Hello everyone, here are some questions I have about arc flash policies. I am thinking mainly in terms of 70E and 1584, and I'm in the industrial world (mining, power generation, etc).

1. Many sites have a rule that says "no live work". If indeed that rule is followed 100%, what need is there for an arc flash study? (Besides letter of the law.)

2. I often see that the arc flash safety rules are applied as if the doors were open, even when they are not. So the AF boundary, PPE requirements, etc are applied whenever the equipment is live regardless of the presence of exposed conductors. What do you say about this policy? The inconsistency there is this: if we are acting like switchgear, MCCs, panelboards, etc are dangerous with the doors closed, why do we not put labels on motor terminal boxes, cable tray, etc? (I'm not arguing that they are not dangerous with the doors closed, just pointing out the inconsistency.)

3. I've been to sites where you have to wear the PPE, including maybe a bomb suit, any time you enter the AF boundary when the equipment is live. I've been to sites where you don't need any special gear in the AF boundary unless you are operating a breaker, and then you wear FR clothes and a face mask if you are operating a breaker. And I've been to other sites that require no extra PPE unless a door is open to live conductors. I think all of these are in compliance to 70E, but it's crazy how differently they interpret it. Thoughts?

4. Is there a way to determine the hazard of an arc flash that happens when a door is closed? With the variety of equipment and manufacturer terms like arc resistant, arc proof, ducted, top venting, etc, how can we determine the hazard for someone standing in front of elec equipment when the door is closed and latched? I imagine we could have a statement by the manufacturer that says something like "when the door is shut and secured, the IE levels are reduced by 80%" for example. Instead of applying the hazard conditions of an open door scenario to a closed door scenario.

5. Another nuance to this is when you have a door open to a control voltage compartment that is "isolated" from the higher voltage, but not as protected as if the door was closed and you were just standing in front of the gear. I'm thinking specifically of Eaton medium voltage MCCs. The lower door opens to the MV breaker, and the upper door opens to the control wiring and the back of the motor protection relay. It's often necessary (especially during commissioning) to be working the control voltage compartment while the MV below is energized. Another similar situation is when you get a 480V MCC with an integrated PLC or networking cabinet. There is usually the main horizontal bus behind the control cabinet back wall. Should the electrician be dressed in full PPE according to the AF label whenever they are in the controls?

6. Do you put labels on every MCC starter bucket, one on each vertical section, or one for the entire lineup (or 2 if you have different IE levels for the MCB vs feeders and starters)?

7. Do you calculate and make labels for PLC cabinets, network racks, etc with 120V? (Every site I've gone to has excluded these from the study-- seems inconsistent to me.)

8. Is there an official interpretation of IEEE 1584's comment about applying the line side IE levels of a main breaker to the entire lineup if it is not adequately isolated-- aka escalating faults?

I have talked with a few different engineering companies who do AF studies, and I've gotten a variety of answers about how to interpret it. Some say insulated bus and sheet metal compartmentalization of any type is adequate isolation, and others say- no, there needs to be a manufacturer's statement that it is adequately isolated, which means they mitigate it by getting MLO gear and use a standalone breaker to feed it, or buy something like Schneider's ArcBlok. What do you guys think? What is best practice?

You can order a switchgear with arc chutes, 1/4" steel doors with 40 bolts around the perimeter, and the manufacturer still won't take the liability of saying that it will prevent faults from escalating to a line-side fault. Yet many engineers doing AF studies are willing to decide that a regular, 40 year old MCC with a sheet metal wall separating the main CB from the horizontal bus is enough.

And please let me know if I have any false assumptions. I consider myself intermediate in this-- I know a lot and have some experience, but I am by no means an expert.

I abbreviated some terms--
MV - medium voltage (usually 4160V for my comment)
MLO - main lug only (incoming cables land directly on the bus, no integral main breaker)
CB - circuit breaker
MCC - motor control center

Thanks everyone!

A few thoughts - others may add to this. Also, this is not any formal interpretation or engineering opinion, just a few things to consider (legal disclaimer)


The process of creating an electrically safe work condition per NFPA 70E requires absence of voltage testing which is considered live work. I have a few cases where a system was believed to be de-energized but wasn't. One resulted in a huge arc flash and debilitating injury when they "thought" it was de-energized.


Two things here. Doors open statement applies when an arc flash hazard exists. Examples: Energized work being conducted, interaction that could result in an arc flash etc. IEEE 1584 calculations are based on doors open because unless the equipment uses an arc resistant design, there is no way to know if the doors will remain closed during an arc flash - blast pressure.

