Capacitor Bank failure mode?

[steve66]

The caps were off line.

2nd to the last post on page 6.

 

[mxslick]

OK I have leaned some more, I ended up talking with the guy who installed this capacitor and was there when the fault happened. The reason I could not find a problem is that the bad capacitor bank has already been removed and the one I was looking at has never been energized.

So what was the original failure mode? That should give you some insight as to the safety of re-energizing this new bank.

Since the motor loads cycle on and off, it may be possible as was mentioned by others that the right combination of running and not running motors causes a resonant condition. That's why individual PF correction at each motor, though costly and somewhat impractical, may be the way to go. Either that or put up with repeated failures of this cap bank.

Like others have said it looks like the 200Kvar of this bank is about right for the application.

If I was in your position, I would hook the bank back up and throw the breaker..from a safe distance just in case.

The PQM will give an instant confirmation from the power factor readings if it's working right.

The harmonics are surprisingly high. Where would they originate from in a relatively benign system?

How about from all the computers in the checkstands and the barcode scanners and inventory systems? And maybe even electronic ballasts.

 

[iaov]

At the point of resonance, the voltage across the capacitor rises towards infinity causing the capacitor to fail. It is a frequency thing not an applied voltage issue.

Is this a parralel resonant circuit being formed by load inductors(motors) and the cap bank??

 

[iwire]

I took it that the wires were disconnected from the breaker and Bob was checking things out and getting some suggestions as to what to look for/check prior to hooking it back up.

That is exactly the case, I have not switched the capacitor on at all yet.

You all have convinced me to do my 'due diligence' here and go the extra mile ............ besides I have the spare time too.

 

[iwire]

So what was the original failure mode?

Breaker turned on for the first time and immediately there was fireworks, then darkness then the generator came on line. Then my to buddies had to go change their shorts.

That's why individual PF correction at each motor, though costly and somewhat impractical, may be the way to go.

But it is not an option that is even remotely going to happen.

Either that or put up with repeated failures of this cap bank.

After talking with the guys on the scene it is my opinion the cap was shipped with a fault and my buds where lucky they closed all covers tight before they turned it on.

Cutler Hammer replaced the unit free of charge without explanation of the cause.

If I was in your position, I would hook the bank back up and throw the breaker..from a safe distance just in case.

That is pretty much where I am heading but I will see what data is recorded for a week.

How about from all the computers in the check stands and the barcode scanners and inventory systems? And maybe even electronic ballasts.

Pretty new store, mostly florescent lighting with electronic ballasts along with all the other items mentioned.

 

[LarryFine]

Breaker turned on for the first time and immediately there was fireworks, then darkness then the generator came on line. Then my to buddies had to go change their shorts.



But it is not an option that is even remotely going to happen.




After talking with the guys on the scene it is my opinion the cap was shipped with a fault and my buds where lucky they closed all covers tight before they turned it on.

Cutler Hammer replaced the unit free of charge without explanation of the cause.



That is pretty much where I am heading but I will see what data is recorded for a week.




Pretty new store, mostly florescent lighting with electronic ballasts along with all the other items mentioned.

Bob, if you want to do a bit of work, you could always separate the leads, energize one cap (or small group) at a time at first, then connect them all back together.

That would minimize the *POOF* factor, if you know what I'm saying.

 

[JWCELECTRIC]

Breaker turned on for the first time and immediately there was fireworks, then darkness then the generator came on line. Then my to buddies had to go change their shorts.



But it is not an option that is even remotely going to happen.




After talking with the guys on the scene it is my opinion the cap was shipped with a fault and my buds where lucky they closed all covers tight before they turned it on.

Cutler Hammer replaced the unit free of charge without explanation of the cause.



That is pretty much where I am heading but I will see what data is recorded for a week.




Pretty new store, mostly florescent lighting with electronic ballasts along with all the other items mentioned.

Don't forget your PPE before you throw the breaker, and at a last ditch effort have your buddies tie a rope around you to pull you out if it gets real bad (Tip from old time Electrician).

 

[mivey]

The harmonics are surprisingly high.

As compared to what? The voltage THD looks ok and as for the currents, I suspect when iwire gets the ISC/IL ratio, that they will fall close to, if not within, acceptable limits.

