I heard it suggested that a multiplier could be used to add more conservatism to the calculated incident energy from an arc flash study. i.e. add 5 or 10% to the calculated number.

Here is this week's question:

Would you use a multiplier for the calculated incident energy?
Yes
No
It Depends

I voted for It Depends.

If ArcPro is being used there is a multiplier required for some situations. Also if IEEE 1584 is used for incident energy in a manhole, I will increase the incident energy as this is an untested configuration on what happens when IE is reflected back?

Unless some kind of exception circumstance exists (the above comment about manholes, for instance), I wouldn't use any kind of multiplier as I see the equations being conservative already.

Not equipped to perform my own incident energy research, the liability of departing from given formulae could go badly either way.

I guess I should have been more specific. I meant the final incident energy whether with arc pro plus its multipliers or with IEEE 1584. Good comment Barry.

Agree with Barry...it's already very conservative. The recent threads also got me to thinking about other issues in general...

First, we calculate the incident energy at a single point which is best described as roughly under the chin at a midpoint between the center of the chest and the center of the head. This is promulgated on the "Rule of 9's" which suggests that major (>2nd degree) burns in this area are more likely to result in a fatality. The rest of the body will be either closer (especially hands/arms) or farther (legs, rest of body) from that point.

Second, assuming that one is escaping either intentionally or "propelled" away from the arc over a period of two seconds until they exit presumably the arc flash boundary, the incident energy is rapidly decreasing at a rate of roughly distance squared.

Yet in both cases, we measure ATPV as the Stoll curve 50% passing criteria on a cloth sample suspended perpendicular to a model arc at various distances, and we require uniform PPE equal to the aforementioned point in space assuming the person never moves.

In my mind with respect to the "uniform PPE" criteria, this makes it at best a "survival" standard. No attempt is made to make any guarantees about the extent of injuries outside of the face/chest area. Reminds me of a Metallica song about a person that had lost all 4 limbs and was blind and deaf. And with respect to the "2 second rule", shouldn't we be accounting for the increase in distance vs. time if the "2 second cutoff" is going to have any validity at all? It seems like the "2 second rule" paradoxically compared to the intent of IEEE 1584 is specifically modelling the scenario of working from a bucket where escape is not possible or we should linearly decrease it at some reasonable value (2 meters/second, as in the "escape" criteria in moving machinery safety standards).

Second, getting to the heart of it, then what would be the basis of the "conservative" adjustment? I look at it this way. According to the IEEE 1584 standard itself, there is a 95% success rate by wearing PPE with an ATPV equal to the incident energy calculated by IEEEE 1584. The outliers are due to the "long tail" distribution of arcing currents, and due to the "sigmoid" ATPV estimate. With regards to the former any small multiplier (10-15%) is going to have little if any material effect because the nature of a long tail distribution is that only large (say double or triple) factors provide any significant improvement. With regards to the latter, looking at actual Kinetrics reports on PPE shows that 100% of the samples passed at a point roughly 0.5-1 cal/cm^2 lower than the published ATPV regardless of the manufacturer or ATPV (same result at 40 cal/cm^2 as at 8-10 cal/cm^2), which would argue at worst for an additive increase. Finally we have the practical evidence that despite going on over a decade of "field testing", I know of no case where PPE selected according to an incident energy calculated using IEEE 1584 failed to provide adequate protection. Thus it seems that right now, the methodology seems to be overly conservative given that it is more likely that the end users will wear the PPE incorrectly than being injured due to being the first statistic to "test" the 95% numerical value given in IEEE 1584.

In the case of a vault, that's pure guessing. I would never recommend doing any sort of guessing when it comes to a safety engineering study. I believe Kinetrics already has a "vault" mockup and it would not be terribly expensive to make one for testing, and guessing should not be an alternative to actual testing. Again, 10-15% seems hardly meaningful for a vault scenario where the big issue aside from the "2 second rule" is when radiative energy bounces off the walls in the relatively enclosed space of a vault.

So in summary...I'm not an advocate of adding multipliers for "conservative" purposes. There is no basis for it.

