PROTECTION COORDINATION

[reactive]

Which fault current is to be used for coordination of IDMT overcurrent relays - 1/2 cycle symmetrical or interrupting current (4-5 cycles, 50Hz). I am using the IEC 60909 method of determining scct currents (EDSA software).

I have been through many standards (IEEE color books) and cannot find a definitive reference to this.

In depth discussion would be welcomed.

 

[DanDel]

If the overcurrent relays are tripping C/Bs, then the calculated interrupting fault current for that C/B should be used for the end point of the relay curve on the TCC graph.
The interrupting value, while it is based on the symmetrical fault current, usually takes into account the manufacturer's test procedure for the interrupting rating of the C/B, among other factors. This results in a value that is usually greater than the symmetrical fault current. The important thing about relay curve coordination is the time difference at the high current end of the curve, usually required to be a minimum of 0.3-0.4s for electromechanical relays(per IEEE 242). You might be able to cut this time in half for electronic relays, but this is a good rule-of-thumb.

 

[jbartos]

Suggestion: Please, notice that medium and high voltage breaker protect and operate somewhat differently then low voltage breakers. 1/2 cycle tends to be linked to low voltage breakers and 4-5cycles tends to be linked to medium or higher voltage breakers.

 

[Bung]

Depends on the protection. For relay op times around a cycle, you need the 1-cycle values, for slower relays (standard inverse on long time levers) the current will be lower. But maybe you should take into account the integrating effect in the IDMT relay as the fault current decreases. More to think about - effects of CT saturation on the current the relay actually sees; the relay algorithm (filtered for fundamental only, rms value of total including harmonics or what)and things like dc offset. I tend to prefer to use the "standard" recipe for doing the calcs as a starting point, but temper it with the realities of the situation. (eg a 50/5 CT on a system with a 15kA fault level ain't gonna tell your high burden CO relay very much at all, so the number you use for the fault level is a bit of a non-issue!)

Bung Bung
Life is non-linear...