 

[mivey]

Once you get the available short circuit current (I_SC), you can consider what is acceptable using IEEE 519 standards for THDv & THDi (actually the recommendations at the service point):

Voltage harmonic limits for 69 kV and below: Individual Harmonic = 3%, THDv = 5%

Current harmonic limits for 120v to 69kV as a percent of maximum load for different I_SC/I_load ratios:
I_SC/I_load < 20: for odd harmonic numbers below 11 = 4%, TDD = 5%
I_SC/I_load between 20 & 50: for odd harmonic numbers below 11 = 7 %, TDD = 8%
I_SC/I_load between 50 & 100: for odd harmonic numbers below 11 = 10 %, TDD = 12%
I_SC/I_load between 100 & 1000: for odd harmonic numbers below 11 = 12 %, TDD = 15%
I_SC/I_load > 1000: for odd harmonic numbers below 11 = 15 %, TDD = 20%

I'm sure your meter will give you the harmonics details, and may even find TDD for the time period measured. If not, THDi is given as a percent of the fundamental, I1. We know that I1 = Irms / sqrt(1 + THDi^2 ). If we figure that the THDi is almost completely made up of the frequencies of the 50th order and below, we can get:

TDD = THDi * Irms / sqrt(1 + THDi^2 ) / Maximum fundamental load current

Using your data, and 550 amps for the max as an example, you get:
TDD_a phase = 0.05993 * 519.1 / sqrt(1 + 0.05993^2) / 550 = 5.646%
TDD_b phase = 0.06891 * 484.2 / sqrt(1 + 0.06891^2) / 550 = 6.052%
TDD_c phase = 0.05942 * 509.6 / sqrt(1 + 0.05942^2) / 550 = 5.081%

There are many ways to get the maximum fundamental load current. One way is to estimate it using the historic billing data. You could use a sample data set taken during the period of maximum load. You might also figure the feeder is sized for the maximum load, but that may not always be the case (like when built for expansion or equipment has been removed).

 

[mivey]

When you complete your data gathering, you should have some charts showing the order of the current harmonics. As Jim mentioned in an earlier post, you want to make a resonance check, but I think he left off the square root.

The resonant harmonic order = h = sqrt(kVAsys / kvar)

where kVAsys = the short circuit kVA at the capacitor
and kvar = the capacitor kvar

If h is near one of the major contributors (like the 3rd, 5th, 7th, 11th harmonic), then the capacitor bank will increase the distortion.

 

[mivey]

That is pretty much where I am heading but I will see what data is recorded for a week.

zzzzzzzzzzzzzzzzz

 

[Jraef]

Removed... failed to notice that there were 10 pages and my comment was already addressed...

 

[Jraef]

Different issue.
You mentioned that this is a new store and only has typical fixtures and refrigeration. But a lot of newer refrigeration compressors are now coming with VFDs built in by the manufacturers. If so, they will be major sources of resonnance. You may have to open up the control cabinets to see if they do, I've noticed that some don't adequately inform people of what's inside, they think it's some sort of super trade secret or something.

 

[mivey]

I've noticed that some don't adequately inform people of what's inside, they think it's some sort of super trade secret or something.

That is funny how a manufacturer will do that. I just saw a "How It's Made" episode where they said the # of stitches the manufacturer put in the glove was a trade secret.:-? Ummm, here's another trade secret: most of us can count.

As far as electric equipment, most of us can look inside. Besides, it's not the end user who will spill the beans. The competition, who is well trained at reverse engineering, can figure it out anyway. Why not help the end user and give them the information that may be tough for them to get, but relatively easy for the competition to get.

Anyway, back to the regularly scheduled program.

 

[iwire]

zzzzzzzzzzzzzzzzz

I have the data but have not had a chance to do anything with it, they (the office) shipped me off shore for a couple of weeks. (OK only 8 miles of shore but it is a world of difference.)

 

[iwire]

Different issue.
You mentioned that this is a new store and only has typical fixtures and refrigeration. But a lot of newer refrigeration compressors are now coming with VFDs built in by the manufacturers.

No VFDs in this store and the stores I work in that have them they are almost all in bypass mode by the end of the first year.

 

[zbang]

I just saw a "How It's Made" episode where they said the # of stitches the manufacturer put in the glove was a trade secret.:-? Ummm, here's another trade secret: most of us can count.

They can call it a a trade secret all they want, but unless they actually try to hide something, 'tain't one. That's why some manufacturers will put decals on top of common ICs on their circuit boards. Don't ask why I know these things .

 

[iwire]

OK guys you all thought this was done. :smile:I have finally downloaded the data and would like to get some opinions about what it shows.

I have saved it as a word document for veiwing here but it is too large so I would be glad to email it to anyone that is intrested.

 

[techntrek]

No VFDs in this store and the stores I work in that have them they are almost all in bypass mode by the end of the first year.

Curious as to why?

 

[iwire]

Curious as to why?

As far as the refrigeration contractors have told me it will either be because the unit broke down and the repairs where deemed an unnecessary expense or in other cases they could never get the control programing set up satisfactorily.