variation of speed of synchronous motor due to frequency variation 5

[RAgrawal]

friends,

there r 2 big synchronopus motors of 16 MW and 12 MW in our plant which drives compressors.
The Grid here is not very stable thats why frequency variation is from 48 Hz to 51.5 Hz.Due to high frequency, speed of motor increases which causes it to get overloaded. To reduce the loading outlet damper requires throtelling which causes the production loss.
I want to know, is there any device available which can be used in this situation

 

[peterb]

Sorry, but speed of a synchronous motor is by definition locked to the system frequency. Is there anything that can be done by your supply utility to better control the grid frequency?

 

[electricpete]

I agree with peterb. One other thought ... it sounds like your motor is driving a fan. Of course if you can vary the fan blade pitch, that would allow you to compensate for increased speed without requiring higher power.

 

[Lewish]

Synchronous motor speed is indeed locked to supply frequency. However, the point here is, the grid is varying within allowable limits. This sounds like a big air compressor for a manufacturing plant or mining operation.
This small a change in supply frequency should not affect motor operation. I would look for other causes.
Are your phase currents and voltages balanced?
Do you have high harmonics on the supply lines?
What else have you checked?

Lewis

 

[buzzp]

I agree with Lewish. SOunds like your somewhere that supplies 50Hz so the utility is providing adequate frequencies. If the motor is affected by this variation in frequency then it is likely sized tightly to the work required. What is the nameplate data on the motor? You may be able to turn up the overload setting without causing problems with the motor, if this is what your referring to as overloaded. Otherwise you may have to deal with the loss of production. Is there a gigantic drive out there capable of handling this? I don't know, I would be surprised if there is. What is the voltage?

 

[electricpete]

I would like to offer respectful disagreement with the last two posters who stated that the supply variations you are seeing are insufficient to cause an overload.

50hz to 51.5 hz is a 3% increase in speed. Using centrifugal pump/fan laws, tthe power increases in proportion to speed^3

A 3% increase in speed corresponds to ~9% increase in steady state power and an approximate 18% increase in stator I^2*R heating temperature rise. On top of that, you may have to add the transient power required to accelerate the inertia from the lower speed (particularly if you swing rapidly from 48 Hz to 51.5 Hz).

 

[Marke]

Another potential problem that you have is at 48 Hz, if the supply voltage is close to the 50Hz rated voltage of the motor, the iron losses will be increased significantly and that could cause an excessive temperature rise and shortening of the life of the motor. Every ten degreeC rise in temperature halves the life of the insulation!
Mark Empson

http://www.lmphotonics.com

 

[RAgrawal]

To all its a air compressor motor, 11 kV voltage level and designed to operate at 50 Hz. so electricpate has a point that 3 % variation in speed will cause the change in loading of motor in great extent.

 

[buzzp]

The motor is designed to operate at 50Hz, yes I agree. My point was the utility is supplying frequencies that are deemed acceptable by regulating agencies(at least here in the US). There is nothing the utility will likely do for you. Depends on the utility. I never disagreed that it would cause an overload. Your situation is a difficult one and your best bet is to try to convince the utility that they need to take action. Depending on where this power is generated and how, there is likely little they can do to help your situation. I am thinking your going to have to deal with the loss of production or turn up the OL setting. Based on my experience with motors, 70% of the time, the OL tripping is due to incorrect settings or improperly sized motors. Now granted, this motor is not anything that can be easily replaced. At the same time, you still need to continue with your process. You have to decide which is more important, protecting the motor or the process. Again, take a hard long look at all of the electrical parameters that can affect the motor current (unbalance, low or high voltage, etc) It is much easier for the utility to give you better balanced voltages or voltage closer to nameplate than it is for them to change the frequency. You may gain enough here to eliminate the OL tripping on frequency variation.

 

[electricpete]

The frequency data I mentioned (60-60.1hz) was obtained from a frequency transducer associated with one of our generators connected to the grid. However I now realize that the utility must maintain 60hz long-term average frequency so that clocks run accurately. I conclude that our transducer must be out of cal by approx 0.05hz and the actual band is probably something like 59.95-60.05.

 

[electricpete]

Weird - I posted a correction to an earlier message of mine which apparently never made it. Let me try to recreate that earlier message:

Buzz - Sorry for misinterpretting your meaning. I value your input on this board as I'm sure the others do. I only ask questions when I think I can learn something.

In the spirit of learning and exploring differences...
The range of frequency variation mentioned is very high in my experience. My indicated data from ERCOT grid in Texas, USA shows that frequency stays in a tight band of 60-60.1 [probably is 59.95-60.05] from December 2000 to present with only a few brief transients outside this range.

Are there any standards which govern acceptable ranges of frequency variation? (Don't tell me NEMA MG-1.... the allowance for frequency variations is not intended to imply motors should be allowed to operated above rated horsepower due to high frequency)

 

[Marke]

I believe that there is a British standard that calls for the frequency to be within 0.5 Hz.

Certainly from my experience, that is the limit normally applied for short term variations. Mark Empson

http://www.lmphotonics.com

 

[electricuwe]

RAgrawal,

I do not know where your plant is situated, but according to the large variation in frequency it seems to be situated in a small grid or in a developing country. Maybe you bought the equipment from an Eurpoean or American manufacturer who didn't know about the large variation of frequency at your site and designed to the standards applicable at his location.

The only solution which seems to be possible is to replace the rotor with three-phase wound rotor, install a small converter (3% of nominal power in your case) and operate it as double fed asynchronus machine.

 

[Lewish]

electricpete - Looks like I need some more education. Could you show me the math that says a 9% increase in power results in a 18% increase in heating. I would like to see what I am missing. Watts is watts the last I knew, whether they be power or heat. Thanks.

Lewis

 

[buzzp]

Ill have to check on standard for freq variation by the utility. I agree that this wide freq variation is definately uncommon here in the US. Some motors are rated for 50/60Hz? If this is the case with this large motor then it is near the bottom of its frequency range and will cause more problems than if it were a variation off of 60 Hz since the FLA goes up slightly with a 5% decrease in freq. On the other hand, my EE handbook is showing the OL capacity of the motor will increase as the freq goes down as well. If this is true, then he should be able to turn up the OL setting.

 

[electricpete]

Lewish
Here was my reasoning:
- Current is approximately proportional to motor output horsepower (particularly near full load). => 9% increase in output horsepower equates to 9% increase in current.

- I^2*R heating is proportional to the square of current. => 9% increase in current causes 18% increase in I^2*R heating.

Temperature rise is proportional to losses (heating).

Approx 20% of full-load losses are constant (no-load) loss and the other 80% vary in proportion to I^2*R. As a rough approximation the losses (temperature rise) near full load can be approximated proportional to current-squared.

Watts is watts surely is true... but I^2R watts are not proportional to output watts.

 

[electricpete]

Perhaps the compressor manufacturer can give instructions for partially unloading certain stages of the compressor to accomplish a reduced power consumption at constant speed.

 

[Lewish]

electricpete - I will have to give you a star for that explanation. Took a little thought with a calculator, but I see how you get there. Thanks.

Lewis

 

[electricuwe]

For information on converting a synchronous machine to a double fed asynchronous machine, as I have suggested in an earlier post, look at:

http://leme

http://www.epfl.ch/ME/publications/1998/speeddrive_1998.pdf