Capacitor Bank Feeder Sizing

[HamidEle]

I am sizing a feeder for a capacitor bank( 600V, 180KVar). Per Code, the size of breaker can be sized to be >1.35 rated current (173A). So I chosed 300A Circuit breaker. But I was told by the capacitor bank vendor, the RMS current( considering Harmonic and faults) is 265A. Though 300A is large enough, they prefer to choose 400A circuit breaker since the Maximum RMS current is 379A.
It is a filtered bank with some inductance. I can understand the RMS current is higher than the ratd current when the bank traps certain harmonic, but it should be limited by inductance. But 400A seems to be too high to me.

 

[waross]

A suggestion;
First remember that a capacitor bank increases its current as the square of the voltage. If the grid voltage is 105% of nominal the capacitor current will be 110.25%. But, the impedance of a capacitor bank is also frequency dependent. The impedance of a capacitor at 180 Hz is 1/9 the impedance at 60 Hz, with resulting higher currents. If you have harmonic voltages on the system I would trust the vendor.
A caveat however; If this is an installation where long distances or for some reason a much greater cost to install the larger breaker, I would do some investigating, as you are doing, before going ahead and spending the money blindly.
By the way, if the capacitor bank is an assembly with a rated current, I would tend to use 135% of the rated current. I would trust the manufacturers rating more than the vendor's suggestion.

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

 

[HamidEle]

Waross, thanks for your valuable suggestion.
This filter is tuned to 4.5x60hz. ~10%harmoonic voltage distortion is expected. grid voltage is 110% of nominal.
Wondering if the current flowing in 4.6x60 Hz is considered normal running current?

 

[HamidEle]

Waross,
Considering the inductance in the capacitor bank, the impedance is calculated based on R+j(2*3.14*297*L-1/2*3.14*297*C). The actual current flowing in the filter is less than the current without inductance.

 

[LionelHutz]

It sounds to me that this capacitor bank/harmonic filter is rated 180kVAR when connected to a clean 60Hz, 600VAC source.

I'm not sure if that is the way to rate this type of capacitor bank since I'm not really sure the harmonic trap capability has a kVAR associated with it or how that would be rated.

When trapping harmonics the current will be higher than the clean 60Hz source current. I suspect that you will find the capacitors in the bank add up to more kVAR at a higher voltage but when you correct for your 60Hz, 600V source they are the equivalent of 180kVAR.

So, I would trust the bank/trap manufacturer. They did engineer the capacitor bank.

 

[HamidEle]

Good Night, Lionel,
I agree with you when it comes to trapping harmonic current by the bank. However, if we add an inductance to the bank branch, ETAP Software gave me a lower current than without the inductance, which makes sense to me. It seems to me, the bank can trap some current, but not much due to the impedance distribution in the system. At tunning frequency, the current flowing to the bank should be higher, but it will be limted by the inductance as well. The bank manufacturer came with 280A RMS, including 173A fundamental and Some harmonic. It seems too high to me.

 

[waross]

Based on the manufacturers rating of 280 A, I would size the feeders at 280 A x 135% = 378 A minimum. 400 Amp breaker.
That takes care of work. However, for personal interest can you call the manufacturers engineering department and ask how they determine the current rating of the filter? Please share the information with us if you are successful.

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

 

[LionelHutz]

The very basic theory is that the capacitor/inductor filter is designed to be a lower impedance than the source at a certain frequency. Typically it's designed to be a low impedance at 5 times or 7 times the fundamental frequency since those harmonics are typically the worst offenders in a 3-phase system. If there is a harmonic current at that design frequency then the harmonics current will flow to the trap instead of the source.

Did you model the trap with a non-linear load also on the system? As a minimum, you would need to model the source, the trap and a harmonic producing load (try a rectifier). I think you are approaching this from a "60Hz" only view point.

 

[HamidEle]

Hi, Guys,

I communicated with the vendor and their response is as below,

The 280 Amps expected current is not quite accurate. It
is the sum of the 3 RMS currents from the 3 different filters. It is not a straight linear addition.

We are dealing with 3 different tuning frequencies of the 5th 7th and 11th harmonic orders.

5th 7th 11th
the fundamental currents are
122 40 20
The RMS current without consideration for import harmonics are 137 45 22
Import currents are 161 57 30
please note that the Import, Export RMS and Fundamental currents all
contain the same fundamental current.

The sum of the harmonic currents only would be the square root of the
RMS current squared minus the fundamental current squared
Fund
0 0 0
Export
62 21 9
Import
105 41 22

The import and export harmonic currents would also add as a true RMS number
Sum of harmonic currents = square root of the (export harmonics squared
+ import harmonics squared)
sum of harmonic currents
122 46 24