I voted for No. This is because if we look at the IEEE 1584 final incident energy equation, there is already a multiplication factor embedded in the equation as Cf "Calculation Factor". For system voltages above 1 KV, the calculation factor is 1.0. This is because for system voltages above 1 KV, the confidence factor is 95%. Confidence factor is the comparison between actual incident energy measured during test and theoretical mathematical equations solved for incident energy. That means for voltage above 1 KV, 95% of the tests conducted have calculated incident energy equal to or greater than measured by practical test instrument. Thus we have calculation factor of 1.0 for system voltages above 1 KV.
For system voltages less than 1 KV (inclusive), the calculated incident energy was less than practical test result and therefore a calculation factor of 1.5 is used to get the confidence factor of 95%. That is the reason we have 1.5 Calculation factor for system voltage 1 kV and below.
So our great knowledgeable committee and working members of IEEE 1584 have already taken care of multiplication factor for us. What I look in the report for incident energy is the clearing time of the protective devices. For example, if there is an incident energy of 20cal/cm2 over 2 seconds (120 cycles), that may appear like a slow cook, but if there is the same incident energy of 20 cal/cm2 in 16.67 ms(1 cycle), that is an arc blast which can have pressure of tens of thousands of pounds per square feet, shrapnel from such blast can fly at more than 700 mph & sound waves can exceed 160 dB. And we must remember that we don't have equations to deal with arc blast pressure.

With you so far on the "correction factor" to some degree but this part is not true. Let's break it down into a claimed pressure, speed, and sound argument. In terms of sound that sounds pretty loud but I can't verify it one way or another but the other things are very far fetched.

In say a 36"x36"x36" enclosure, we'd have 27 cubic feet of air. One cubic foot of air weighs about 0.08 lbs., so we have a mere 0.08*27 or about 2 lbs. of air in a fairly large medium voltage enclosure, and not all of that air is going to be pushing a projectile. It's honestly just not enough mass to really do much of anything, much less push it to a speed higher than the speed of sound especially when the pressure dissipates very quickly.

It is much more plausible and more likely that shrapnel could be propeled by magnetic forces. Plasma from arcs has been measured around 300-400 feet per second, which is around 200-300 MPH. This is pretty fast but nowhere near "over 700 MPH". Also, the material that is being magnetically accelerated is plasma which has virtually no mass. So I highly doubt that solid material can exceed the speed of the plasma being flung around near the arc core.

Finally there's the pressure argument of "10's of thousands of pounds per square foot", or roughly some multiple of 70 PSI (10,000/144 = 70). I highly doubt this for one simple reason. Military studies on the effect of concussive explosions on the human body show that starting around 20-35 PSI (depending on what study you read) causes a fatality by rupturing internal organs. Thus we would expect that if this were the case, we would find the number of fatalities due to arc blast to be extremely high, vastly exceeding the 10:1 ratio of serious injuries to fatalities that we see today, and the extent of both the damage to the victims and the surrounding area would be much worse than it is. Effectively an arc blast would not be survivable. But I can't point to a single arc flash fatality case that was not due to thermal burns.

Furthermore I doubt that even more than a few PSI (perhaps a few hundred pounds per square foot) is even possible for some simple reasons. Generally most industrial equipment flies apart with even small 1-3 PSI pressure rises. I've seen it several times from working around thermal processing equipment when occasionally due to poor maintenance and operating practices, it goes "poof" and things fly apart. The pressure per square inch turns rapidly into thousands of pounds of pressure on a given piece of equipment overall, but not necessarily per square foot. Measurements from pressure sensors top out at a few PSI at most. And again, no cases of people dying from rupturing internal organs. Arc blast is no different. Recent modelling efforts by CIGRE for arc resistant gear show that arc blast is all over with within 1-2 cycles and it is a linear pressure rise followed by a rapid pressure decrease once the equipment ruptures either along controlled pathways (as in arc resistant gear) or in an uncontrolled manner. Once the pressurized air is released, the pressure rise is done. And the pressures reported by their testing are on the order of a few PSI, not hundreds.

In other words, the purported danger from arc blast is totally overblown because quite simply, Lee's theoretical equation for arc blast is totally wrong. The major dangers from arc blast are loss of hearing (ear drums blow out at around 1 PSI), and being knocked off your feet/into another object/out of a manlift/bucket truck.

This is why I explain to most people that even though gas/oil/coal "explosions" (aka rapid combustion) can be very scary and you can have flames shooting over your head (and I have), and it can blow brick out (very weak in the transverse direction), blow large sheet metal seams apart, knock siding off a building, and rapidly converting square ductwork into round ductwork, the actual amount of pressure (PSI) required is really not that large and that most industrial equipment is so "open" relatively speaking that there is a far greater hazard from a blow out while putting air in a truck tire than there is from large industrial burners. That's not to say that uncontrolled fire or any of the other effects I've just described is a good thing because it's not. Its just to say that its not as bad as what it sounds like, and when the electrical equipment can easily fit inside the thermal equipment with room to spare, it's not as big of a threat as it sounds like.

I agree with PaulEngr, barry ..no multiplier, already conservative.