Motor Current Oscillation 3

[Masbibe]

Hi to you all.

I am interesting does anyone have expirience in commissioning big induction motors supplied via VSD that workink in scalar control mode (V/f).
I had commissioning 2.85 MW motor, 660 V, 50 Hz. Drive is ABB ACS800.
Uncoupled motor was worked very bad in scalar with some high oscillation of current and in DTC everything was OK.
You can see in attach graph of uncoupled motor in scalar.
Does anyone know what could be reason of oscillations.

Best regards,

Milovan Milosevic

 

[Skogsgurra]

The oscillation is around 6 Hz - if I understand correctly. That is a frequency that could be the result from rotor flux, rotor inertia and stator flux forming a resonant system.

The possibility has been discussed before in this forum and e-pete has done some simulations.

Does the frequency change if you reduce excitation (lower stator voltage)? If it is the rotor swinging in stator flux, then the frequency shall decrease if excitation is reduced.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Masbibe]

Thanks Skogsgurra for this respond.

I didn't try with reducing stator voltage so I dont know and I am not any more in this factory so I can not try.

But this method with reducing stator voltage is only usefull to see if the problem is with this resonance, and what to do if this is true and you must run motor in scalar.
How can you decrease this oscillations.

For example some my colegue had commisionned HV motor with LV drive with step up transformer. So all recomandation is to run drive in this situation in scalar. They also had some problems with oscillations. Also they had problems with very very high current on very small frequencies but this was probably because transformer is almost short circuit for drive on small frequencies.

Do someone have experience with drives with step up transformers.

Milovan Milosevic

 

[Skogsgurra]

Yes. We used a 600 kW Siemens Masterdrive with a 500 V to 6 kV step-up transformer to start a large (think 6 or 8 MW) synchronous motor once.

We got bad oscillations and eventually had to switch to torque control and limit torque to positive (no braking torque allowed) to be able to control the motor.

It was a rather tiresome experience - especially after we found (somewhere on page xxx) in the manual a sentence saying that synchronous motors could not be run with that VFD.

But, we had to do it - so we did it. After lots of hours...

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Masbibe]

Skoksgurra,
Drive was probably thought that he run iduction motor ?
So you run motor in torque control, not vector speed control.
On which frequency was oscillations.
My idea for this step up application is to have current limit that is linear function of output frequency with offset on zero frequency (motor is running pump).

Milovan Milosevic

 

[electricpete]

I doubt it will help, but the type of resonant oscillation that skogsgurra referred to in his first post is well described in the figures here:

http://lipo.ece.wisc.edu/1970s pubs/T07.pdf

It completely ignores any supply/control system dynamics, just assumes a constant-frequency sinusoidal supply and which frequencies would cause resonance for given motor parameters.

=====================================
(2B)+(2B)' ?

 

[Masbibe]

Skogsgurra,

Sorry I written your name wrong. I was typing on mobile phone.

Milovan Milosevic

 

[Skogsgurra]

LOL!

That is really one of my least problems right now!

BTW, it doesn't mean anything offensive in our language (Swedish) and I hope it doesn't in yours either...

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[Masbibe]

electricpete,

Thanks for this paper, but it has to much mathematics and process control theory for me, but I read it.



Milovan Milosevic

 

[electricpete]

A summary of the article is that a motor system fed from a fixed-frequency supply has eigenvalues that are a function of the supply frequency and of the motor parameters. (the frequency of the supply would be fixed for a given simulation or analysis, but varying between simulations/analyses). The unstable regions are where the eigenvalues have have a zero real part or a positive real part…i.e. the mode is not damped. If you were experiencing this type of problem, it should go away when you vary the supply frequency (for example the motor in the article was only unstable when the supply frequency was around 0.3+/-0.1 times the base frequency). Note that in general the frequency of this unstable oscillation will be different than the supply frequency (it is not like a resonance where the system vibrates at the excitation frequency).

If you are interested, I can calculate the eigenvalue for your uncoupled motor. What I would need is:
number of poles
full load: amps, power factor and efficiency
half load: amps, power factor and efficiency if available
no-load: amps, power factor if available
motor rotating inertia
frequency of the supply at the time you experienced the oscillations.
(I don’t need locked rotor torque or locked rotor current).

The output of the analysis would be:
frequency of oscillation
damping associated with oscillation… (indication of whether stable or not and how much margin).

There can also be some what-if sensitivity analysis…how does the eigenvalue change when you change the inertia or add a load torque or change the supply frequency or voltage.

I don’t’ know much about vfd’s, but as far as I know scalar control is a simple open-loop control with constant supply voltage magnitude and frequency (as long as the setpoint isn’t changing). In that case, Krause’s analysis would seem appropriate for scalar control .

By he way, that is the biggest low voltage motor I have ever heard of! (2.85MW, 660V).
Are you sure you have the rating right (2.85MW, not 285KW).
Is it a squirrel cage induction motor?


