Situation: Switchboard where the incident energy calculated on line side of main was 66 cal/cm2. Barriers preventing contact between each section were specified and confirmed installed.

1584.1 Section 8 Arc-flash hazard calculation location considerations EXCERPT -
"Using the same logic, main circuit breakers in switchboards and panelboards generally are not considered to provide arc-flash protection for their switchboard or panelboard because there is normally no complete barrier between their line side connections and the possible location of an arc-flash in their enclosures."

My questions :
1. Does the addition of a barrier present an opportunity to include the main breaker when calculating incident energy in all the other sections except for the main section?
2. Why or why not?




Some of my thoughts on this.
1. The barrier will prevent accidental contact of line side main breaker when working in adjacent cubicles.
2. My understanding of the intent behind excluding the local main breaker in arc flash calculations within switchboard or panelboard is that accidental contact could happen on the line side of the main breaker. The highest potential incident energy exposure within a piece of equipment should be considered and when line side of main is exposed, we exclude that main.
3. What is the chance of exposure from an incident happening on the line side of the main breaker while in, or operating a breaker on an adjacent distribution cubicle assuming the Main Breaker cubicle is closed and isolated by a barrier that would prevent accidental contact on line side of main (in other words no live exposed parts with the higher incident energy level)? In other words we are eliminating the chance of accidental contact line side of main in an adjacent cubicle by adding a barrier. Section 8 - IEEE 1584.1 2013 seems to support this contention.
4. That said, if someone has the main section open, and the barrier is not designed to prevent propagation of an arc then I would say that the entire lineup would be of the higher Incident energy just for the duration the main section is open. Written procedures or other procedures (warning signs, kirk key etc…) may be in place to prevent main and distribution sections to be open at the same time.

In sum, I would tag the Main Cubicle only at the higher level, Tag distribution sections taking into account the internal main within switchboard and finally ensure that there is either warning labels and/or written procedures to prevent main and distribution sections being open at same time.

Direction is not right. It's not a matter of controllling whether or not an arc is triggered. That is what an arc flash risk analysis is all about. IEEE 1584 is only concerned with the hazard part of the equation...what happens IF an arc flash occurs. The issue here is one of propagation. As the air temperature increases from an arc, the dielectric insulation of the air breaks down and it becomes more conductive. Generally speaking single phase arcing faults propagate to 3 phase faults within 1 cycle (16 ms) as an example. And similarly arcing inside a panelboard will often move to the line side of the breaker within an enclosure and that's what IEEE 1584 is telling you.

It was thought that insulated bus and barriers betwen phases would prevent arc propagation as I described. This has been thoroughly disproven in actual lab testing. If you have separate ENCLOSURES such as in switchgear then arc propagation does not occur. But in panelboards categorically you rate it according to the incident energy on the line side of the incoming cable, ignoring all the breakers in the panelboard.