Noah wrote:

Thanks for the reply, this is their website I have just recently found.

http://www.arcflashtables.comThe article seems to have disappeared since conversion to the new web site and I believe that Jim is going to put out a revised version, but this is probably similar to a simplified calculation procedure.

The IEEE 1584 incident energy calculation is driven by 4 variables:

1. System voltage.

2. Bolted fault current.

3. Arcing time.

4. Grounded or ungrounded.

5. The equipment type (switchgear, MCC, panelboard, etc). There are effectively 4 "classes" of equipment even though the table has more entries.

6. Sometimes, gap although this quite often is set by equipment type.

Out of these, we can safely say that factors 1, 4, 5, and 6 are generally "constants" or that we have a small number of standard cases. That leaves items 2 and 3 to work with.

Now, the standard approach is to calculate everything and then determine the incident energy and thus the PPE required.

Working it backwards, we start with PPE which givens an incident energy. Then if we produce a table for say "all 480 V MCC's", if we use some arbitrary maximum trip times such as 5/10/20/30/60/120 cyccles, we can back-calculate the maximum fault current possible for a given arcing time and incident energy. Or conversely we can set the maximum available fault current and calculate a maximum trip time.

This results in a considerably faster method for calculating incident energy, one that can easily be done "by hand". At least at first blush this appears to be the case. And getting a short circuit value, especially a conservative one, is extremely simple to do. For instance with a rough guess at X/R ratio and transformer kVA and %Z, one can assume wiring has zero impedance and calculate a worst case short circuit current. If there are motors involved, some small adjustments to the value are eeasily accomplished. This takes less than 5 minutes to do by hand and I have personally used this method (basically the ANSI short circuit method) for years.

But, there's a huge problem with this method. Quite often, incident energy INCREASES as available fault current (arcing current) DECREASES. I know that this sounds absolutely backwards. After all, incident energy is approximately related to power so a decrease in current would result in a decrease in power. However this misses the bigger picture issue. Inverse time devices (breakers, fuses) take longer to trip as current is reduced. This effect is definitely not linear in nature. So quite often the rate that the arcing power (arcing current) is decreasing is far less than the rate at which the arcing time is increasing.

Note also that this is not an absolute rule. There are times, especially with long cable lengths, that the expected thing happens and reductions in current result in reductions in incident energy. But the majority of the time, this is not true.

So notice the trap here. The simplified approach results in much higher arc currents and thus fast trip times and thus paradoxically, lower arc flash ratings.

However that being said, quite often I will use a simplified method anyways. If I use a simplified, conservative value for arcing current and calculate the incident energy with an arcing time of 2 seconds, I can immediately predict the maximum, worst case possible arc flash. Its just that I can't reliably assume anything about short circuit current with any degree of accuracy to the point that I could use a simplified approach beyond looking at incident energy from an extremely conservative point of view.

In another direction, I can do some extremely simple things with a given arc flash calculation. Assuming voltage, equipment design, and arcing current are fixed, and if I'm only making changes to a protective device, they are, I can then assume that time is linear (and it is) and simply do ratio analysis to figure out how to adjust the protective devices to achieve a given incident energy.