Modelling of capacitors

[Adam1980]

Dear Members,

I am doing a dynamic study which involves load shedding system and a number of filters that remains connected to the network after the bulk of the load is disconnected due to under-frequency. The filters are then disconnected due to overvoltage settings.

During the period of overvoltage and since the filters contribution is large which leads to increase of votlage and hence an increase in filters contribution and so on, the filters are reaching a maximum of 9 folds their rated power because of these events.

Since the filters are modelled as an impedance load this behaviour is expected from the model however how much realistic it is?

My question is:
is there any limitation of the VAR contribution of a C with a definite Farads value? as a consequence does this mean that the filter contribution would increase as a function of U^2 as long as the U is increasing to a certain limit then stop or does it increase as U increases no matter what is the value?

Thank you.

 

[magoo2]

A capacitor is a fixed impedance. There is no nonlinear characteristic. At some value of voltage(U), you'll get a breakdown of the capacitor.

 

[Adam1980]

Thank you magoo2 so it is expected that the filters would have an inverse time protection for the overvoltage correct?
do you have an experience with the settings of this protection?
Thanks.

 

[lume7006]

Hello,

According to IEEE or IEC standards for capacitors, there are limits under which capacitors may operate:
110% of nominal voltage
135% of nominal current
135% of rated power

These limits are for each "can", not precisely for the whole capacitor bank.

Typically, capacitor banks (or filters) have a Y or double Y configuration.
There could be some protection schemes based on:
unbalance on the center of the Y
over/under voltage on the capacitor bank terminals
Overcurrent on the capacitor bank terminals (this protection may be using inverse-time characteristics or also definite time)

If as you mentioned, voltage is increasing there will be a moment in which the overvoltage will exceed capacitor limits and then, shoud be disconnected.

Another aspect to consider is related to nominal voltage of each "can", because it is typically different from the operating voltage. I mean, if you are using a capacitor bank at 34.5 kV (e.g.) then, each can may be of 38 kV, this will depend on the design and manufacturer.
What I want to clarify is that the overvoltage on the system may not be an overvoltage on each can

Regards