The IEEE 1584 Equations can produce some very large incident energy levels - usually dependant on the protective device clearing time. That is one of the reasons the "2 second" rule/cutoff was created.

What is the largest calculated incident energy you have seen in any arc flash study/report. (with or without the 2 second rule - simply what is the largest number you have seen.)

I think this question should be tempered by maybe excluding the main incoming protective device which maybe on the secondary side of the incoming transformer. Usually the primary side is under control of the utility and therefore, I typically exclude that as the client has no control over what the utility does.

That being said, if I allow the default to be 1000 seconds, I have seen incident energies going to over 11k as there is no protective device on the primary side of the incoming transformer.

Additionally, I have had one utility give me a range of 7.5 to 100 MVA for their fault contribution and then 2 years later change it to 4.5 to 100 MVA. Needless to say, the AFH did change on some buses to worst than before and negated some of my recommendations to reduce the AFH.

Just my 2 cents worth.

There were some sites that had AF around 50 cal on the LOAD side of the Main breaker... After doing some revisiting, and turning the breaker trip settings down (which could be done without effecting the coordination) it dropped the values to around 5-12....

And a good 2 cents it is. I left this question open ended because there are so many variables just like you described. Rather just complicating the question, I left it simply as, how big is the number and people can explain if they wish. - Thanks for your info.

btw, my largest incident energy was 205 cal/cm^2. Transformer secondary, no Inst. Pickup - all the usual reasons.

Seen some large ones, for the same reasons mentioned already. But my favorite was at several large industrial plants (A name you would all know) had a bunch of stuff labeled HRC 5.

Everyone is concerned with big numbers and rightly so but if you look at a fault (or heat source) long enough, no matter how small, you can get some pretty big energy number using 1584. As an extreme example, you can get over 40cal/cm2 if you stand in the sun long enough (1.96 calories per minute per square centimeter); I can get 40cal of accumulated heat in front of my fireplace. Neither of these are seen as dangerous events assuming your wearing sunscreen and not playing in the fire. The point is everyone talks in terms of accumulated heat numbers, maybe we should be looking at how fast the heat is produced (cal/cm2 per unit time) for a specific event to better understand all the dynamics associated with the AF event.

We’ve been taking a little different approach and further analyzing these “bad actors� to see if they’re a function of very high current over a very short time (big boom, very bad) or a result of much lower fault current over a very long time (burning fault). In my opinion, one short coming of the 1584 methodology is that the results and associated equipment labels do not give the user any distinction between these very different events. For example, you may see 30cal/cm2 for two locations. One is the result of say 50kA in a few cycles while the other was only a few kA but over seconds…from the observer I believe these would be two very different events. Extra caution and possible mitigation may need to be considered for the first location as opposed to the second.

Great! You nailed it. The 40 cal cut off seemed like a good idea at the time however, as you pointed out, the equations are based on total accumulated energy. It's a carry over from earlier work. Just because you exceed 40 cal doesn't necessarily mean an explosion.

I received one of my worst sunburns ever, many years ago snow skiing - red, puffy and painful. It took many hours of "energy exposure" (and the stupidity of not wearing sun screen) for this occur. The same burn can occur in a matter of cycles given a greater energy per time. The existing IEEE equations do not account for this but Heat Flux will be the next "big thing" i.e. energy per area per time.

IMHO you are looking at this in a more practical manner.

That’s a scary thought…where’s everyone else at? We cannot be the only ones who see it this way.



My point exactly but how do you get everyone else to understand this?



I going to assume “heat flux� will not be addressed as part of the 1584 revision, or will it? If not, then what?

Heat flux is slowly creeping its way into the picture. As far as IEEE, the new equations seem to still be a long way off. The next phase is taking a long time.

I do address this in the book coming up in an effort to force the issue.

Hopefully as more people begin to talk about the issue, they will begin to recognize there is a difference between a 40 cal boom and a 40 cal long sizzle.

I believe a great number of people know this already, the problem is where do you draw the line between an explosion and a slow sizzle. I guess all I can say at this point is:

....to be continued.

Given the question of how large of a number, I have a result of 11,760 cal/cm2; however, this is not at a point where anyone would actually ever work energized. The situation is a fault on the primary side of a 12,470/480V transformer of a double ended substation with the 480V tie breaker closed. Actual clearing time calculates to 311 seconds. The sequence of events is that upon experiencing the fault the 12,470V feeder breaker to the transformer with the faulted primary opens in a few cycles. Then the opposite side of the double ended substation feeds the fault. Because the fault current is feeding through the impedance of two transformers, i.e. stepping down on the west bus and stepping up on the east bus transformer (or vice versa) to feed the faulted primary, the arcing fault current is reduced to a point where it misses the instantaneous and short time of every protective device. Using the two second rule the energy calculates to 75.7cal/cm2. The situation can definitely be improve through setting changes; however, it is on the low side of my priority list. I know that sounds strange, but we never actually close the tie; and if we ever do, it will be because one of the two transformers is out of service due to a transformer or feeder failure.

So if the Arc Flash Protection Boundary was ever calculated, it would be...

...in the next State?

Your results were one of the points of my questions. How (unrealistically??) large of a number can be attained with the equations.

So far you are in first place for the largest Ei.

Seeing as how the new equations are a long way off do you think at least more guidance will be provided as part of the 1584 revision? As an example, maybe a shorter cut off point, something less than 2 seconds, for low voltage or light weight equipment.

We’ve seen several examples and associated papers of independent testing where either the voltage is too low to allow an arc to continue after the trigger wire burns out (typically at 208V) or the equipment just does not have enough fuel (copper) to feed the fire for any duration beyond a few cycles.

Largest incident enrgy

The question also points to my greatest frustration with the Table Method - No task is ever Dangerous even though as many people have pointed out the secondary of large transformer commonly calculates above 40 cal/cm2.

Naw - I'm from Texas. We're big enough that the next county will be fine!

That's amazing. I almost added ...unless you live in Texas or Alaska. Who would have guessed.

All tasks at a location where the table footnotes are not met could be considered Dangerous because out of spec of the table.

I have two different studies from the same facility (two different services) that were extremely high one at 58,468 cal/cm2 @ 18� with an AFB of 27,398� = 2,283’ and the other at 15,648 cal/cm2 @ 18� with an AFB of 11,196� = 933’. These are real scenarios and in areas where some would (and actually had – prior to the study) worked live. This is 480V Switchgear (without a Main) fed by 2500KVA Utility Transformers.

Threading the Needle Reference

See my posting "Threading the Needle" in the Software for Arc Flash Studies section. I would have posted it here but attachments are not permitted in this section.