Do you take every last panel off to verify the wire sizes? Or do you assume the contractor used a wire adequately sized for the breaker? I've seen it done both ways. Usually when the panel is newer we assume the contractor sized to code. When its older gear or a smaller outfit we will verify every panel with our eyes.

Thanks,
Mike

How can you honestly verify wire sizes? Even if I use a plastic nonconductive caliper, there is enough variation between compact and normal conductors among other things that measuring it is questionable at best. And frequently the lettering on the cable has faded even within a year or two so badly that it is nearly impossible to read it.

As to the assumption that the proper conductor size was used based on the breaker, boy that's a can of worms. Why would you assume that the "correct" as in I guess Neher-McGrath cable was even used in the first place? And how would you know? What if the installer required an assortment of sizes and simply sized everything up to a single convenient size or used whatever is on the truck or left over? What about "odd" sizes such as 3/0 that are rarely used and usually scaled up to more commonly available sizes? What if it is upsized based on the fact that some place unknown to your immediate area and inspection there are multiple conductors in a single raceway so it must be derated and sized up from the usual 3 conductors in a single raceway? And if you miss any of this you end up screwing up the short circuit calculation

Conductor sizing in my experience whether measured or assumed is ALWAYS challenging to say the least. Your two answers hardly even scratch the surface and this is a fairly "deep" topic since it drastically affects OCPD opening time.

I generally casually walk the building first then go back to my office and build the model first draft based on the existing drawings, then perform a detailed survey. My detailed survey involves removing all panel covers to get breaker data AND verify wire sizes to the best of my ability. Most of the time I can read the printing on the insulation. When I can't read the jacket I can usually judge it pretty accurately just from experience and compare what I see to what the original drawings show. If what I see looks like the old print I go with the old print size.

Yes, we check the wire size and length.

We verify wire size as well. I would say 95% of the time we can read the gauge of the wire on the insulation without issue. We also try to verify length of cable run best we can.

We will try to read the writing on the cable. If it is worn off, then we use the plastic nonconductive caliper.

Sounds like everyone here is trying to do it right. Some of the larger equipment firms "assume" wire size with a caveat statement in their study that lists or notes assumptions have been made. This reduces their data collection time and their worker's exposure but may compromise the study.

The best study would verify wire size since as Paul said, it really can affect clearing time of the device and line impedance calculations in the software.

All our studies are done by our engineers and they gather their own data in almost all cases so the study is as accurate as it can be.

Hugh Hoagland
e-Hazard.com

Nice conversation.

Another thing to consider before assuming a conductor size is NEC Article 110.14(C. This was introduced in the 1993 NEC and it basically states that the conductor temperature rating used to determine the ampacity can not exceed the device terminal's temperature rating. i.e. 75 degree lug rating means a 75 degree maximum ampacity rating.

What this effectively did was limit using a 90 degree ampacity and instead the 75 degree ampacity rating is used. (slightly different for 100 Amps and less)

Pre 1993 (and later years for those that did not get the message) conductors may not be sized as they are today.

An example - assuming everything is normal, ambient temp, number of conductors in raceway etc:
4/0 Copper - 75 degree ampacity = 230 Amps
4/0 Copper - 90 degree ampacity = 260 Amps

Pre 1993, it was not unusual for someone to use a 4/0 for a 250 Amp circuit. Today that would be a 4/0.

Also, conductor sizes may not match the protective device rating if the conductor size was increased due to voltage drop.

There are a few other exceptions too.

As many here realize, determining conductor sizes can sometimes be a challenge.

I really like that everyone here seems to be singing the same song and understand the importance of having good impedance information in the model, including the wire size, raceway and length. My process if very similar to bbaumer:



I typically build the model based on the existing information provided by the client, collate my forms and then do a walk down of the site with a qualified electrical worker allowing us to get in all the panels. However, if they don't have ANY study today, I want to minimize my (and their) exposure to energized parts. While during the survey we will be typically opening panels that haven't been opened since they have been installed, and sometimes with little to no maintenance.

Instead I will do something similar to what Hugh mentions.



For the first pass of the study we will try to visually verify all cables that can be seen without removing covers to expose us to energized parts. If we can't see the label we will use calipers like Robert does. For a large facility this means that there may be a lot of cables that we have to "guess" at. This is noted in the model and then we will ask the client to create work orders to gather this information during normal shutdowns; and if there is to be energized work in areas that are affected by these limitations that they then get the information with the IE as we calculated and re-run the study for that area of the plant.

This allows a balance between exposure to the hazards during data collection and an accurate model. It also helps me communicate to the client that this is a continuous process, an arc flash analysis is not a one and done activity.

Regards,
JM

We have the table based approach to address exposure when no study exists. Hugh and others did a pretty extensive data collection effort on about 50+ bonafide arc flash cases. Even following the table based approach (as it exists as of 2015) offers decent protection most of the time and even for those cases where workers are under-rated but still wearing arc rated PPE of some sort, it works quite often. So as maligned as the table based approach is, it's not half bad.

Now if someone were to use the graphical approach presented at the last ESW on the circuit breakers and fuses we'd have absolute worst case analysis and achieve IEEE 1584 empirical model results with much less effort. I think we could in all likelihood have a better table method but it would take a lot of time pulling all the data together.

Thanks Paul.

Can you link to the presentation that you are talking about? The paper that it was based on should be on IEEE Xplore.

JM

Basically take the TCC from a particular OPCD and calculate incldent energy at every point given the TCC. Then find the maximum value.