Will increasing the copper area of AC motor winding affect its VFD operation? 4

[edison123]

We have planned to rewind AC cage motor with increased copper area with the same original no. of turns and original winding connection. So, the winding inductance will remain the same while the winding resistance will come down by about 20%.

This motor with increased copper area along with another motor with original winding is run via a single VFD.

Will increasing the conductor area in one motor affect the VFD operation and the load sharing between the two motors?



Muthu

http://www.edison.co.in

 

[electricpete]

> the load sharing between the two motors?

IF they are driving the same shaft in tandem and with the same applied frequency and voltage (those are all my assumptions) then they obviously have the same slip. The slip characteristics will determine the load sharing. Identical rated motors with identical slip characteristics would share the load equally.

As a first approximation slip is proportional to Pout/[V^2/R2]. So it is very sensitive to the equivalent circuit parameter R2 for rotor resistance, but not particularly sensitive to the equivalent circuit R1 for stator resistance . You can develop a more complete expression for slip from equivalent circuit that would include R1 but R1 would play a much smaller role. My guess is no noticeable change in load sharing.


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

 

[electricpete]

> You can develop a more complete expression for slip from equivalent circuit that would include R1 but R1 would play a much smaller role. My guess is no noticeable change in load sharing.

Attached is an algebraic excercize which includes the more complete/complicated expression. It examines the effects of equivalent circuit parameters on the torque speed curve.

One result of interest is halfway through (the middle of page 5) where we evaluate dT/ds at s=0 (i.e. the slope of the torque speed curve at s=0 which is representative of the slope at low slips)

The result is
dT/ds (s=0) = [Vs²*XM²] / [R2*wsync*(XM²+R1²)]
(btw, can TGML format an equation to show a numerator above a denominator?)

The only place that R1 shows up is in the demoninator, but it is "added" (SRSS) to XM, which is orders of magnitude higher than R1 during normal operation, which means R1 has negligible effect.

As long as R1 << XM then R1 has not much effect on the low-slip torque speed characteristic
The condition for insensitivity of the low-slip torque/speed slope with respect to changes in R1 is:
R1 << XM = j*w*Lm

This conclusion is most defensible during normal operation where w is high.
During ramp-up with very low w, then XM=j*w*Lm is smaller and the conclusion may not be as solid.
So IF there is any load sharing problem created by change in R1, then I think it would only be at very low frequencies at the beginning of rampup.

I should add one more caveat - all of my analysis has been steady state or quasi-steady state and with a simple (not modeled) control system. So if there are load sharing problems during transients or as a result of the control system, I wouldn't be able to foresee those with this type of analysis. My gut feel is that you are fine to add stator copper, but I have almost no experience with vfd's, so... hopefully someone else can provide you additional input.


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

 

[Drivesrock]

Depending on the VFD there may be several types of control available - simple scalar to a flux vector control, (however the different manufacturers call theirs).

Two or more motors in parallel with flux vector control is possible as long as the motors are 'the same'. Same rating motors may vary up to 5% between them I was told once by a motor manufacturer - but that's ok. I've done several applications of multiple motors on a VFD in the different control strategies. Please let us know the application.
In scalar control the VFD is just a source of Volts and Herts and each motor does it's own thing.

From the practical point of view, whenever a motor is replaced for 'same' though, I always recomend that any auto-tuninng or motor model calibration functions be performed (as may be available in the selected control mode) as rewound motors can come back 'different'. Some VFDs when replaced, may need this too as the calibration run can determine some drive power circuit characteristics that it needs to 'know'.
With a different stator winding resistance (although same stator inductance) there will still be a different voltage drop across the stator - but how much depends on the inductor part. A VFD in flux vector control will need to know this - especially those that may track Rs as many do Rr (Rotor resistance) for compensating.

In your case of two 'nearly same' motors in parallel (but if/how they are mechanically connected we don't know) I think that the future operation of them with the VFD could be checked / accompanied at the beginning - motor currents.

Best Regards

 

[waross]

My thoughts; More copper should give better efficiency.
What about harmonics? Will a larger cross section result in increased eddy currents in the conductors? Will this be worse for harmonics? Will a lower resistance will result in a percentage of the harmonic losses being shifted to the supply conductors?

Bill
--------------------
Ohm's law
Not just a good idea;
It's the LAW!

 

[edison123]

Thanks pete. Agree at lower frequencies the stator X/R ratio becomes less but that goes for the other motor also, right? So, the effect of lower R should not be significant?

Thanks DrivesRock. I will find out what type of control the client uses. These are motors used in oil rigs where they drive the drills. They are not driving the same drill so 'load sharing' term I used may not be right. Good point about re-tuning the drive for the new R. Will inform the client.

Bill. Yes, the idea was to reduce stator copper loss. Not increasing the conductor size per se, just increasing the no. of conductors. So, the eddy current loss, which even otherwise is very low, doesn't play a big role here.



Muthu

http://www.edison.co.in

 

[zlatkodo]

As for the motor itself, you will get:
- lower resistance ie. less heating and more efficiency
- same harmonic content
- same FLA and NLA
- higher SF

ACW

 

[edison123]

One more question.

Can a single VFD run two motors with different KW/HP? Do you auto tune the drive for each motor first before you connect them in parallel?

Muthu

http://www.edison.co.in

 

[edison123]

Paging ET drive gurus, Keith, Jeff et. al.

Muthu

http://www.edison.co.in