Capacitor Bank Harmonics

[DTan]

What sort of harmonic current levels are going to cause capacitor failures in power factor correction banks? We've had a lot of capacitors fail recently and are trying to figure out why. According to a Fluke 43 power quality meter, the capacitor bank had a voltage THD of about 1%, and a current THD of about 7%. 480/277V system. The capacitors that have failed have ranged in age from ones that are 10+ years old, to ones that are only a few months old.

 

[GoldDigger]

My guess is that the damaging factor is the extra current, although mechanical forces ( vibration) could be an issue too.
For third harmonic content, the current for a given harmonic voltage will be three times the current as for the same voltage at the fundamental frequency.
But higher harmonics are proportionally worse.
So it would be a big help to get actual harmonic levels instead of just THD.

Tapatalk!

 

[DTan]

Here's what the meter said. THD varied between 6.7-7.0% while I was taking the readings
RMS: 160.8
1: 160.5
3: 0.5
5: 7.5
7: 4.7
9: 0.9
11: 5.2
13: 1.7
15: 0.1
17: 2.0
19: 0.9
21: 0.3
23: 0.8

 

[GoldDigger]

This is the current reading, yes?
The high content of higher harmonics, all the way up to 11th concerns me, but I do not have the expertise to interpret them in terms of potential damage.
Waiting for someone else to jump in....

Tapatalk!

 

[junkhound]

With the part number of the failed cap, go to the web (or directly to the mfg) and get the ESR vs. frequency of that cap, many cap mfg also supply max current ratings for different frequency ranges.

Use the cap and harmonic data and put into a FEA analysis program like PSpice or similar and see how much heating you are getting.

Edit ps: if the caps are on the end of a long line or on the output of a high leakage impedance transformer also see if you happen to be sitting at a LC resonant frequency, that can drive the voltage above the cap rating.

 

[Besoeker]

Here's what the meter said. THD varied between 6.7-7.0% while I was taking the readings
RMS: 160.8
1: 160.5
3: 0.5
5: 7.5
7: 4.7
9: 0.9
11: 5.2
13: 1.7
15: 0.1
17: 2.0
19: 0.9
21: 0.3
23: 0.8

OK.
Got the kVAr and voltage for the unit you got these readings on?
Just as a comment in passing, 5th, 7th, 11th, 13th etc (6n?1) harmonics are those usually resulting from variable speed drives.

 

[GoldDigger]

Would line side reactors be useful if
A. That is the problem and
B. They are the source?
Or would it be better to address the problem at the capacitor bank?

Tapatalk!

 

[DTan]

Those were the current harmonics.

Here's the nameplate of the bank. 125KVAR at 480V.



It's a single stage controlled by an external contactor that doesn't seem to be chattering and is on more or less constantly since we don't have enough correction capability to correct the service beyond about 85-90% when production equipment is running.

There are a fair number of VFDs (almost all under 10HP), but they don't make up a large portion of the total load. Maybe 10% of the total.

 

[Besoeker]

Those were the current harmonics.

Here's the nameplate of the bank. 125KVAR at 480V.

It's a single stage controlled by an external contactor that doesn't seem to be chattering and is on more or less constantly since we don't have enough correction capability to correct the service beyond about 85-90% when production equipment is running.

There are a fair number of VFDs (almost all under 10HP), but they don't make up a large portion of the total load. Maybe 10% of the total.

From the nameplate data the measured PFC Irms is somewhat greater than the nominal current that would flow.
Even the measured fundamental components is greater.
Harmonics really wouldn't account for that.

On the subject of VFDs, if all are on the same supply with no phase shift, the effects are cumulative.