hey. Better late than never I hope. :-)

Anyway back to the plot.

Tomy, In answer to your question why a standard UPS would not do. The application I have in mind is not really a UPS for when the power is out, but instead to take advantage between the cost of electricity at night (3.4pence/KWh) when compared to the day (12.5pence/KWh). (well actually it's just to forfil my quest for knowledge and something to play with)

I'd like to store enough energy over the night to accomodate my needs during the day. Batteries are out because of the poor charging efficiencies and the limited number charging cycles. The cost of replacing a battery bank large enough would be extreme.

I have found this company:

They make brushless DC motors of 6KW or 12KW ratings (and might even be scalable beyond that). They look very attractive. Await to hear back for a price. They have speeds of upto 6800rpm and voltages as high as 72v. I need to do some calculations to see how much energy can be stored at that speed (allowing for a minimum useable generation speed) but it would make a good feasability study.

I have bought a small hobby DC motor and controller off eBay to have a play with.

Steve, I thought of using gearing to increase the flywheen speed and store more energy (e=mv²) but I decided against it for two reasons. 1) the flywheen UPS is intended to run 24hrs a day so a drive train would be a component which would need regular replacement - hence cost. 2) Over coming the initial inertia would require some fancy gearing to get the flywheel started. It would be like trying to get your car rolling in 5th gear! If anyone has any comments please let me hear them.

Michael, I am well aware of the energy density and am confidant in my understanding and engineering skills. I would envisage the flywheel being stored in a cylindrical metal tube, which would be evacuted to reduce windage. This tube would be buried underground with an access hatch above. Hence the meed for long term maintainance free operation.

I my dreams I'd like to store about 25KWh or usable energy.

Thanks for all the replies! Speak to you sooner this time! ;-)


I have a shed that I use as a workshop/storage room. It has a single turbine ventilator on top. That supposedly keeps the shed cooler, but I can't say I can tell, during summer. :-)

Anyways, that thing is ALWAYS spinning! And I'm thinking to myself... "why not harness all that wasted rotational energy?" Shoot, maybe even harness the heat that builds up inside that shed, too!

Even if you mod the ventilator to generates some infinitesimal amount of energy... that's STILL free energy! Get a rack of batteries (or whatever) and start saving up all that free energy for "a rainy day", so to speak.

Are there any sites that show how you can mod a turbine ventilator for such a purpose?

And, from what I've read, you can use a DC motor as a generator? How? By turning it the opposet direction it turns when being powered by DC or...?

Reply to: luposian(at)

Change the (at) to @ to reply. Thank you.



The answer to your question is yes you could generate power from your shed vent, however the amount that it would generate would be so small it's probably not worth it. By the time you take into account the losses in the circuit and the effeciency of the motor i doubt that you will get enough to even light a torch bulb. Batteries take a certain ammount of energy just to hold their charge.

Yes you can use a DC motor as a generator. The simplest is a standard brushed DC motor as it produces DC voltage that can be used to charge batteries or power circuits with very little electronics needed. Other motors such as brushless DC motors or AC induction motors can also be used as generators, but they will produce an AC output which will need converting to DC before it can be used to charge batteries.

Your DC motor will have two terminals, lets call them A and B. Say you connect +ve to A and -ve to B and the motor turns clockwise, then when you use the motor as a generator; when you turn the motor clockwise A will be the +ve output. Simple. The voltage at the output terminals is proportional to the speed of rotation.


Hi There,

I want to run brushless DC motor as a generator. But I am wondering that how much power it can deliver? Because rotor has fixed magnetic field and if we run this rotor at fixed speed it will generate fixed terminal voltage.

And if we connect load to this terminal voltage, what will happen? Rotor speed will go down?

I haven't done this test before; please guide me in this matter.

Many thanks,

hi Gaurav
It should be able to generate close to its rated power at a certain RPM.

If you conect load at the terminals the current will increase and that will mean an increase in torque felt as a force oposing the rotation.

