I have a question on what everyone has been seeing and general opinions.

In some cases we have a service entrance switchboard (480Y/277V). In this switchboard resides a main disconnect breaker. Most times the Hazard at this switchboard is Extreme Danger. Now, if there is bussing between this switchboard and an adjacent panelboard or MCC, generally the back sides of the enclosures are wide open. Beacuase of this, the threat of a flash traveling from the adjacent MCC or panelboard may damage the main disconnect. If the main disconnect becomes damaged it is unable to clear the fault and thus the risk hazard rating for the adjacent MCC or panelboard would be the same as the main switchboard. Now, if the switchboard and the adjacent structure are provided in completely seperate enclosures and conduit is ran between them in lieu of bussing, the risk hazard category for the MCC or panelboard may be reduced because we can assume the main overcurrent device will clear the fault. So, if this is the case what is the thought of providing conduit nipples or just an opening with bushing to run the wire between the structures. Would this also be a practice to mitigate the hazard for the adjacent MCC and/or panelboard or would you treat this simular to an open enclosure given in the senerio above? Would you use a fire caulk to fill such an opening? Do you believe this would help mitigate a hazard? If so what would you use?

Inserting upstream separate main breakers is a very common strategy to alleviate the problem. The biggest problem with this strategy is usually space....as in there isn't any. As an example on a recent job I had, the circuit breakers were molded case units that were mounted on the panelling of a dry transformer (7200:480, 1 MVA, all copper so impedance=low). Never even mind that the calculated incident energy value was. The downstream MCC's were main-lug only, and they were very old (1970's vintage) GE MCC's and pretty much everything was worn out.

When I replaced them we put in molded case breakers on the transformer with electronic trip units, and put in MCC's with full molded case circuit breaker main incoming sections rather than main lug only. Now the entire MCC (even though it's an 800 A rated MCC) is around 0.8 cal/cm^2, even on the main circuit breaker. The molded case breakers on the other side are still hazard "ridiculous" or whatever you would call the >100 cal/cm^2 range but we don't care. We can always power off the 7200 volt side fairly easily where the incident energy is only around 18 cal/cm^2.

In similar fashion, the primary for the 7200 volt sub is 22.9 kV. In this case we installed fused cutouts about 20 feet off the ground. It's a bit of a reach with a hot stick but doable. In this case, the increased distance makes the arc flash hazard again almost a nonissue. We were putting in air break cabinet mounted switches before. As of this year I stopped that practice because it's just not practical to mitigate the arc flash hazard (it's a mine....I have dozens of portable substations, and we replace a half dozen a year).

Thank you for the reply. I agree with your points but I don't think it fully answered my question. I attached this diagram to help illustrate what I was refering too.

Let me expand as well: The key question is how much seperation is required between the main and (in this instance) the MCC. Do you feel it makes a difference if larger conduit size and quantity is used (i.e. 1-2"C vs. 3-3" conduits for larger parrallel sets)?

Current modelling is based on thermal radiation only. No consideration is given to hot gas, plasma, etc. Never mind rupturing a panel (except for the ANSI test for arc resistant medium voltage gear but it's purely a functional test).

So although your question is a valid consideration, I don't think that there is currently a good answer. From a practical point of view though the open area available through the panel nearest the incident (even if the hinges do not fail) has much more surface area than that through the conduit hole full of wire. There likely will be some passage of hot gas through the opening but not a lot.

I am interested to see what others have to say on this. I agree with you on installing the main breaker in a separate enclosure. This is what we recommend on all new installations. On existing equipment like you describe, I think it depends on the IE on the load side of the main breaker. If your main breaker will trip fast enough to limit the IE to a Category 0 (default working distance), I don't think that arc will be big enough to damage the upstream breaker. If its a Category 1 or above, I would consider if the main breaker is within the arc flash boundary or if an arc can travel along an un-insulated bus or cable to the main breaker.

Just a thought.. But wouldn't the average bus way on a typical MCC disperse any shrapnel, heat, etc. better than a series of 1/2" conduit nipples? The bus bars are usually in a larger space (top, bottom or back) and shouldn't the forces should be disipated in that space??...

This issue is another variation on the same one with regards to whether or not enclosures themselves provide any significant protection...in other words, is there any additional protection offered when performing some function on non-arc resistant gear with the doors closed. The answer is of course yes. However, the problem is that there is no design data to determine what the result is.

In a similar way, there is broad agreement that the arc flash boundary is not a hard and fast rule. Being on the other side of a cinder block wall may provide significant protection. However again, no design data is available yet.

As to the references to heat and/or arcs travelling down a wire. It is well known that especially at higher voltages plasma is magnetically propelled along a pair of conductors like a rail gun. However quantifying this is not so easy.

Heat is also likely to travel quite far. However, it depends on the thermal conductivity (in a dynamic sense) of the conductor, since in this case there is not a whole lot of time for the thermal energy to travel. Funny thing about high temperature thermodynamics by the way...radiative thermal transfer is roughly 10 times more efficient than conduction or convection. It's the reason that heat shields around cement kilns and the like are usually nothing more than 1 or 2 sheets of corrogated roofing with 1" spaceers (unistrut) in between. I''d venture to say that so long as the adjoining walls of the enclosures don't rupture, unless the heat is transferring by convection (hot gases expanding throughout the area), the adjacent enclosure is probably fine.

But again, this is just my engineering "gut reaction" talking. I have no test data to back up this claim other than looking at the charred remains of equipment in the past.