Another significant change to the 2018 Edition of IEEE 1584 is the addition of more enclosure sizes to those that were used in the 2002 edition (sometimes thought of as “default values�.

Also, a new enclosure size correction factor (CF) calculation was added to adjust the calculations for actual enclosure dimensions.

The new 2018 equations are based on an enclosure sizes normalized to 20� x 20� x 20�. The enclosure size correction factor CF can be used to adjusted the calculations for other sizes. If the actual enclosure is larger than 20� x 20� x 20�, the correction factor CF can be determined and then used to calculate a more accurate (and lower) value of incident energy and arc flash boundary. Not including the correction factor would result in higher (more conservative) values. Shallow enclosures were discussed a few weeks ago.

At first look, this appears to add a degree of complexity to the arc flash study. However, based on how studies were performed with the 2002 IEEE 1584 equations, many may ultimately try to simplify the process. It has only been a few months since the 2018 edition was published and people are still trying to understand it all so let’s see where the thoughts are with this week’s question.

How will you/consultant/company address enclosure sizes?

Use 2018 IEEE 1584 default sizes
Keep as 20 x 20 x 20 inches (508 x 508 x 508 mm) (worst case)
Use actual enclosure size measurements
Not sure yet
Something else (please describe)

Seems like a client question.

First off generating an unrealistically high result is NOT, I repeat NOT conservative. In fact we don't even need to do that. We just run around and label everything "category dangerous" like some software companies do and they don't have their licenses on the line. If that was the case then why waste time on IEEE 1584 anyways? We can just use Lee for that! Overprescribing PPE and requiring PPE where little or no hazard exists decreases visibility and dexterity which increases the likelihood of human error and thus leads to an increased probability of an arc flash and thus an injury.

Second let's consider incentives here....WIIFM. For the PE there are twin incentives. Going with a high number, even an obviously wrong one, is generally considered the "conservative" approach though with arc flash we know how that works when it comes to using say ANSI bolted short circuit values...garbage in, garbage out, and dangerously so. The twin incentive for the PE is that the vast majority of the billable hours are in the field survey hours and that adding CF simply increases the overall billable time. Thus the incentive very much favors doing the calculation. There's no money in not doing it.

The incentives for the client/customer/end user are a bit more of a mixed bag. First off using CF decreases the results, sometimes substantially so especially when it comes to old medium voltage results. Lower values means less cumbersome and less need for PPE so it lowers risks, improves attitudes about arc flash, encourages PPE use, and improves the time it takes to do a job. In short there are no non-financial negatives to doing it. This is juxtaposed with the fact that on a new study the cost is a wash. It takes very little time to use a plastic ruler and a tape measure when you're already having to go through just about every cabinet in the first place. Many of the results are going to be identical anyways (standardized switchgear and MCC enclosure sizes for instance) so it just becomes a book keeping exercise to keep them all straight. On a 5 year review though it depends on how the review specification is done. If the 5 year review is essentially just redoing the initial survey, it changes nothing. If it entails a simple review (has anything changed) then the cost is substantially less if surveying is not done, and the client has already been following the IEEE 1584-2002 results for years so there is little financial incentive at this point to change. This is the #1 driver for not using CF.

So in summary the engineer is incentivized to use CF. The customer is incentivized financially not to use it during a review but there are lots of practical non-financial incentives to use CF. So once again it all comes down to money. As the guy going into the cabinets, I'm all for using CF. I distinctly remember seeing absolute crap come out of SKM for instance when running numbers on 35 kV class gear where it produced ludicrous results where in a plant with 50 years of history and dozens of substations, we had never actually seen anything nearly as crazy as the results SKM produces for >15 kV when it defaults to Lee.

Anyone got a source for a non-conductive tape measure that reads in English units?

I've only been able to find variations of these: https://www.amazon.com/Sekisui-Non-cond ... pe+Measure

I voted: Use 2018 IEEE 1584 default sizes

I am assuming this will be the default for software.

Like in the past, select the class of equipment and the default IEEE values are used - enclosure size, gap,working distance.

If you want to change any of the default values you can but I don't believe most would. It would seem unrealistic to go measure each enclosure, cubicle etc. The cost would skyrocket.

I voted for using the actual sizes. I've done several projects that were a follow-up to an arc flash study in which the plant wanted to reduce the arc flash hazards. The main goals is to get that incident energy below 8 cal/cm2 level so they don't have to wear the AF suit & hood. If they can get it below 1.2 cal/cm2 even better.

