All of the posting below precludes arc resistant gear which is substantially different in construction.

Perusal of the tables in 70E shows that when doing operating activities (switching such as opening/closing breakers or disconnects) with the doors closed and latched, the ratings are either PPE 0, or PPE 2. The ratings are significantly less than potentially possible (>40 cal/cm^2 is possible yet the tables yield a value of 0 or 2).

This by itself is quite interesting. In a doors closed/latched with properly maintained gear scenario, the major danger is offgassing around the door seal during an event. Second, the PPE requirements come from ASTM F1959 which is a test based on direct exposure to an 8 kA arc and the thermal energy from this type of event. Arc blast and other secondary effects are not addressed in 70E, nor in ASTM F1959.

There are essentially 3 different "scenarios" under which PPE is tested to ASTM standards. First is for firefighting. Second is potential exposure to flash fires. Third is exposure to arc flash events. Although the testing methods and techniques are all obviously similar (such as comparison to the Stoll curve), the methodologies are adjusted to the specific scenarios under consideration.

Finally, the incident energy calculation performed in IEEE 1584 is under an open/exposed arc scenario. Moving arcs as well as shielding from arc flash events such as by the door of an enclosure are not considered. So the results of the IEEE 1584 calculation at least directly would clearly not apply to doors closed/latched scenarios.

Appendix F in 70E at least hints to the idea that the tables were developed under a risk assessment based procedure which would fully explain why we arrive at PPE 0 or PPE 2 scenarios irrespective of the potential result produced from an IEEE 1584 calculation. Unfortunately, the specific reasons for the risk reductions taken in the tables (if indeed this is how the values were arrived at) is unfortunately undocumented.

Questions:
1. Does anyone know what the risk assessment assumptions made were to develop the tables?
2. Since "switching" (aka operating) activities cannot be analyzed using IEEE 1584, is the table value appropriate? Why or why not?
3. Is selection of PPE under ASTM F1959 appropriate for switching activities, or should the test procedures for flash fires be used instead?

The method used to develop the task tables is described in 130.7(C)(9) FPN No. 1.

This is a false and potentially unsafe assumption.

Unless the doors have been specifically tested for containing an arcing event there is no assurance they will remain closed. While there is some standard testing that addresses enclosure integrity during a 'normal' fault interruption (i.e. UL AIC ratings for enclosed devices), there are very specific tests for containing an arc flash event. Equipment built to contain an arc flash event is generically called Arc-Resistant and has requirements for reinforced doors and fastening hardware.

The consensus in the electrical industry (and as directed per NFPA 70E-2009) is that ASTM F1506 is the standard to look after to select PPE garments (other ASTM standards apply for face protection (F2178) or arc resistant rainwear (F1891)).
Those standards all reference ASTM F1959 for the test procedure to qualify the material as arc resistant, but the final marking on the PPE is F1506 or F2178 or F1891, not F1959.

The same garment may also meet NFPA 2112 (flash fires protection), but without proper testing you cannot assume it. The characteristics of the heat exposure are different, so the protection needed may be different.

Excuse me for being cautious and not accepting a flippant answer (read NFPA 70E, despite the fact that it doesn't actually give anything other than a hand-waving answer), or one which is clearly intended to hock arc resistant switch gear while failing to provide a real answer.

Arc resistant gear is not the answer if you operate outside of a medium voltage world because there is no accepted standard for arc resistant gear and thus no independent testing lab certification. You are entirely at the mercy of the manufacturer and their claims, accepting a lot of liability for your organization if you put any faith in this stuff.

If you do operate in the medium voltage world, then it's kind of a joke because the fact is that it is getting rare to see hand operated devices such as a pressure bolted disconnect these days except for no-load switches on overhead lines or puffer switches with substantial springs and mechanisms that place you well clear of even the 36" boundary. The reason that it's not the answer is because most end users are tending more and more to go with vaccuum gear (2.4 kV-35 kV) or SF6 (35kV+) due to the substantial cost advantage. This type of gear automatically has an electric operator of some sort and adding a remote operator is trivially easy, far less than the cost of arc resistant gear. Why place your faith in arc resistant gear when you can place yourself outside the arc flash boundary at substantially lower cost unless there is a space issue?

When I do a risk assessment, I have solid reasons for a severity or frequency rating. I use OREDA, IEEE Gold Book, HEART, actual documented plant data, or some other documentation for determining frequencies of occurrence on an event. That is how a risk assessment is done. You can pull numbers out of thin air, but it won't stand the test of time as the risk assessment gets scrutinized and rescrutinized over time.