Depending on how powerfull is the driving person/turbine/engine/flywheel behind the shaft this could result in a decrease in RPM or not.


I want to be sure I understand. Brushless DC motors (like computer fans) can be used to generate power when force is applied to their shaft (Like a windmill). Am I right? I've played with small cheap DC motors and gearing systems to spin them fast and get useful voltage, but never tried with brushless. Thanks!

By George Younkin on 20 September, 2006 - 12:45 am

There are 2 kinds of dc motors. One has a wound field that is seperately excited, the other design has permanent magnets for the field flux supply. You can rotate the motor shaft and produce an output voltage. This is a practical test to determine if the motor is operaing properly. For a dc motor with a magnetic field, you can operate the motor in the constant torque range up to the rated output. For a wound field dc motor, you can operate the motor thru the rated output. Then if you weaken the field, you can increase the speed in the constant hp range.

A bldc motor is usually a synchronous 3 phase motor where the comutation of phase currents is controled in the servo amplifier. As with the dc motor, you can rotate the motor shaft and check the phase voltages to see if the motor is operating properly. The voltages should agree with the motor specification at rated speed. This is how we can check the voltage constant Ke of the motor. However, the phase voltages will be ac so you need to use an ac voltmeter.

George Younkin


I would like to purchase a 12V Brushless DC motor for an engineering project. But given it is a 4980KV motor how do i know what is the mass of the load (i.e. flywheel) can it actually rotate? How do i go abt calculating it? Can anyone assist me. thanks!


By George Younkin on 30 September, 2006 - 12:14 am


You can calculate the inertia of a flywheel with or without a hole from the attachment included.

George Younkin

By George Younkin on 30 September, 2006 - 5:17 pm

In my email to you about flywheel inertia, I find that you cannot send attachments. If interested email to me.

George Younkin

No. there is no reason that a brushless motor would work in inverse, The drive circuitry is not symmetrical in the sense that it would operate
as an AC to DC convertor simply because it works the other way. It might well be feasible to use the actual motor as an alternator with different electronics.


By Anonymous on 19 June, 2007 - 12:33 pm

Get a 24 volt DC motor and then get a 24 volt inverter, and there you go. That's what I am going to do, but am trying to find a 12 volt DC brushless motor.


Your idea of a metal tube that is earth shielded is a good one. Some lessons are learned only the hard way. There was a spin pit at an aircraft engine overhaul facility in which rebladded turbines were spun up to some speed above their rated operating speed to make sure everything was properly attached before the engine was reassembled. Lots of angular momentum in these tests!. The pit was lined with lead bricks to absorb energy from a catastrophic failure. The first failure caused lead to shoot straight up out of the pit, rupturing its cover and the ceiling above it. Redesign of the pit included a top ledge surrounding an inward sloped truncated cone to force debris toward the axis where its would dissipat its energy with other debris from the opposite side of the pit. There is a lesson here, and as you are confident of your engineering skills, I am sure you will understand. Confinement of debris is one step, energy dissipation is the second step in safety design.

Jim Rock, PE

By Curt Wuollet on 2 April, 2006 - 12:35 pm

There are lots of people doing this already. Study the physics of spinning massive objects at high speed. The reason they limit the RPM is not that the speed advantage is lost on them. Among other things, you have to get approvals to install in many parts of the world and their confidence in your engineering might be less than yours.


By Jitender Bedwal on 16 September, 2009 - 2:36 pm

You can use a CVT or increase the gear ratio step by step. Btw, i am also designing, a CVT for the exact same purpose, but only the CVT. [:P]

By Steve Myres on 27 May, 2005 - 1:52 pm

Since you've already identified several types of motors that would work in your application except for the speed, why not add a speed increasing stage? This would also allow the motor to run at it's design speed when motoring (storing energy in the flywheel). I would suggest a timing belt
drive as they need no lubrication, like to run at high surface speeds, and are very efficient.