So if I take a little bit of time to get some dimensions and use that to get a more accurate result that is sure to get a reduced incident energy, it's time well spent, and certainly less than re-engineering the plant electrical system. For most equipment like panelboards, switchboards, switchgear, and industrial control panels, it doesn't take that long. The only place where it may get tedious is MCC's. Most of the time they only want one label per MCC section, so maybe that defaults to the smallest cubicle in the section or you use the full section height. Haven't figured out the best approach to MCC's.

Ideally, software companies will start putting dimensions in their libraries for the bus equipment. For example, SKM has UL 67 200A panelboard as a library bus equipment. If that just selected a 20"W x 5.5"D x some height, it'd save a lot of time.

I think the bigger issue will be deciding which electrode configuration to use on equipment that has multiple configurations.

I also agree that the dimensions should be used when ever possible. Wearing an Arc Flash Hood and suit is very uncomfortable and usually very hot. We as professional engineers need to do what we can to make the electrician's job easier and more comfortable.

For new facilities, the majority of equipment submittals will have dimensions on the drawings. In the field, it only takes a minute or two to make the measurements and write of a form or enter them directly into the PC.

We are currently measuring the smallest cubicle or enclosure in the MCC, Switchgear, and Switchboards. Then the entire MCC, Switchgear and Switchboard will be labeled with this worst case enclosure. Except for the main and bus tie cubicles, we are not measuring each cubicle in the MCC, Switchgear, and Switchboard and then producing a separate label for each cubicle. I do not believe that our clients would want to pay for this. However, we will be flexible and if a client wants labels on each cubicle, then we will bid the job with this extra cost.

I think you are WAY undersizing your dimensions for the MCC's for a couple reasons. As you know we look at worst case which is on the bus side of the fuses or breaker within the bucket, not the load side. The arc flash that you are modelling represents an arcing fault that occurs across the bus at best maybe across the top of the breaker/disconnect terminals but more typically it represents an arc across the vertical bus bars that usually travels down the length of the bus bars and fires at the bottom of the bus bars behind the little door at the bottom. The MCC is effectively all open from that point of view. So a lot of MCC's are going to be 20" tall by 20" deep by 84" tall.

If you are modelling the dimensions as the bucket itself then you need 1 label per bucket and the overcurrent protective device is the MCP/breaker/fuses because this is a load-side consideration which is generally very low in the first place due to the fact that the load protective devices will be very fast acting (as fast as possible) and very low trip settings.

See this article for examples of what actually happens in an arcing fault in an MCC:
http://www.neiengineering.com/wp-conten ... s-2012.pdf

There is more of this out there but it's a little harder to find good examples.

This also points out what is essentially a fallacy in the IEEE 1584-2002 equipment table that gives "MCC's" as an example and uses the 20x20x20 enclosure as the default size. That size might be appropriate for a combination starter enclosure but not an MCC which is more like 20" x 20" x 84". It also points out one of the errors made frequently when it comes to arc flash boundaries. The arc flash boundary and the working distance are the distance from the potential arc, not from the front of the enclosure. So if the working distance is 18", it is measured to the bus at the back of the MCC bucket which is usually about 12" away so the actual working distance from the front of the bucket is more like 6". This also affects arc flash boundaries in the same way.

See also the original (Part 1) article here espousing the typical theories about for instance insulated bus which are proven to be false in practice as per Part 2 above:
http://www.neiengineering.com/wp-conten ... s-2012.pdf

Thanks for the input Paul. I think that mostly answers the MCC height dimensions question. Use the full height. So let's finish this discussion for everybody's benefit. Most MCC sections are divided into two, the front where the starter buckets go, and the rear where the horizontal and vertical busbars go. So lets finish out the other two dimensions.

Depth = full depth of MCC or
depth only of the busbar compartment or
worst case of depth of busbar compartment/starter compartment?

Width = Full width of the MCC section or
width of starter bucket (which equals full width of section minus wiring compartment)

Agree. I should have said "highest incident energy" which I did with this week's question. Conservative depends on the complete context.

To Everyone: Good conversation. As many are recognizing, there is quite a bit to be sorted out regarding the application of the new 2018 edition of IEEE 1584. We went down this road back in the early days after the 2002 edition was released and incorporated into software.