Fixed Speed Induction Motor Harmonics

Hello, I am wondering if it is possible for an induction motor in a system could produce harmonics? I had often considered induction motors a non-linear device kind of like a rotating transformer, but I'm told they should only be considered a linear device.

The system in question is an oil well using an electric submersible pump (induction motor driving a fluid pump). The motor is provided power through a three phase cable, let's say about 2500-3000 feet long. The cable is connected to a switchboard at surface, that is fed via a step down transformer from a 60Hz line.

The system voltage is around 1400VAC, and the amps are around 30 amps. The man taking measurements has recorded 3.96% VTHD and 12.4% ITHD. The current harmonic distortion is made up almost entirely 5th order. From what I understand, there are several sites similar to this one on the power line, but he is only seeing this result on a few of the sites. We want to know why the ITHD would be this high. I've attached a photo of the measurement trend.

We have considered possible saturation as an issue, is there anything you folks would check out? Let me know if more info would be helpful!

SWB harmonics.jpg
The simple answer is yes! But it's normal, especially, if the liquid is oil. Unfortunately, it's virtually impossible to eliminate (but can be reduced somewhat) and it should be considered when sizing the pump-motor! But, the most important factor is voltage at the motor!
If you would like additional information, then please answer the following:
1) What is the Down-hole cable construction like? Is it 3 x single conductors, 1 x 3 conductors arranged in a flat configuration. or 1 x 3 conductors arranged in a triangular configuration?
2) Size of motor in HP?
3) What is the motor on-load Amperes compared to its nameplate rating!
Regards, Phil Corso ([email protected])
Ps: An-aside... Deepest oil well I worked on was 10,000 feet... in the Florida Everglades! Now you know oil has been taken from Florida for almost 75 years!
Phil, thank you for the quick response. What we know is that the motor is just over 50% loaded and uses flat 1x3 cable.
I had the depth way off, it's actually closer to the 7800ft figure.
1) Did you know the the diameter and of both pump and motor are in the order of inches, while their lengths in 10's of feet?
2) The cable is not symmetrical, that is, their core to core separations are not equidistant !
3) The core distances to "ground" are also asymmetrical !
4) The ambient temperature is in the order of hundreds degrees.
5) The ambient pressure is well above normal, depending of the location and depth of the oil-find!

Short story, the motors are not "typical", and electrical demands lead to the phenomenon called "Negative Sequence" thus, resulting in the "harmonics" noted. BTW, Negative-Sequence is calculable, easy math using ANSI/IEEE, more difficult math using IEC !

The operation noted meets the ANSI/IEEE and IEC standards covering Voltage limits ! There is no such standard for Amperes. And, both cause elevation of the motor's carcass temperature! But, they are designed to operate under the conditions noted! Many attempts have been made in the past to come up with ways to control source phase-to-phase voltages on the surface to minimize Negative-Sequence ! Some successful, most not!

My recommendation has always been to ensure Surface-to Down-Hole cable voltage-drop, should never result in the motor voltage being below it's nameplate rating ! One way is to calculate the required size... then select the next largest cable size ! The other way is to make sure the supply can be "adjustable" via taps on the supply transformer !

Think of the results of motor failure... 7,800 ft of cable, motor, and pipe has to be removed, and then put back in th hole! Ain't cheap !

Regards, Phil Corso
Phil, thanks again for the great reply.
Yes, I'm very familiar with the ESP assembly and proper surface equipment setups, etc. Also familiar with the quirks that come with thousands of cable feet and long, skinny motors! Been pulling, running, and repairing the things for years. But, I have not dug deep on the power quality elements but for the last 2-3 years or so.

Our main "puzzle" was the discovery that the motor being powered by the switchboard was showing unexpected 5th order harmonics. We've come to expect that kind of thing on VFDs, but didn't expect it on a switchboard. Although we recognize it's not a real problem, we are more just curious on what might be causing the harmonics, particularly since it was only seen on a few wellsites and not others.

I think you began to answer that discovery with your "Negative sequence" comments above. When you say that, the negative sequence would imply the braking torque created by the 5th harmonic, right? I'm trying to dig to the "why" the harmonics exist, and not necessarily that the "just do" exist.

My thought was that at a relatively reduced load, the harmonic currents reduce in absolute value, but increase as a percentage of the fundamental waveform. Particularly with current harmonics, at reduced load, ITHD will appear high. So possibly if the motor was de-rated to match the pump load a little better, we could theoretically see that ITHD drop as a percentage somewhat. What is your perspective on this?
cl302i (makes you sound like a Robot)
Your comment about "Braking Action" is correct. My physical representation is that of two 3-ph motors, one big, the other small, connected on the same drive shaft ! The big one is wired properly for direction, the small one is wired in reverse!
I'm also puzzled by the Harmonics chart. Can you provide the actual frequencies ?
Another thing for you to remember... Negative-Sequence also impacts the cable impedance, altering its effective Voltage-Drop! I would ask for six low-V, impedance checks... A-B, B-C, C-A, A-G, B-G, and C-G. And, a vibration check!
Regards, Phil Corso

Yes, these are all very important points. Sometimes there are simpler "solutions" which can cause something to happen in a way you do not expect. I have found, read and saw voltage readings close line to line and still have a large variation phase to ground so sometimes when looking for"noise" I will also take readings or capture waveforms line voltage and phase to ground voltage. Also the ratio of voltage to current harmonics are generally around 6 times so a 5% voltage harmonic can generate a 30% current harmonic and a 5% voltage imbalance and unbalanced current of 50%. It is the current causing the rotating field which is pulling the rotor around. Also, when I was very young I worked with an OLD electrician who switched L1, L2, L3 by moving them all one location in the same direction and his unbalanced current problem was not as bad as it had been before. Maybe he got lucky but it is the old lessons we remember the longest. If you discover any of this to be interesting or informative just ask.