Steve Myres, PE
Automation Solutions
(480) 813-1145

By Curt Wilson on 28 May, 2005 - 8:41 am

A few points to consider:

A properly controlled induction motor can be used as a generator over a wide range of speeds. People do it all the time, even if only to get good deceleration. If you apply proper vector, or field-oriented, control, then commanding torque in the opposite direction of velocity will automatically make it generate (just as commanding torque in the same direction as current will automatically make it motor).

This type of control of an induction motor is really very similar to that of permanent-magnet brushless motors -- a permanent-magnet motor just has zero slip and zero field (magnetization) current. In our own controllers, you would just set these two parameters differently for the two types of motors.

By the way, simple inverter control will not do the trick with any type of AC motor -- you will need smarter control techniques.

One possible advantage for the induction motor is that if you remove the magnetization current, there is no generator action at any speed, and therefore no voltage at the terminals. For people doing long-term energy storage in flywheels, this is an important safety consideration (although they tend to use reluctance motors, in my experience).

I agree with an earlier commenter that you will need to rectify the generated AC power (from either type of motor) and then invert it back to 50/60Hz AC. This is for two reasons. First, to maximize the energy storage for a given size, you will want the thing spinning fast enough that it would put out a lot more than 50 or 60 Hz. Second, as you pull power out of it, it will decelerate, and so not produce a constant frequency. (With induction motors, you have some ability to play with the slip to maintain constant electrical frequency as the mechanical speed varies -- another potential advantage of the induction motor (and one used in wind turbine generators for many years) -- but I doubt that it would be enough to prevent the need for rectification and re-inversion.

Curt Wilson
Delta Tau Data Systems


I have some anecdotal evidence for you. I actually have a brushless DC hobby motor spinning a brass flywheel (about 80g) at about 22KRPM in a little project I've been building. When we disconnect the power the system keeps running, powering a 12V computing system for about 30 seconds. The first time we unplugged it we shook our heads and wondered what was going on. Then we had that "Aha" moment...

The brushless motor controller we're using is a Castle Creations Phoenix.


I see where you were running a brass flywheel with high speed motor. I am looking for a high speed motor with a flywheel for a gyroscope project. I would be intrested to see how you coupled the 2 together and where did you get the parts.


Hello Dean,

How you convert AC three phase to DC?

Thanks a lot!

Bridge rectifier. 3 phase 600VAC will give you 800VDC.

I am working on an engineering project in which I need a very efficient generator for cheap. It has to be effiecient because the motor will be turned by human effort. As an rc airplane hobbyist, I know that dc brushless have very good effieciency. This would be great except that it takes a very complex controller to operate these. I initially thought that the controller would not be able to run backwards, but your post gave me a glimer of hope. I also have a castle creations phoenix controller rated for 25 amps. I was wondering if you could give me some more information on you findings and verify that you can actually spin the brushless hobby motor and get useable dc power through the controller, thanks.

By Craig Zeyher on 2 December, 2006 - 3:02 pm

The motor you describe above (12V -22krpm) along with flywheel seems to be what I'm looking for could you tell me the source of such a motor? I too would also like to know how you coupled the flywheel. Any help would be appreciated.

Sincerely Craig

You can increase the efficiency of the system by reducing the friction of the air on the surface of the flyweel. This friction is extremely significant at high speeds, and easily much more than the bearings' friction.

This can be done essentially in two ways:

1 - Seal the system airtight and create a vacuum. This one is cheaper because all you have to do is get a small vacuum pump and a pressure sensor and maintain a 70% vacuum or more.

2 - Seal the system airtight. Replace the air with a 'lighter' gas such as helium. This one is somewhat easier to do as it is easier to keep helium at ambient pressure in a recipient than maitain a vacuum, but you have to get a small tank of helium and replenish the system ocasionaly if the seal is not 100% perfect or every time you do maintenance.

