We recently found a situation where we had to look at the IE of a disconnect that was located several hundred feet away. The disconnect then got moved several feet. The difference in IE between the two locations was ~15cal and .5cal. The reason was the current was about the same as the short-time setting, one to the left and the other to right of the setting.

The point of this is we saw a huge variation in IE based on estimated data (after all isn't everything we do estimated to some degree). Such a small change resulted in huge differences in IE. One being "safer" and other being "more hazardous".

So my question is: why doesn't 1584 work on a range of values - say +/- 10% of the arcing or bolted currents and evaluate the worst case across the range instead of a select number of discrete points? It seems to me finding the worst-case on a range would prove far more valuable to the person standing in front of the equipment with a poky metal object.

Looking at the tables in 1584-2018 the coefficients are at 10^-5 or -6 when my utility contribution is +/- 20% and my cable length is equally as varied.

Just looking to find a reasonable answer to a complex problem that doesn't cost fortune but saves lives!

It's not IEEE 1584. And it's really not very accurate beyond about 1 to 2 digits. In your case the dramatic difference is because your circuit breaker calculation jumped significantly in opening time.

To elaborate a bit, when the arc flash energy is low, that is because the breaker is tripping instantaneously, the tripping time is very low. When it jumps up, that is because the breaker is tripping out in the long time delay (overload) region. This is because the increased cable length decreases the fault current which causes the breaker to trip out in the overload region.

Correct me if I'm wrong but doesn't IEEE Std 1584 cater for this by taking the higher of the two incident energy values calculated (first from the arcing current [Iarc] and then a reduced arcing current [Iarc_min])?

Section 4.5 of 1584-2018 states the following...


See the attached image which illustrates a big difference in arc durations for arcing current and reduced arcing current.

In our studies this often happens, we always recalculate the incident energy with the Iarc_min, we prefer to do it in that way, we think is a safer practice for the user