UNBALANCE PROTECTION OF CAPACITORS 1

[144x]

IS IT POSSIBLE TO USE A MEASURING CT CORE FOR UNBALANCE
PROTECTION OF A DOUBLE STAR CONNECTED CAPACITOR BANK.
I'M ASKING THIS QUESTION BECAUSE A CONTRACTOR CLAIMS THAT
BECAUSE OF HIGHER ACCURACIES FOR THIS KIND OF CT CORE IT IS
BETTER FOR DETECTING SMALL CURRENTS.
ANY IDEA?

 

[scottf]

It is quite common to use a metering accuracy core for unbalance sensing in capacitor banks.

For typical protection applications, relay accuracy cores are sized/designed with the main factor being performance during transients, i.e. system faults, where the CT needs to stay accurate up to many times nominal rating.

For your application, the current that will flow through the the CT in the event of a capacitor failure (unbalance) is generally quite small and for discussion's sake is a "steady-state" current. I.e. the neutral current will rise from almost 0A to 1A or so. In general, protection rated CTs will work fine for this application, if the rated current is very close to the current you expect to flow in an unbalance situation. However, you should not use a 200:5 A protection rated CT to try to detect 1A flowing...this will not work well.

From a design/cost perspective, it is generally more economical and better technically to by a metering rating CT with a suitable ratio. In the ANSI world, it is quite common to see a 5:5A ratio used with an accuracy rating of 0.3B0.5 or so. With the low rated current, you should look at reducing the rated burden as much as possible to save cost/improve ease of design. The ratio you use really depends on the capacitor bank parameters and the type relay you're using.

As I'm sure you know, you need to use a CT insulated to the system voltage. If you have a line to ground fault, i.e. an entire back failure, the CT in the neutral will float up to the system voltage.

Hopefully this helps.

 

[busbar]

http://standards.ieee.org/reading/ieee/std_public/description/relaying/C37.99-1990_desc.html

INDEX may have applicable information.

 

[144x]

isn't there any possibility for the measuring core to saturate if there is cascade failures of capacior cans?

 

[scottf]

144x-

There could be a chance of the measureing core saturating if there is a cascade failure, but it would still have an output current up to a certain level, which would serve the activate the relay. IKeep in mind the relays used for this application are set to trip if the neutral current exceeds 0.5A or 1A, for example.

If there is a full failure of the capacitor bank, generally that would be picked up by over-current relays on the line.

You cannot pick up this small change in current from the line side, as the relays would not be able to be set that sensitive, without constant tripping.

The main purpose of unbalance protection is to detect failures of single capacitors/sets of capacitors that can only really be detected by sensing the relative small change in the current in the neutral.