=====================================
(2B)+(2B)' ?

 

[Masbibe]

electricpete,

Motor ratings are OK so 2.85 MW, 660 V, 3000 A, 50 Hz, 992 rpm, power factor 0.86. This is all parameters I know right know.
You dont need to struggle whit those equations to find critical frequency.
This oscillations was appeared only on some frequencies (somewhere between 10 and 15 Hz) and only during acceleration or deceleration (motor was uncoupled).
Also on faster ramps oscillations was smaller (probably because of faster passing through this critical area).
I can attach some pictures that illustrate this if you are interesting.


Milovan Milosevic

 

[Masbibe]

electricpete,

I was reading some your old post in which you said that oscillation of torque during direct on line start of induction motor is because of DC component in stator current. Are you sure in this.

Milovan Milosevic

 

[electricpete]

.

I can attach some pictures that illustrate this if you are interesting.

Yes, I am interested.

You dont need to struggle whit those equations to find critical frequency.
This oscillations was appeared only on some frequencies (somewhere between 10 and 15 Hz) and only during acceleration or deceleration (motor was uncoupled).

The the oscillation only occurred at certain frequencies is consistent with the Krause paper (there are certain zones of instability, primarily 0.3* base frequency for his motor).

The objective of the study would be to determine the damping factor of those modes to check whether they are in fact predicted to be unstable. Whether the oscillating frequency matches observed oscillating frequency (for a given supply frequency) would be a factor that could be used to corroborate or validate the model..

I already have a program ready to go and have done it before for another motor discussed in this other thread:
thread237-249262

I was not able to post the full results of the eigenvalue sensitivity analysis in that thread, but I have attached them to this thread. It starts with a numerical sensitivity analysis of the eigenvalue. From the numerical sensitivity analysis we can see that the eigenvalue frequency is proportional to voltage, inversely proportional to square root of inertia, inversely proportional to the sqrt of sum of leakage reactances. This suggests a simpler model which matches the behavior of the eigenvalue as shown on slides 3-5 (not as good as full eigenvalue analysis, but a little more intuitive).

But there is not really enough data to get a good model. Do you have motor inertia and full load efficiency?


=====================================
(2B)+(2B)' ?

 

[electricpete]

I was reading some your old post in which you said that oscillation of torque during direct on line start of induction motor is because of DC component in stator current. Are you sure in this.

I am sure they are related somehow.

Attached is simulation of start of a 2250hp motor.
On slide 1 you can see the 60hz torque oscillations which die out at approx t=0.8 seconds.
On slides 5, 6, 7, you see the stator phase currents.

Let’s analyse the shape of the phase currents:

If you follow the simpler approximation discussed in the other thread by Rockman, we can say we have a sinusoidal component plus a decaying dc component that is gone by around 0.2 seconds.

But if we take a closer look, that dc component is not simply decaying, but instead it goes up and down while decaying for the period from t=0 to t=0.8 seconds. I am not sure what the exact origin of that is (I am inclined to think it may be related to rotor frequency because the “frequency” of that decaying component starts slow and then increases, which matches what we expect from rotor current frequency).

At any rate, the time period of the torque oscillation in slide 1 matches the time period of the superimposed decaying oscillating dc in the phase currents in slides 5,6,7, so they are somehow related.

In my mind it is expected because we expect steady torque only during balanced conditions and we expect oscillating torque when unbalanced. Up until 0.8 seconds the phase currents are never balanced so we have oscillation. Maybe there is a better / alternate explanation (?)


=====================================
(2B)+(2B)' ?

 

[electricpete]

Just in case I wasn't clear, the key thing I was pointing out was that the period of the torque oscillations (0-0.8 seconds) is the same period as which the currents have an added oscillating/decaying dc component.

=====================================
(2B)+(2B)' ?

 

[electricpete]

Do you have motor inertia and full load efficiency?

Whoops.. sorry - we've got voltage, full-load current, power, and p.f....so efficiency is computable. How about inertia?


=====================================
(2B)+(2B)' ?

 

[Masbibe]

electricpete,

I dont have data for inertia but I was calculate it according to graphics that I recorded.
So I get this values: 105, 107, 100, 100, 110, 98, 111, 99 (all in kg*m^2).
This is values from differents accelerations and decelerations that I record.
So you can use mean value for inertia and that is 104 kg*m^2.



Milovan Milosevic

 

[Masbibe]

electricpete,

On this picture you can see where is unstable area of motor.

Milovan Milosevic

 

[Masbibe]

So in picture in previus message biggest oscillations was on ramp that start with 10 Hz and end with 12 Hz (somewhere in 80 sec on x axis).
On next picture you can see zoom of this oscillation.
Period of oscillations is 0.16 sec.



Milovan Milosevic

 

[Masbibe]

On this picture you can see that there is no any problem on frequencies between 20 and 50 Hz. Motor works perfectly.

Milovan Milosevic