SCIM dynamics during loss of utility

[ZeroSeq]

Hello everyone,

I've been pondering the following scenario, and after modeling it in ATPDraw, I'm not exactly trusting my results:

Immediately following a utility outage at an industrial facility, will a squirrel cage induction motor's stator voltage (air gap flux) and its slip (rotor speed) decay at a rate which is a function of the other loads on the system?

The motor in question is approximately 10,000HP (13.2kV) and has an open circuit time constant of 1.2 seconds. ATPDraw shows that the decay rate (voltage and slip) with and without the remaining loads is almost identical. I'm not sure I believe this (also could consider the model erroneous..). Intuition tells me that the non-infinite impedance of the system loads would have to draw real and/or reactive power from the motor, resulting in phase angle decay and voltage magnitude decay, respectively (beyond that of the o/c damping and frictional damping).

Any thoughts?

 

[davidbeach]

Inertia of the driven load is probably the major factor. Having done real world testing of a fast motor bus transfer scheme, I've seen that individual motors will drop out at separate times while all connected together between loss of one source and connection to the other. If the electrical system governed then they'd have all reacted the same, but they didn't.

 

[waross]

Possibly the worst case:
A 400 HP wound rotor motor direct driving a fan.
The contactor flashed over phase to phase effectively short circuiting the motor terminals.
The motor slowed down faster than the fan.
Bent motor shaft.
Bent fan shaft.
Exploded coupling, parts all over the machine room.
I suggest that the distance and the impedance between the motor and the fault location will be significant.
Also, the frequency of the motor contribution will be less than system frequency and will be decaying.
As the motor slows, the voltage, the frequency and possibly the impedance of the feeder will all be dropping.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[electricpete]

I assume your other loads are motor loads.

The motors will diverge in frequency during coastdown.

Interaction of your motor with another motor at different frequency would result in instantaneous power flowing one way during portion of your motors cycle and the other way during another portion of the cycle. It would be somewhat chaotic (we can't really use the terms real and reactive power in this situation), but over a long enough period of time it would probably average to zero energy transferred between motors. So I can believe the long term speed decrease rates are roughly the same, but I’d expect more ripples showing in the speed traces when other motor loads are present.

As for voltage, if there is another motor present whose terminal voltage is decaying faster than yours, I'd think it would likewise have a chaotic effect but over a period of time I'd think it should accelerate the rate of decrease of your motors flux and voltage. (maybe all motors if analysed alone have similar voltage / airgap flux decay rates to each other?)


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

 

[ZeroSeq]

Thanks for all the responses.

With respect to just the terminal voltage (airgap flux) for a moment:
-if the motor is disconnected from the source voltage with no other loads on the system, the airgap flux should decay as a function solely of the open circuit L/R as the motor terminals looking into the system see infinite impedance
-if the motor is disconnected from the source voltage with other loads on the system, the airgap flux should decay as a function of system impedance (loads) as well as the magnetizing+referred rotor impedance.

My question is thus: Do you think it is acceptable to analyze the system referred to in the second bullet point using the steady state motor equivalent circuit?

In this case, the "open circuit" time constant will be the system impedance (loads) in parallel with the magnetizing branch, both in series with the rotor referred Z.

Or, would it be necessary to model this system using the arbitrary reference frame model as shown in "Analysis of Electric Machinery" (P Krause)?



Electricpete, you bring up an interesting point. I agree that as multiple motors on the same bus with source disconnected with then have instantaneous transfer of power between them. Can you elaborate as to why you cannot look at this as real and reactive? Would the "real power" transfer be a function of the phase angle differences and the q-axis impedances of each motor, whereas the "reactive" power transfer would be a function of the voltage magnitude (airgap flux) difference and the d-axis impedances of each motor?


Thanks

 

[aussiejohn2]

G'day ZeroSeq,
The equivalent circuit you describe seems fair...except that the other motors are not just plain impedances - they have fluxes of their own (often represented as dependent voltage source which from very distant memory is what Kraus has).
JD.

 

[waross]

What you say about the flux decay is valid for a motor that is disconnected from the supply so that the motor terminals see infinite impedance.
However when the grid fails and leaves the motor connected to other loads things change.
The motor becomes an induction generator.
In many cases the load current will induce self induction.
An example of this may be seen when a long, down-hill conveyor is driven by diesel power, or when the stated conveyor is to be provided with runaway protection.
A resistance bank may be connected across the motor terminal sized to absorb the regenerated power in the event that the connection to the grid is lost.
When the possibility exists for a sustained overhauling load, it is prudent to size the motor to withstand both the current and duration of an overhauling load.
In your case, your calculations are valid for a disconnected motor with no impedance across the motor terminals.
As the impedance across the motor terminals decreases, a point will be reached where the motor self-excites and becomes an induction generator.
Now you must also consider the decay of the energy of the load inertia.
As the impedance across the motor terminals decreases further the current may exceed the rated current of the motor.
As the current increases, the resulting torque will increase and the rate that energy is extracted from the load inertia will increase. As shown by the real world instance, the reaction torque and the resulting rapid de-celleration may be so severe as to cause mechanical damage.
Have you considered capacitive reaction?

How do you deal with this in the real world?
I would design for the worst possible case.
The action of an induction generator running at rated motor speed will be close to the action of a synchronous generator.
I would analyze the system considering the 10.000 HP motor to be an equivalent sized synchronous generator.
This will yield a conservative result.
Once the load on the generator is determined, you can consider the speed decay as energy is extracted from the load inertia.
Again, the results will be conservative.
As the speed drops I expect that the accuracy of this estimation will be less. However, in the real world this should help you to estimate the worst case limits within a few percent.

Capacitors (power factor correction) will facilitate self induction.
Synchronous motors may facilitate self induction.

Immediately following a utility outage at an industrial facility, will a squirrel cage induction motor's stator voltage (air gap flux) and its slip (rotor speed) decay at a rate which is a function of the other loads on the system?

Yes, I agree.

The motor in question is approximately 10,000HP (13.2kV) and has an open circuit time constant of 1.2 seconds. ATPDraw shows that the decay rate (voltage and slip) with and without the remaining loads is almost identical.

This is possible under some conditions of plant load and the inertia of the load driven by the motor.
Remember that this will apply to all the motors left in parallel with the motor in question.

Bill
--------------------
"Why not the best?"
Jimmy Carter