The second part is if there isn't an arc flash hazard - no interaction, then the arc flash boundary doesn't exist. i.e. simply walking down a corridor with equipment nearby doesn't trigger an arc flash boundary.


Some of this depends on the severity of a prospective arc flash. PPE is required for any part of the body within the arc flash boundary. That is the area where the incident energy exceeds 1.2 cal/cm^2 which is the threshold for arc rated PPE. It also depends on the likelihood of occurrence of an arc flash - NFPA 70E provides a table for this. Equipment condition, condition of maintenance play a role in this decision.


Arc resistant equipment is designed to offer protection to the worker when the equipment is properly used, doors closed etc. I saw 15 kV arc rated switchgear being test several years ago. I was in another area of the test facility conducting my own testing and didn't know they were about to test. When it happened it sounded like the end of the world but.... It was a successful test.


This would be part of a risk assessment. What is the likelihood of an arc flash occurring if no one is directly interacting with the power side of things.


There isn't a standard for this but most will place at least one label and maybe two depending on the physical size of the equipment and whether they are using the main device to define the arc duration for the feeders. There is nothing to prohibit a label on each cubicle but the front of the equipment will be very crowded with labels.


Technically NFPA 70E applies to 50V and greater but what you cite is common.


This requires care and judgement. For switchgear, many have one calculation for the main based on the clearing time of an upstream device - often a utility device that may be quite slow. The the arc duration of the feeders are based on the main.

I agree with Jim's responses.



A couple of points:

To clarify, arc flash PPE is required for work within the AF boundary. We can debate as to the definition or "work" but as Jim mentions, 70E table 130.5(C) provides guidance an assessing the likelihood of occurrence of an AF event. In general operating electrical equipment in good repair, with covers in place, and in a manner consistent with published instructions (e.g. operating a breaker in a panel or switchboard), may not be considered to have a likelihood of occurrence.

There is a separate set of rules and PPE requirements for shock hazard, which apply for work within the restricted approach boundary of energized equipment.

One must be specific as to the type of hazard being addressed and review both sets of requirements.

Thank you so much for the very insightful responses. I will take a few days to process this.

On the philisophical side of this discussion, it is useful to remember the hazard reduction hierarchy of controls:

From most effective to least effective with examples
1. Eliminate the Hazard (no live work)
2. Substitution (remote work)
3. Engineering Controls (arc flash detect and trip systems)
4. Administrative Controls (procedures)
5. PPE (bomb suit)

However, each of these is also fraught with the possibility of human error, especially any control where procedures are involved. There are many incidents where a procedure to eliminate the hazard (i.e. de-energizing the equipment and testing that state) was not followed correctly, sometimes with catastrophic consequences.

It's easy to become complacent with following the rules and guidance of standards and there are numerous techniques to overcome this such as checklists and written job hazard evaluations. Well meaning but misguided or too narrow focused application of "safety" rules can result in greater hazards. Managers with little or no training in electrical safety can make arbitrary workplace rules that cause problems. As an example, cut resistant gloves were mandated for all workers in one circumstance I was involved in. However, the electrical conductivity of those gloves was not considered, resulting in a serious arc flash incident when an electrician attempted a commonly used procedure to measure live current with a clamp-on meter.

Equipment designed and installed before the current generation of arc-flash and electrical safety standards were developed is frequently next to impossible to diagnose and maintain without considerable forethought as to how to complete the work with minimal exposure to hazards. Those of us who have been in the trade for many years constantly have to carefully rethink and retrain ourselves to eliminate old bad habits such as flipping open buckets without proper analysis and protection. This can also become a serious impediment to testing and maintaining equipment properly when it is "just too much hassle" to follow proper safety procedures when servicing old equipment. For example, we had an incident where a piece of equipment (critical live spare) was left out of service for months because no one could be bothered to put on arc-flash gear to measure whether the contactor was making good contact on all three phases and providing full power to a VFD input. Yet again, the manufacturer neglected to provide a diagnostic fault indication for the loss of an input phase, the assumption being that electricians could easily test for this condition with a meter as recommended in their manuals.

There is also a great temptation that much be resisted to bypass procedure when no one is looking which some companies try to reduce by monitoring security cameras in electrical rooms. However, this is only effective if it is continuously monitored and policed. Typically the video is only reviewed when there is an incident.

Each company must take responsibility for providing safe work environments for their workers. It is not enough to rely on standards and procedures without also providing practical solutions, procedures, training and enforcement to complete each of the required work tasks safely. By all means, interpret and apply the standard, but also consider the possibility of human error at each and every step.