Effect of Capacitor Switching Transients on Customers

[p5t1]

All we know that most of the capacitor switching transients ( 1.0 to 2.0 pu peak) last between a few miliseconds to about 10 or 20 miliseconds, time for which the transients are almost damped completly. Then, according to the ITI (CBEMA) Curve, most of these transients are outside the "No Interruption in Function Region" (i.e. outside the safe region for a customer.) Then, the question is: Why is it very rare to hear about a damaged customer equipment ?

p5t1

 

[alehman]

The impedance of the system between the capacitors and the utilization equipment reduces the magnitude of the transients seen by the customer. Or maybe the CBEMA curve is too conservative??

 

[p5t1]

I agree with you respecting resistance will damp de transient. But how do you apply the ITI Curve to an oscillogram of the transient (either a measured or simulated transient).

 

[mc5w]

There have been recorded instances of capacitor switching transients that would produce false triggering of silico controlled rectifiers. If the rate of change of voltage across and SCR is high enough, it will produce a capacitive current flow in the junctions that acts just like the current that is injected through the gate.

The way that an SCR is implemented on an integrated ciruit is to connect a PNP and an NPN transistor collector to base collector to base. Not only does it have the gain similar to a Darlington transistor but it is also a regenerative device which means that its gain after a few microseconds is at least a 100 time that of a Darlington.

Usually, these instances were confined to a single plant where power factor correction capacitors would cause false triggering of SCRs in a motor drive or furnace control.

Mike Cole, mc5w@earthlink.net