Cause of motor failures 3

[jmcs]

The information I'm looking for would be a chart of the temperature rise and long term effects on the motor insulation of a motor running with a voltage imbalance of .9 to 1.1 percent and current of about 5 percent. I have a customer that has had several motor failures in the past two years (same motor). The motor repair shop reports that I have read say the problem is voltage spikes, transients, lightening, etc. I have run several surveys with a power quality analyzer and have not seen any RMS spikes, etc. even during lightening storms, (motor has lightening arrestors and surge capacitors). I have seen the voltage imbalance of 1% and current of 5%, plus waveform faults with peak voltages of 580 volts phase to neutral (480Y/277 volt system), but they only last one cycle, sometimes 2 or 3 of them within a 15 cycle time frame. I think it may be capacitors switching on the primary. There is also a system voltage drop of approximately 5% every 30 to 40 minutes as the result of another motor starting which causes a 20% drop in current on the motor in question followed by the inrush current as the motor regains speed. The motor current also fluctuates up and down 4-6 amps quite rapidly. I have also caught a few partial phase losses which would single phase the motor for a short time. The question I have is could all of these things combined over time deteriorate the insulation to the point where a small voltage spike destroys the motor?


If anyone is interested in seeing them, I have bitmaps of voltage and current time plots and the waveform faults.

Thanks to anyone who can help

 

[electricpete]

Motor current imbalance is typically 5-7% of the voltage imbalance that caused it... so your 5% current balance and 1% voltage imbalance numbers match well.

EPRI and Richard Nailen's books will tell you that the temperature rise due to an imbalance is approximately twice the square of the voltage imbalance. 1% voltage imbalance typically would increase the temperature rise from by 2%. A 1% derating in motor load (current) would compensate for this effect on an overall heating level (although local and rotor heating might remain a concern). That is apparently the basis for the NEMA derating curve which shows 1% derating for 1% voltage imbalance. (4% derating for 2% imbalance, 9% for 3%, and 25% derating for 5% voltage imbalance).

That does not seem to agree with the way I would compute the derating... namely a 5% current imbalance should result in approx 10% increase in heating on one phase which should require perhaps 5% current derating to maintain life. I'm not sure why the difference.

Please note that as I have used it above the term imbalance is defined as max(|Vi-Vaverage|)/Vaverage. It is also usually approximately 1/2 of the highest minus the lowest reading expressed as fraction of Vaverage.

There are a few more exotic formulae you can use to get a really good estimate of the "negative sequence" component based on the three magnitudes (assuming symmetrical phase angles), which is the more relevant number. Imbalance as computed above is used only as an estimate of negative sequence. Let me know if you want that formula.

The change in life based on EPRI/NEMA/Nailen approach for 1% imbalance is fairly modest. But if you want to sharpen your pencil and compute a mathematical change in life based on thermal effects alone, then you can calculate the increase in temperature rise as above and use the 10-degree rule... insulation aging rate doubles for every 10C increase in winding temperature.

In transformers it is well known that thermal aging often weakens the transformer insulation mechanically over time and then the winding often fails during a through-fault. The cause was not really the through-fault but the thermal aging. This is probably the most common failure scenario for transformers.

Similar scenario's certainly could apply to motors (thermal aging over life increases susceptibility to failure during a spike). In particular thermal aging can allow the windings to start to loosen and vibrate, this could degrade the turn-insulation. Turn insulation is usually what fails during a voltage spike. I have never heard of this scenario and I'm not sure how likely it is.

Certainly your voltage spikes are cause for concern. I guess there's lots that can be said about attacking that problem... I'll only suggest that if you can't eliminate the voltage spikes, consider surge supression for larger or more critical motors. You might also give some thought to what kind of requirements you can put in your motor purchase/rewind specification to get motors less susceptible to this known stressor in your environment.

 

[jmcs]

Electricpete,
Thanks for the info. Something strange about the current imbalance, it exists even when the voltage is perfectly balanced. Any thoughts?

I would like the formula. Thanks

 

[electricpete]

During voltage sags, the motor current should go up.

