VFD Motor Losses at 0Hz

[7carisfast]

We are looking at keeping a VFD turned on but have it output 0 Hz in order to keep the motor warm without the use of space heaters. No external process conditions will turn the motor.

My assumptions are when we magnetize the motor at zero speed, we don't get the "conventional" iron losses neither we get the rotor losses, neither we get friction nor
windage losses. The only losses we have in the motor are the resistive losses in the stator.

Typically stator losses are 1-1.5% at nominal current. Magnetizing current for big motors is around 30% of the rated current, because resistive losses depend on the square of the current the losses the losses would be only .09 times 1-1.5%., which equals to .1 to .15 percent
of the rating of the motor. Any motor should be able to handle that loss and it should reduce also the thermal stress the motor.

Assuming I have a 330KW motor at 600 Amps @ 400V, that would mean the motor would only draw 0.6-0.9 Amps or 330-500 Watts worth of heat.

Can anyone confirm my assumptions are correct?

 

[jraef]

Using the VFD at zero Hz will not keep it warm enough to do any good. When VFDs are used as motor heaters it is done via the DC injection circuit, placing DC across 2 of the windings. From my experience, on a motor that big you will need around 50A of DC current going through it to keep the windings warm enough to avoid condensation.

 

[ScottyUK]

500W is pretty typical for a space heater rating on a machine that size. I guess the difference is that the space heater runs hot at 500W and can raise the temperature of the motor while the motor winding dissipating 500W will be fairly cool.


----------------------------------

Sometimes I wake up Grumpy.
Other times I just let her sleep!

 

[sreid]

If the VFD is a vector drive, the rotor will remain 100% fluxed at zero speed.

 

[7carisfast]

Yes. It's an ABB ACS800, Sensorless Vector.

 

[itsmoked]

All you need to do is make sure the motor is a few degrees warmer than its surroundings in this 0Hz state. If you can't feel a difference or measure one then you will need to add more heat somehow.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ozmosis]

I wouldn't run the motor at 0Hz with the ACS800. I'm sure you can take it out of vector mode but then I'd be asking why have you got an ACS 800 as this is a high performance drive for high perfromance applications and running at 0Hz will just cook the motor I would think, as sreid says.
What is the delta temp of your motor when stopped? Have you got a problem with condensation or is it a possibility?
If you have one, fit heaters, it is what the are there for. It is not worth screwing up your motor for a relatively low cost.
Using an injection of DC is a solution but again, it can be difficult to maintain. Maybe there is some form of logic in the ACS800 that could bring in a timer for DC injection "pre-heat" before it starts rather than keeping it on all the time.

 

[sreid]

The sensorless drives I've used only work at a 10:1 or 20:1 speed range. They don't work at zero Hz.

A vector drive (using rotor position feedback) does work at 0 Hz and keeps the rotor fully fluxes. This does not overheat motors designed for vector drives (no air flow).

 

[ozmosis]

I think that 'open-loop' with the ACS800, the control below about 5~7Hz is simply not defined. I don't think it's a case of it 'not working' other than 'it not working very accurately' and therefore it could be anyone's guess as to what happens:underfluxing or overfluxing. Hence the reluctance to suggest this as a viable option.

 

[niallnz]

You could use the adaptive programming feature of the ACS800 to set your own flux reference for zero speed heating operation. This will let you control the current into the motor. On a motor this size it would be nice to have PT100s meausring the stator temp, the ACS800 can use the PT100s to via the adaptive programming to control an external heater or set the flux reference to maintain a temperature.

Cheers
Niallnz

 

[itsmoked]

That would be slick!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[ozmosis]

niallnz
It does sound interesting but are you able to alter the flux reference 'on the fly'? as I would imagine what Zcarisfast requires at zero Hz is probably quite different to what is required for 'normal' running.

 

[niallnz]

As the ABB advertising says "it's like having a PLC built into the drive". The adaptive programming as standard gives you some space to join up 15 function blocks (AND, OR, COMPARE, PID, timers, etc) and some user constants to make the drive perform some logic in its operation. I've used it to do pipe line filling on an irrigation system, so that the drive runs at a constant speed for a time before changing to pressure control.

Short answer is yes, you can change the flux ref on the fly.