Automatic Capacitor Bank - CT location 2

[rmt131]

Hello all,

For installing an automatic capacitor bank, I was wondering if it makes a difference where in the power system the sensing Curret Transformer (CT) is located? Does it matter if it's 'upstream' or 'downstream' of the capacitor bank tie-in? Should it be right next to the capacitor tie-in or can it be somewhat removed from it?

The reason I ask is I have one system where a capacitor bank was added to an existing system. Due to space issues, the tie-in for the capacitors is through a 'sub' MCC which is being directly supplied from the main site MCC (all 480V, max 400-600A rating). I was told that the CT for the capacitor bank should be located as close to the supply transformer as possible, so we placed the sensing CT on the main site MCC supply, some distance (approx 10m) from the capacitor tie-in point. DLRO tests show good connectivity between the main MCC buss and the 'sub' MCC buss.

This capacitor bank is giving us some problems in comissioning and I have reason to suspect that the power factor isn't being sensed correctly, so I just wanted to be sure that placing the CT so far physically from the capacitor tie-in isn't causing our problem.

Comments appreciated!

Thanks,
rmt131

 

[windie]

I would think that the CTs and PTs for controling the capacitor bank should be as close as practicable to the point that you need the power factor to be controled. That is, if you need to control the power factor at the tie to the utility, then install the CTs and PTs as close as you can to the utility tie point - if this is not practical, then you may need to compensate for any devices that use significant amounts of reactive power (transformers, induction motors, reactors) between the point that you are measuring at, and the tie to the utility.

 

[rmt131]

Thanks for your input, windie!

I didn't consider the fact that the correction bank controller needs to see both current and voltage at the same point to detect the power factor (duh...). There are also quite a few medium-HP motor branch circuits physically tied into the MCC buss between the CT location and the capacitor tie-in point, so I could see there being a slight difference in instantaneous waveforms between the two points imposed by the inductive loads. Since the controller assumes the measurements are being made at the same location, and calculates it's measured power factor accordingly, no wonder nothing's coming up right!

We are in the process of changing the CT location to match the capacitor tie-in location (unfortunately, moving the tie-in closer to the utility is not feasible), and we'll see if this helps any. If I can get the bank switching in and out as I expect it to, then we'll see about adjusting it for being so far removed from the utility...

I'll let you know how it goes.

Thanks again,
rmt131

 

[rmt131]

Aha!

Further site investigation revealed that the CT was on Phase B rather than Phase A, while the controller was hooked up to monitor Phase A for voltage. No wonder the controller kept trying to switch them all in...

Phase B was chosen by the field installation team because it was the most heavily loaded phase (technically correct procedure). Unfortunately, they then didn't realize that this 'shift from normal' has to be reflected in the phase rotation/connection of the capacitors... oh well, it seems to be working now...

Thanks for your help!
rmt131

 

[mc5w]

But, locating the CT and PT at the sub MCC means that the bank can really only absorb the reactive power that is connected to the sub MCC and not create a condition of capacitive power factor at light load.

An alternative is to put a clamp-on CT on one of the service wires. You are also allowed to have an extra service switch beyond the 6 disconnect rule for power monitoring equipment ( which the capacitor controller is ) which would allow you to have the PT in the same area.

If you locate the CT at the service and the PT at the sub MCC you should not have too much error if your voltage drop is reasonable. In this configuration you might have to tell the controller a slight "lie" as to the desired power factor just as long as it disconnects all capacitors at light load.

 

[rmt131]

Thanks mc5w!

I forgot to mention, we left the CT at the utility entrance, we just switched around the phases of the capacitor tie-in (therefore, the PT phase) at the sub-MCC. So the controller is detecting all reactive site current, and since the DLRO testing between the main MCC and the sub-MCC was good, the capacitors shouldn't be seeing much of a voltage drop from the utility.

If we start having problems during light loads, however, I'll keep in mind to tweak the target Power Factor on the controller.

Thanks again!
rmt131

 

[mc5w]

At light load a capacitive power factor can cause a voltage swell that will break stuff particularly electronics. The capacitor controller should have an overvoltage relay that will disconnect the bank during a voltage swell.

Good Luck!