Minimum current to sustain arcing current

I am using PTW's Arc Flash software which has a default parameter "Cleared Fault Threshold" which is the percentage of initial arcing current which when interrupted results in arc being extinguished.
The default setting for the "Cleared Fault Threshold" current is 80%. ie. the assumption is that the remaining 20% of arcing current can not sustain the arc. Using Warrington's empirical formula for arc resistance, an arcing current of 1.74kA across a 'Gap' ( arc length ) of 25mm has a resistance of only 0.0208 Ohm. I am therefore reluctant to assume that the arc can not be supported after the ‘Cleared Fault Threshold” current is interrupted.
Does anyone with arc experience know what magnitude of current will sustain an arc in a system where the 'Gap' is approx. 25mm and the phase to phase voltage is 415V, 50Hz ?

I am also using PTW, and lowered the threshold down to 75% for my calculations.

PTW will do a piecewise calculation from each source and sum up the results to give you the final result. (I think.)
One problem that I had was that it would often result in very high incident energies due to synchronous motor contribution. It would give me two seconds on a single large motor (we have some VERY large motors.), using the full motor short circuit contribution for the calculation, when the upstream breaker tripped much quicker. This would result in very high energy levels. This also would not take into account the motor decrement over the time period. By lowering the threshold, I think I have simulated both effects reasonably.

It is important to realize the other aspect of what constitutes the clearing of arcing current. If you do not draw the line at 80/20 and are looking to push the envelope to 95% or more, then you are looking at clearing times well above 2 seconds.

May we could define this percentage for different systems. For example, systems with available fault current more than 60kA can have 90% "cleared fault threshold" value while systems with less available fault currents can have 80%.

WDeanN: I am not sure what you mean by "would not take into account the motor decrement over the time period." If you've used the ANSI standard to do your short circuit calcs then you are definately using the multipliers for your motor short circuit current (to obtain withstand ratings). If you've disabled these multipliers so that you could use your 75% cleared fault threshold value then you are probably not doing it right.

Regards.

The multipliers are in use.
The problem with most of the software is that it does not account for the decay curve of the motor contribution. It assumes the full motor contribution (with multiplier) for the full time period of the calculation. This results in large short circuit currents for the full time period (2 secs) with no decay accounted for. By limiting the fault contribution cut off percentage in SKM PTW, I am "simulating" the decay effect. The motor contribution is still there, and is accounted for, but effectively dies when the upstream breaker trips.

In SKM, you specify the number of cyles to include for motor contribution. This effectively accounts for decrement.

Does it stays constant for that number of cycles, or does it follow a slope towards 0?

It stays constant, so is somewhat conservative. You could adjust the number of cycles to get an equivalent I²t as a sloped characteristic. Most motor contributions do not contribute that much to the IE anyway, so I don't think it's worth it.

This was in a recent upgrade, you are right. I could probably raise it back to the default of 80% with minimal effect now. I have set the synchronous motor contribution for full short circuit contribution for 6 cycles, then the contribution drops to 100% of FLA. According to some studies I have run, and graphed, this should be conservative.

As I said, we have some LARGE motors at my facility. Several, yes Several, over 20,000 hp. They make a difference!