This is one thing that the Technical Committee did that really irks me. As of today, I think I have even more evidence that this editting of the tables should NOT be done.

Following virtually all existing risk assessment methodologies, what we have to consider when looking at risks is the severity of the risk, and the probability. If the severity is not very great, then we may find that even if an incident occurs fairly frequently, the risk is acceptable. For instance even though I might get a paper cut once in a while, it is an acceptable risk so that I don't bother wearing gloves while doing paperwork in the office.

On the other hand, as the severity goes up, correspondingly, the only way that it becomes acceptable is if the probability of occurrence is very low. As a general rule of thumb for instance most industries consider "1 in a million" to be remote enough to tolerate even a fatality.

So then we come to the tables in 70E. In some cases, the probability of risk is so low that there is no reason to wear PPE, such as opening and closing disconnects or circuit breakers in equipment that is properly designed, installed, and maintained. Hence the reason that it is a "zero" in the tables.

On the other hand there are some activities which are inherently risky such as inserting buckets into energized MCC's or working on energized 480 VAC conductors in those same buckets, especially if the reason for doing it is to make a repair. The probability of an arc flash is pretty high.

Then we get to the "in between" situations. Opening bolted covers for instance. There is a chance that a pinched wire is present or something laying against the panel, to say nothing of dropping a fastener down inside. But the probability of an arc flash hazard is somewhat lower than the above case.

The Technical Committee addressed this by publishing reduced PPE requirements. BUT, this doesn't make any logical sense. The severity is the same. Only the probability of occurrence has reduced.

So it got me to thinking. Right now we have a way to assess the incident energy required for a single consequence: a second degree burn. What we don't have is a way to look at other consequences such as first aid or fatalities. If we are going to consider reduced PPE then I'd like to see not only the risk of a second degree burn decrease, but also and more importantly, the risk of a fatality.

I got to thinking about how to extend the methodology out to these cases. When assessing burn victims, triage procedures use the "method of 9's" to calculate the percentage of body surface area that is burned. At 10% second and third degree burns, the victim is sent to a burn unit. At around 30-40%, there is a significantly probability that left untreated, the victim could die. The success rate with treatment is not as well known. A percentage is used simply to take into account different sizes and shapes for people. The average adult human body has 1.73 m^2 of surface area.

The area of most concern is the head/chest area, where even small areas can become fatal. As long as only a small area is of concern, we can look at a circle. Inside the circle, the burn will be second degree or greater. At the edge of the circle is the threshold, or 1.2 cal/cm^2. Aha...we can calculate that.

Taking 10% of an average human, or 0.173 m^2, we can calculate the radius of a circle with this area as 0.235 m, or 235 mm. Now if we know the normal working distance (which we do from IEEE 1584), then we can calculate the incident energy at the center of the circle, the place that we normally consider as the incident energy. First, we need the new working distance which we get from Pythagorean theorem. For 600 V class MCC, it's 455 mm. Recalculating for the new working distance, we get 512 mm. Using the de-normalizing equation from IEEE 1584 for working distances we can arrive at a calculated incident energy. If the edge of the circle is 1.2 cal/cm^2, then the center of the circle is 1.46 cal/cm^2...not much different.

This doesn't do much good. We can work this backwards too to estimate how much total body surface area would be impacted if we decreased the PPE from 4 cal/cm^2 down to 1.2 cal/cm^2. The result turns out to be 125% of total body surface area. It's a result, even if it's nonsensical.

In conclusion then, even a reduction in PPE by a single category does not result in an acceptable level of risk if the probability of an arc flash occurring is reduced, unless the probability is so low that it reaches the point where any injury is a tolerable risk such as 1 in a million.

This suggests that the tables should be modified to show the required PPE for the actual calculated incident energy if there is anything other than a remote level of risk for arc flash, and zero otherwise.

I certainly invite others to comment on the logic of this. I've long disdained the rationality behind the tables in 70E, and this approach seems to make much more rational sense in my mind.

Nobody responded to this. I'm not an engineer or an electrician, just a writer, but I think I see what you're saying. Are you saying that the table should just give recommendations for PPE based on the incident energy? And that it shouldn't get into judgment calls about altering the PPE based on the probability of the incident? Is that it?

No, not at all.

There are standards (IEC 61508, IEC 61511, ANSI TR1, the RIA standard) which are general standards for performing risk assessments. This is also true in NFPA 70E which keeps trying to put an assessment procedure in the appendix but nobody has put in a clear one yet.

In all of these methods, you evaluate the consequence of something bad happening. That is what the arc flash analysis in IEEE 1584 does. It tells us that if the worst happens, how bad it can be. IEEE 1584 only evaluates the threshold for a second degree burn. I pushed it to third degree/fatalities and basically showed that there is little difference (it's an exponential curve).

The second step in a risk assessment is to look at the chance that it will actually happen. You look at each task and evaluate what the chances are that something actually does go wrong. In risk assessment methods, they look at how often the task is done, the chance of avoiding injury (if possible...for electrical, this is not possible), and so forth.

For any tasks where the consequence and the likelihood of it happening are deemed unacceptable (usually shown in a matrix), the task must be re-evaluated to find ways to reduce the risk to an acceptable level. In the case of arc flash hazards, this could be modifying the equipment to reduce the potential incident energy, or modifying the task or equipment to move the worker out of harms way. The final solution which is the least desirable is to use PPE to reduce the potential consequence. PPE is always considered to be the least desirable outcome in all risk assessment methods.

In the case of 70E, most folks are going about it making the assumption that the probability is always unacceptable. Very few are even attempting to look at ways to reduce the hazard or the likelihood. And they are jumping straight into wearing PPE as the only viable solution unless there is no PPE available.

In the case of the tables, the 70E Technical Committee did step 2. They evaluated the tasks and the likelihood of an injury. In SOME cases especially in the 2012 edition, they determined that the likelihood was so low that they mostly eliminated the PPE requirement (PPE level 0). In other cases especially in previous editions, they just sort of magically reduced the PPE requirement without any practical justification at all, no showing of a reduced consequence.

i keep kicking around the idea of presenting this as a full blown document somewhere. I just haven't had the time to do it. In my mind, I would simply adopt one of the existing risk methodologies (in the end, they are all nearly identical) and apply it to this particular problem as a proposal to rid ourselves of the botched risk assessment method in the appendix.

The tables would stay, AND they should show "PPE 0", just as they do in the 2012 edition whenever the likelihood vs. consequence is deemed acceptable. In all other cases, the PPE rating should match the incident energy level. I'm not 100% sure but I believe that this is exactly how the 2012 tables are written though.