Power Factor Capcitors on Low Voltage Power System

[sjmgd997]

I have a situation where an engineer has placed a 3 phase automatic tuned capacitor bank rated for 150kVAR. It is intended for power factor correction.

The service itself is 480V, 3 phase, 3 wire, 1600A. The 1600A switchboard feeds (3) 480V MCCs, (2) of which have VFDs on them.

Here are my concerns:

1) If I wanted to model my power system in simplest terms, I believe what I need are:
a) Equivalent R-L (per-unit values) of impedance for my 3 MCC's all connected in parallel. (2) of the MCCs have variable frequency drives on them, the 3rd MCC is dedicated ONLY to emergency loads and has no drives on it.
b) A voltage source to represent my 480V switchboard (which feeds the 3 MCCs).

I ask this question about modeling BECAUSE we I am concerned with a possible resonance point between the drives (on MCCs), the remaining power system impedance, and my source.

I believe that I could apply an elementary circuit analysis to see which multiple of 60HZ gets us to a resonant point.

2)In this case I have a standby generator. The standby generator does NOT parallel the utility (in that it's a break before make) type of transfer system. Do I have any form of an issue with the interaction between the power factor correction capacitor and/or the drives in that I'd have some form of residual voltage from the power factor correction capacitor that could be impressed on power system components EVEN AFTER the transfer from normal power to emergency power happens?

 

[redfurry]

Others may be able to provide better info, but I suspect that you will have to use a more complicated model.

If you are interested in solving a harmonic problem, you will need to know the impedances of the loads at each frequency (may not be constant if you have VFDs). A single lumped load impedance may not be realistic.

The problem with harmonic analysis in general is that it often requires you to provide input data that is difficult or impossible to obtain

 

[sjmgd997]

I see your point. I figured I could start with a "gross" model of the 3 impedances and then add complexity from there.

The other part of my question is what of further concern- that is, when I switch the power system off and attempt to run on standby power, is there any form of residual voltage I need to be concerned with?

 

[Thealanator]

I had a similar situation. We wanted to use a generator to run a 600 HP compressor in the event of a utility failure. The compressor is equipped with capacitors that switch on a few seconds before the main motor starts. I had understood that generators are not fond of leading power factors. None of the other engineers had any insight, so I decided to attempt a start with capacitors in place. The generator breaker tripped, so I removed the capacitor fuses. The next start attempt was successful.

Since you appear to have an automatically switched unit you may be able to modify the switching logic so that capacitors drop out on loss of utility. If you are doing power factor correction to avoid penalties you certainly don't have to worry about such things while on generator power.

 

[waross]

Will this work for you:
Model your system at a selected load level without caps.
Calculate the VARS.
Re model the system with the amount of capacity that the PFC controller will connect for that amount of system VARS.
Re-iterate at a new load level.
As for the generator, if you can run without caps, do so. If there are issues such as the generator hitting the KVA limit because of poor PF, or voltage drop on long feeders, add enough capacity to correct the problem.
If I were on site I am sure that I could devise a way to fool the PF controller when on generator power.
Without detailed information concerning the PF controller, the loads to be expected, the duty cycles of the various loads and a few other factors I won't even try on this forum.


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