NFPA 70E provides hand waving suggesting that this is what they did but the actual substantiation is completely absent from the document.



This is either ducking the question or being flippant.

"The work tasks and protective equipment identified in Table 130.7(C)(9) were identified by a task group and the protective clothing and equipment selected was based on the collective experience of the task group. The protective clothing and equipment is generally based on determination of estimated exposure levels.
In several cases where the risk of an arc flash incident is considered low, very low, or extremely low by the task group, the hazard/risk category number has been reduced by 1, 2, or 3 numbers, respectively. The collective experience of the task group is that in most cases closed doors do not provide enough protection to eliminate the need for PPE for instances where the state of the equipment is known to readily change (e.g., doors open or closed, rack in or rack out). The premise used by the Task Group is considered to be reasonable, based on the consensus judgment of the full NFPA 70E Technical Committee."

So in other words, they did a risk assessment. It is clear (if you run the calculations based on the data presented with the tables) that they did indeed exercise the "1, 2, or 3" number reduction procedure with respect to switching operations. It is also clear because the same tables have substantially higher numbers for what would be considered "working on" activities.

However, this is not sufficient to reconstruct how they arrived at the conclusions reached...the documentation for the risk assessment is missing, and that's what I'm asking for. I don't need the final result...it's there in black and white. I just want to know if the supporting documentation is available.



The test you refer to is ANSI C37.20.7. It is applicable only to medium voltage switch gear. Essentially the test specifies hanging "T-shirt" type material 4" from all the openings, setting off an arc, and checking all the samples to see if they suffered any damage. It's a pure pass/fail test. On metal clad and metal enclosed switchgear, all the elements are already present except adding a vent port on the top or back to vent the pressure wave away from the front face. "reinforced doors" are not in the ANSI C37.20.7 standard...all it specifies is how to test for arc resistant gear. Since the majority of the requirements (such as separating compartments) are already requirements for metal clad or metal enclosed medium voltage gear, the additional cost (about 10%) is relatively minor. No such thing exists in the low voltage world. Even the test is not intended for low voltage equipment. Anyone selling "arc resistant" low voltage gear carefully skirts the issue that none of it is certified to operate correctly, even though they simply take the medium voltage standard and apply it to low voltage equipment. All bets are off if you violate the equipment design AIC.

AIC ratings are highly relevant. To suggest they are not relevant sounds like sales talk to me since this comment is truly dangerous. If you are operating your gear outside it's design SCCR (AIC) ratings, then none of your protective devices are going to be guaranteed to open during a fault. At best you might be able to claim the IEEE 1584 rule of a maximum 2 seconds fault time (which is based on either personnel moving away or being pushed away by the arc blast). This is only considering the arc flash heat equation and does not consider the effect that flying equipment has on personnel safety as the bus bars are ripped out by magnetic forces and ejected from the housing. The whole point of the revisions to UL 508/508A that got the world in an uproar about 5 years ago or so is precisely because equipment was being built that totally violated AIC (SCCR) ratings and several incidents had occurred as a result. NEC also addressed it about the same time by adding new labelling/rating requirements.

The other reason that AIC is relevant is because this concept of using the tables for switching/operations is recommended to determine PPE required by Square D. They are firm believers that the risk reduction assumptions were determined precisely because of the UL requirements with regards to testing equipment for AIC.

I couldn't find substantiation of any sort for the NFPA Task Group's risk reduction decisions (only their methodology), but I've been given more or less an informal "hint" from Square D. Hence the reason I asked the question.

I never said AIC was irrelevant. In fact, I agree that operation within the AIC rating of equipment should be safe and considered as part of a risk analysis program. However an an arcing fault on the internal bussing of a switchboard, particularly with group mounted devices, is not part of the AIC/SCCR testing performed by agencies like UL. So, I have different opinions towards 'tasks' like routine operation of branch circuit devices versus the operation of a main device after 'work' has been conducted inside of the equipment.

NFPA-70E Task Matrix Table Risk Asssessment

The risk assessment that was used to generate the category numbers in the 2004 task matrix table 130.7.c.9 was documented in Annex J of the NFPA-70E-2003 Pre-print. They elected not to include the annex in the final 2004 issue. They made reference to it in the 2008 issue.

Thank you Jeff for the reference.
In 2003, I was a bit too young to be involved in electrical safety. It's good to have somebody remembering that time which is not that far away.
For those interested, the NFPA 70E 2003 preprint is just a Google search away, direct from the NFPA website.