Current imbalance between phases

Are there any generally accepted tolerances for this in a three phase motor? I've seen <10% as a limit from a couple of sources, but I was wondering at what point it should become a concern. I've got a couple motors where one phase is about 15% off the average - I'm just not sure if it's a real world problem. Most of the readings I've ever taken have been pretty tightly together. I'd appreciate the benefit of anybody else's experience. Thanks.
Original Post
The NEMA MG-1 calls for a ratio of 7:1 max current to voltage unbalance. However, that is in a situation where there is a controlled circuit. In the real world, the ratio can be quite significant due to circuit impedance, etc.

There is a quick way to determine if current unbalance is related to voltage unbalance or the electric motor. It is called 'rotating the phases.' In this case, you take all your current readings for phases A, B, and C, then at the output of the starter, move phase A to phase B, phase B to phase C and phase C to phase A. You will maintain the motor rotation.

Re-take your current readings. If the current unbalance moves with the conductors then the unbalance is due to the motor. If it stays with the appropriate phase of the starter, then it is due to the supply.

I think Howard's original post was clear, but a clarification for dhlinva would be "If the current unbalance moves with the motor conductors (T-leads), then the problem is with the motor..."

It's tough to answer your original question. Most of the standard motor specifications have no limits on balance. NEMA has some derating curves for motors based on current unbalance under the assumption that it originates from suppply voltage unbalance. I have two observations:
1 - I have seen that often current unbalance decreases significantly as load increases.
2 - A small voltage unbalance can cause a large current unbalance as discussed bove.

If the motor was in the shop or uncoupled in the plant, I would not be as concerned with a 15% difference as if it were in the plant under load.

Electricalpete is right on. An elementary rule of thumb, voltage problems happen up stream and current problems happen down stream. I have tested a motor on-line and seen a problem with current unbalance. Shut the motor off and measured the motor resistance and all was balanced. Started the motor up again and still had and current unbalance. The infrared measurements at the MCC showed two legs hotter than the third. I informed the customer that I still thought the problem was in the motor or the connection in the J-box(Reason being, only thing down stream of the tester). He went to pull the motor and found the connections at the motor were loose. When the motor was running the leads were vibrating and causing a bad connection.

Use all the tools in the tool box to diagnose the problem. This is a great case study where Infrared indicated a problem and on-line motor current analysis pointed where the problem was.

Can a 0.84% voltage unbalance (470V, 477V and 475V) produce a 12.4% current unbalance (195A, 224A, 249A) in an electric motor? To make sure that the unbalance was not due to the motor we rotated the phases but the current unbalance did not move with the conductors, therefore the unbalance is not due to the motor. Can such a low voltage unbalance produce such a large current unbalance? Further the small voltagr unbalance is mantained when the motor is running or stopped.
At low load, that is reasonable.

Let’s say negative sequence impedance is Z-=(1/7) p.u. (similar to locked rotor impedance)
Let’s say positive sequence magnetizing impedance is Z+ = 4 p.u. (selected to give 25% no-load amps)

Apply a voltage with 0.84% voltage unbalance. i.e. V-/V+ = 0.84%.

ASSUME no-load conditions. We will see the following currents at no-load:
positive sequence current is I+ = V+ / Z+ = 1 / 4 = 25% of FLA
negative sequence current is I- = V- / Z- = 0.84% /(1/7) = 0.84% *(1/7) = 5.48% of FLA

The current unbalance is I+/I- = 5.48%/25% = 21.9%

If motor is under load, then the unbalance will be lower (the negative sequence portion stays the same but the positive sequence portion increases).

Add Reply