The cons of these economizing solutions is that the motor will not cool as well so consider attaching a heat pipe from the motor to an external heatsink.

If the system is cylinder shaped, having the motor screwed on to one of the tops, and with a good contact surface between the motor and this top you can go without the heatsink as long as this top part is a good conductivity metal such as copper or aluminum or even steel and well ventilated.

By George Kaye on 17 May, 2007 - 12:58 am

I remember an article about using a flywheel in a bus for energy storage that would be revved up every so often, It described the flywheel as being made of kevlar so if it flew apart it would just make a mess in the housing. I like the idea of energy storage like this. Maybe you're on to something big!

By Paul J. SPoeltman MSc on 17 August, 2007 - 12:07 am


A suitable option to cope with the high revolutions of a fly wheel would be to use a variable transmission such as used in the automotive industry. The former (van Doorne duuwband) you can set to the 1800 RPM 50HZ primary speed of the AC motor and adjust the fly wheel speed variable.

Another and more safe option is to connect a centrifugal pump and pump up water in a water colummn at night and run it as water tubine during the day.



The water pump solution has a problem which is the low pump efficiency of about 70%, and considering all the losses you would end up getting about 40% of what you put in. Plus the investment would be larger and a huge tank would be necessary, or better yet, a lagoon. Still it was good thinking.

By Michael Griffin on 28 August, 2007 - 11:47 pm

In reply to Vampzorba: This configuration is called "pumped storage", and it is used in electric generating systems on a very large scale
(hundreds to thousands of megawatts). It is used for peaking and grid control. The net efficiency which is achieved varies, but is typically between 70% to 80%. They usually use reversable pump/turbines, so the capital costs are not double that of a conventional generating plant.

By norman mcgeoch on 14 December, 2007 - 12:15 am

Hi Tim and others on the forum,

My name is Norm and I worked in Special Effects for 20 Years. I think it is important that someone quantifies how dangerous a flywheel can be. 25 kWh of energy stored is 90MJ but let's say it out loud, 90 million joules. A kilo of TNT (2.2 lbs for our imperialists) can release roughly 4 million joules. Divide this up and you get roughly 1Kwh=1kg of TNT. A flywheel spinning at 100,000 RPM will release its energy in a catastrophic failure in about the same time as TNT. Buried 300mm down in free earth (your normal garden) 1kg of TNT will produce a crater roughly 2 metres (6 ft) in dia and 500 to 700 mm deep. 25kg would take out your house and probably your neighbours' on either side. As will a 25kWh flywheel if you give it a chance to get out of control...

Good luck with your experimenting but make sure you have your containment and safety contingencies in place BEFORE the accident happens.

By Will Hatch on 19 April, 2009 - 1:01 am

To all the people who are interested in this type of project.

I studied electrical engineering at university and one of my professors is researching this type of system. Any mechanical transmission is not ideal for your use here. You want a synchronous motor here with the magnets on the rotor. Any mechanical transmission will be a constant loss of power and will destroy the overall efficiency of the system.

A hobby brushless DC (BLDC) drive will work to get the system rotating. However to harness the power from the flywheel a 3 phase rectifier is necessary (if you don't know what this is check wikipedia). Advanced systems for this application have the capability of using the drives output inverter to do this.

I would not recommend pursuing high power systems here unless you have a good understanding of the electronics required to drive these machines because poorly designed electrical systems will have high losses (using the same hardware I can give you two drive methods, the advanced one has a 10% performance gain over the basic one). Also as many people have mentioned the mechanical failure of the flywheel is potentially tragic, and I think your family would potentially be upset with me if I didn't mention that.

On small scale this is an excellent learning project. If you are not highly skilled in motor drive systems use a brushed DC motor. Applying a voltage to the terminals will cause the flywheel t o spin. Once you disconnect the power supply there will still be a voltage across the terminals due to the rotation. If you apply a load of some sort across the terminals there will be a current flow and the energy stored in the flywheel will transfer.

Good Luck!