Regarding current oscillations, that can sometimes be associated with broken rotor bars. But for smaller (460v) motors this is not common...and certainly not likely for multiple small motors. Do you attribute the current oscillations to voltage oscillations?

 

[rhatcher]

The motor repair facility may be guessing, but in a lot of cases a pretty solid determination of the cause of motor failure can be made based on the condition of the windings. For example, any overload related failure (includes single phase and voltage imbalance) will result in blackening of the windings, while surge or spike related failures result in a localized fault resembling a hole blown in what would otherwise appear to be a good winding (usually in the end turns). This can be seen turn to turn or phase to phase, although admittedly it is sometimes hard to tell based on the size of the hole. I would recommend that you ask the shop how confident they are of their determination before spending more time looking for other causes.

Any of the conditions you mentioned will have a detrimental affect on winding life. However, the 580V L-N that you measured could be a good candidate. Did you measure it at the controller or the motor terminals? I ask because a fast rise time surge like that may cause a reflected wave in the leads that increases the voltage at the terminals to an even greater level.

 

[electricpete]

My first words in this thread should be corrected to read:
"Motor current imbalance is typically 5-7% times the voltage imbalance that caused it..."

The formula is at work... I'll get it tomorrow.

A current imbalance (especially 5%) in the absence of voltage imbalance would normally suggest a motor problem (or at least some problem in the circuit downstream of the point where you're measuring the voltage).

You mention that it is the same motor that has failed several times... but are there other motors at the same site that have been working well? If yes, that would lead to question what is unique to this motor.

If current on the motor in question is really going down when voltage goes down, that is something quite unusual which demands further investigation.

 

[jbartos]

Suggestion: References:
1. Robert W. Smeaton, "Switchgear and Control Handbook," 2nd Edition, McGraw-Hill Book Co., 1987 indicates in Section 22 "Voltage Unbalance" on page 11-56 the following table pertaining to the induction motor:
Voltage, Unbalance % 2.0 3.5 5.0
Current, Sator, rms % 101 104 107.5
Stator average loss increase, % 2 8 15
Rotor average loss increase, % 12 39 76
It is actually the motor rotor which experiences highest increases that can be made corresponding to temperature increases.
2. J. Linders, "Effects of Power Supply Variations on AC Motor Characteristics," 1972, p 583 is also applicable.

 

[electricpete]

jbartos - good reference.
do you know what they mean by "Current, Stator, rms"? Is that the average value of stator current or the highest phase current? I'm assuming they mean average. That is yet a third formulation for predicting unbalance, and no two seem to agree.


Here is a little bit of Nailen's discussion followed by his formula. I believe he assumes that zero sequence component of the voltage is zero and the phase angles are balanced (120 degrees).

"To avoid the trouble of figuring out the actual negative sequence voltages or currents, NEMA proposed many years ago a simple formula which gves answers close enough for general use. It is
% Voltage imbalance = 100*(max deviation from average voltgage)/(average voltage)

Another simple formula for voltage imbalance gives a still closer answser. It is
% voltage imbalance = 82*sqrt(dVa^2+dVb^2+dVc^2)/(average voltage.... where dVa = Va - Vaverage,"

(dVb and dVb are similar).

 

[jmcs]

Electricpete,
The motor is a 100 hp deep well submersable pump. When the voltage dips as a result of the 50 hp starting the current also dips, but then increases slightly over typical running current then back to typical, which is oscillating. I thought the oscillations were possibly load related but it's a pump that should have a steady load. The building is fed by three 50 KVA transformer connected Y on the secondary with a solid ground. The only motors are the 100 hp, 50 hp, and two smaller aeration fans. During a single phase sag the current on the low phase decreases and the current on the other two phases increase well above FLA.

rhatcher,
The 580 Volt L-N is a peak measurement taken on the Line side of the soft start feeding the motor. It happens fairly often and is associated with the waveform faults that I thought were possible capacitors switching. Someone observed that if that was the case I would only see one fault and there is sometimes 2 and 3 within a few cycles of each other. It must be another type of switch fault.

jbartos, Thanks for the references.

 

[electricpete]

It's an interesting sounding problem.

