Estimating Starting Current in 3 Phase Induction Motor by Single Phasing

[BigMotorGuy]

Can you accurately estimate the starting current in a 3 phase induction motor by applying single phase power to the motor while stopped, measure the current and then multiply this single phase current by a factor? What factor do you multiply the single phase current by to get the total 3 phase starting current?

 

[jraef]

If you apply single phase power to a 3 phase motor that is not moving, it will not start. There is no relative rotation.

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[dpc]

There should be a NEMA starting code on the nameplate. I don't think a single-phase measurement is going to tell you much, but I could be wrong.

I'm assuming you don't have access to three-phase power? 6 times the full load amps will be in the ballpark, unless it is a special motor.

You also need to consider the asymmetrical inrush current, depending on what you are trying to figure out.

 

[ScusaMe]

Uh-Oh!

"jraef (Electrical) 11 Jul 12 15:52
If you apply single phase power to a 3 phase motor that is not moving, it will not start. There is no relative rotation."

U-m-m-m ... close, Jeff; but no cigar.

You can get the motor to rotate IF ... you apply an external rotational force to the uncoupled (no-load) motor shaft. In fact you can even get it to accelerate in the reverse direction IF you spin it in the reverse direction. Of course ... you aren't going to obtain optimal operation from it .... but ... it can be made to rotate, Jeff. I've even seen some motors that had a modicum of magnetization in the rotor, take off on their own ...

 

[dpc]

If mechanically spun up, it might rotate until it overheats and fails or trips the overload relay due to single-phasing.

 

[jraef]

...by applying single phase power to the motor while stopped...

I see nothing about spinning it first.

I'm aware it MIGHT spin in an unpredictable direction, my point was that the initial premise of his query was flawed from the outset and unless that was corrected, the rest of the information would be at best hypothetical and no matter what, virtually pointless other than as a discussion of "what ifs".

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[electricpete]

I tend to agree with dpc. It will be very tough. For one thing, I'll bet you are planning on using a reduced voltage. Lots of non-linearities to consider. Even at full voltage, the differences between 2-phases energized flux pattern and 3-phases energized flux pattern is substantial and again many non-linearities would make it very difficult to predict.

There may be other ways to skin the cat as dpc also mentioned. That code letter on NEMA motor nameplate, sometimes called starting code or kva code. Different than the NEMA design letter (which is usually B for general purpose motors). eplate,





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

 

[BigMotorGuy]

Guys,

All good information, but perhaps a little off from my original intent. I am not proposing to run the motor on single phase (been there, done that....not good)but rather just trying to get to an actual measured starting 3 phase current measurement. These motors are special purpose designed/built motors for cryogenic fluids, so NEMA ratings do not apply. I will be testing these motors using VFD drives (only option for testing) but the end application will actually be direct on line. The customer wants a true inrush current measurement (as opposed to just calculated values) for DOL drive and the VFD drive just does not do a good job at getting to the actual inrush current (inrush is damped due to limited ramp rate of inputs). I remember there being a procedure for applying full rated voltage to a single phase on a motor (yes the motor will not start nor do I want it to) and multipying the measured current in the single powered phase by some constant to estimate the full 3 phase inrush current under normal operating conditions. This might be in NEMA but could not find it.


Thanks for all the input!

 

[electricpete]

I found the following in IEEE 112

7.2 Locked-rotor tests
7.2.1 Current
This test may be performed either to check for quality or to determine performance. When possible, readings
shall be taken at rated voltage and frequency since the current is not directly proportional to the voltage
because of changes in reactance caused by saturation of the leakage paths. When the test is made to check
the quality of squirrel-cage machines, it is possible to omit the mechanical means of locking the rotor by
applying single-phase power of rated voltage and frequency to any two of the machine line terminals of a
three-phase machine. With a three-phase machine, the line current will be approximately 86% and the power
input will be approximately 50% of the corresponding values obtained with polyphase power. The values so
obtained may be compared with those measured on a duplicate unit that has been subjected to a complete
test.

So it sounds like you will have to multiply your measured single-phase locked rotor current by 1/0.86 to get an estimate of three phase locked rotor current.

Note (at least in my view of terinology), locked rotor current does not include that initial decaying dc component. "Inrush current" does. And would be trickier to measure since it depends on timing - angle of closing with respect to the voltage source.

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

 

[electricpete]

One can infer how they came up with the 86% number. A simple model of a motor as three linear impedances connected in wye.


In 3-phase starting, the voltage accross each leg is VLL/sqrt(3).

In 1-phase starting, the voltage accross each leg is VLL/2.

The ratio single-phase/three-phase would be (1/2) / (1/sqrt3) = sqrt(3)/2 = 0.86

I'm a bit skeptical of that linearity assumption (see my previous discussion).

If you ever get any data, I'd be interested to hear how it comes out.

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

 

[BigMotorGuy]

Thanks electricpete.......this is the correlation I was looking for. I agree that the linearity assumpion is suspect, but at this point it is all I have to go on. I will try to have the customer take some high speed data once it is hooked up to his system to verify the assumption........of course the data will be dependant on the voltage sag at site, so some variation site to site will invariably be there.

 

[waross]

And if the motor is delta connected?? (I hope not).

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

 

[electricpete]

Interestingly, if we apply similar simplistic logic to delta winding (consider constant impedance Z in each leg of the delta), we get the same ratio:

For three phase starting:
ILEG = VLL / Z (where ILEG is current in the delta leg).
Iphase = sqrt3 * Ileg.= sqrt(3) * VLL/Z

For single phase test, we get one phase which contributes ILEG and the other two in series contribute ILEG/2. giving
Iphase = ILEG + ILEG/2 = 3/2 * ILEG = (3/2) * VLL / Z

Ratio single-phase/three phase = [(3/2) * VLL / Z] / [sqrt(3) * VLL/Z] = sqrt(3) / 2 = 0.866

In addition to non-linearity, there is the question of interaction between phases. The voltage induced in a given phase depends not just on current in one phase but on current in the other two phases. So the model should include both self and mutual inducatances in all systems which are not balanced (or in sequence network which is a superposition-sum of balanced systems). Since it’s in an IEEE standard, it’s probably right (as far as phase interaction goes). I’d just have to spend a little time to prove it to myself.

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

 

[PersonalProfile]

As suggested above you could use no load / locked rotor test etc, to determine the equivalent circuit from which you can calculate the starting current.


Regards,
Lyle

 

[electricpete]

My reading is that locked rotor (3-phase) is the test that couldn't be done.

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