Is the 50hp motor a single-phase load? Or is the source of the voltage sage something other than start of the 50hp motor?

If you'd like to email your bitmaps to me, I can post them on my webpage so they'll be available for everyone on the board.

email me at electricpeteNOSPAM@technologist.comNOSPAM.
(take out the NOSPAM's)

 

[rhatcher]

I found that page I was looking for.

http://www.joliet-equipment.com/winding_damage.htm

Try it out. It pretty clearly shows the condition of motor windings after different types of faults and has links to some other good info.

 

[jmcs]

rhatcher, Thanks for the link.

electricpete,
I've e-mailed you the bitmaps. Let me know what you think. I've got alot more of them but I think there is a example of most accurrences included in what I sent. The 50 hp motor is 3 phase. Every 30 to 40 minutes the voltage dips because of the 50 hp but there are also other voltage fluctuations and single phase sags not caused by anything at this facility.

 

[electricpete]

jmcs' bitmaps can be found in zip file at:

http://geocities.com/pschimpf/temporary_stuff/well.zip

You'll have to download it to your hard-drive and unzip it.
Sorry - my site doesn't have enough space to store the unzipped files (even when converted to jpg).
I did run a virus scan on these bitmaps and didn't see any problems (I'm not even sure if bitmaps can have viruses, but I checked anyway).

 

[electricpete]

Here is a link to a page with Jeff's files in jpg format. This will avoid the need to download any files, but it might take you a little longer to view all of them if you're on a slow connection.

http://geocities.com/pschimpf/temporary_stuff/jeff_pq_filelist.htm

 

[jbartos]

Suggestions:
The values considered vary. The one I posted above in my previous posting for Stator Current is rms, which means a normal continuous Irms.
Please, notice that voltage and current unbalances are undergoing much bigger scrutiny, in general, than addressed in here, what could be called one approach to the voltage or current unbalance. When, it comes to any deviation from this, it supposed to be explained.
Visit

http://www.access1.net/rwgd1900/VoltageUnbalance.htm

http://www.epri-peac.com/shared/tutorials/brf28tut.html

http://www.epri-peac.com/shared/tutorials/brf29tut.html

http://www.wisconsinpublicservice.com/business/phase.asp

http://www.usbr.gov/power/data/fist/fist3~24/3~24_1.htm

(average voltages addressed)

http://www.oit.doe.gov/bestpractices/explore_library/energymatters/sep1998_root_cause.shtm

(NEMA voltage unbalance curve is included)

http://www.reliance.com/mtr/b7087_5/b7087_5_7.htm

(check Figure 25 for voltage unbalance)

http://www.execpc.com/~hgmurphy/SYSCPAT1.HTM

http://www.pge.com/002_biz_svc/002c1c_understanding.shtml#balance

(See ANSI C84.1(1) for 3% voltage unbalance limit)
etc.

 

If you are using a soft start, check that out. There are many different kinds and some can cause problems when not working correctly.

 

[jmcs]

electricpete,
What is your opinion of the power analyzer data?


jbartos,
Thanks for the qreat links

chrishart,
The softstart was installed by a contractor to solve the problem. There have not been any motor failures since,but it's only been about 7 weeks. I don't believe it is the solution. Any thoughts?

 

[jmcs]

Has anyone had time to look at the bitmaps or jpegs and if so what is your opinion on the cause of motor failure? Am I trying to make something of nothing or could all of these disturbances combined be related to the motor failures?

Thanks,
Jeff

 

[electricpete]

Jeff
Sorry I haven't devoted much time to the waveforms... also power quality isn't my specialty. My gut feel is that you've got some serious power quality problems. Other people may have some more specific info including standards which govern this (although I think that maybe it's not well defined).

Perhaps you could repost your message on the power board with links to the files. There might be a wider audience there.

An even wider audience can be reached at alt.engineering.elctrical.... link is:

http://groups.google.com/groups?oi=djq&as_ugroup=alt.engineering.electrical

and
misc.industry.utilities.electric,,, link is:

http://groups.google.com/groups?oi=djq&as_ugroup=misc.industry.utilities.electric

 

[jmcs]

Thanks for the links. I'll try them.