I attended a seminar today and one of the presenters claimed there is typically a significant (well over 40 calorie, even saying 60 or 80 calorie "if you do the calculation") available incident energy amount in an enclosed metering compartment of a low voltage switchboard.

He claimed this was due to the fact the voltage input for the meters are tapped ahead of the main and are therefore unprotected and even though the wire is only #12 or #14 the wire length is so short there is a tremendous amount of fault current.

I scoffed and questioned that logic thinking the withstand rating of #12 or #14 combined with impedence wouldn't allow a sustained arc flash event at all let alone over 40 calories but didn't want to get into it with the guy as it really wasn't the time or place.

Thoughts?

http://www.litz-wire.com/New%20PDFs/Fus ... 011609.pdf

You need to look at the second equation though since the time is the critical factor. It is a practice in some areas to purposely use a fairly small wire on a pole to connect to the lightning arresters (#1 for instance) since this effectively becomes a fuse link for utility protection in case the lightning arrester fails by shorting to ground.

If you go down the resistance route this is fraught with problems because you have to consider the total circuit impedance. Not saying that it can't be done but hard to make a more generalized statement.

From experience if the statement was actually true then we'd see a huge mess in metering cabinets on a regular basis, which I haven't seen personally yet.

Thanks Paul.

I referenced withstand rating along your same line of thinking, I think.

The equations and curves posted above are for conductor damage. I am unaware of any approved methods for determining the clearing time of a #14 copper wire. Remember also that where arc flash testing is performed, arcs are generally initiated with a fine conductor that completely vaporizes without interrupting the fault.

I disagree with the statement the wire is unprotected since it is ahead of the main. The utility will have protection, even if it's on the primary side with very long clearing times. If this time exceeds two seconds, then cut it off at two seconds per IEEE.

If you are looking at this from the utility side, the NESC has a low voltage table that deals with the IE question. The table is based on testing.

In the arrester example, the utility is likely to have an upstream recloser or reclosing circuit breaker that will give the arc some time to dissipate after the wire separates.

Do not use the method you propose, it is not supportable.

Then the scare tactic consultants need to not propose and throw crap out there like arc blast, arc flash inside low voltage compartments, or the idea that arc flash is "hotter than the sun" with absolutely no supportable evidence either. If there is no supporting evidence (ie, documented cases 2nd degree or more severe injuries in the face/chest region) Garbage in, garbage out.

IEEE does not cover arcs sustained for more than a few inches. At low voltage (under 300 VAC) sustainable arcs have not been demonstrated over more than fractions of an inch even under extreme laboratory circumstances. But we have scare tactic consultants throwing utter unprovable and made up crap everywhere. In fact one of the worst that actually does good work also happens to be the one that put in the revisions for the revised 2018 70E Annex including wild claims about arc flash temperatures and arc blasts. And this is in a peer reviewed standard coming from a so-called expert that is clearly and obviously practicing nothing but scare tactics.

So what is the end user supposed to do when a consultant throws out this kind of bull crap? You can't prove a negative and yet that is precisely and exactly the position that we find ourselves in with these idiot safety consultants that are out there making these kinds of statements.

Don't get me wrong...meter sockets have a proven and very bad track record but there's a big difference between instrumentation compartments and low votlage gear with meter sockets sticking right out of it. EPRI has beautiful pictures explaining exactly what the concern is when it comes to meter sockets as an example.

I know nothing of consultants, and wasn't aware that any were involved here.

By my definition, low voltage includes anything at or below 1000V. Bbaumer's figures show he is not below 300, but is operating at 480V. 480 V will sustain an arc easily.

Conductors are not fuses, they were not designed or intended to interrupt faults.

This is the source of the crap about metering compartments:

http://ieeexplore.ieee.org/document/8003871/

Referring specifically to low voltage equipment (the author distinguishes it from medium voltage switchgear):

"According to the 2014 National Electrical Code [6] and
IEEE Std C37.20.1-2015, 2001 , 1993, and 1987 [7] , there
is a requirement for these transformers to have primary
fuses (identified in Figure 3); however, there is no
standard for the location of these fuses. The line side of
these primary fuses makes a direct connection to the
distribution bus in the switchgear which may have
available fault current in excess of 100 kA, depending on
the supply source. Therefore, theoretically, the same arc
flash potential that is on the distribution bus of the low-
voltage switchgear could also exist in the metering
compartment at the line side of the primary fuses. Figure 4
details this possible hazard location."

"Typically these primary fuses are fed from the
switch gear bus via small gauge wire such as No. 14
AWG . These unprotected runs of wire from the potentially
hazardous distribution bus to the primary fuse line side
terminals are the source of hazard that this paper will
discuss further. Due to the fact that there is no standard
for length of cable or location of primary fuses, the arc
flash hazard and risk assessment can vary in low-voltage
switch gear metering compartments and therefore should
be considered carefully. Incident energy calculations as
well as other arc flash hazard and risk evaluations specific
to metering compartments will be discussed further to aid
in that assessment."

The article then lists 5 tasks from the 70E task table and states:

"Tasks 1 through 4 require the qualified worker to wear
arc flash PPE regardless of the equipment condition .
Whereas Task 5 does not require the qualified worker to
wear arc flash PPE so long as there is no exposure to a
circuit energized to greater than 120 V. In short, taking
measurements on a 480 V circuit or 240 V circuit requires
arc flash PPE. Taking measurements on a 120 V circuit
978-1-5090-5288-2/17i$31 .00 © 2017 IEEE2017 -PPIC-0262
does not, as long as the worker is not exposed to a circuit
energized to greater than 120 V. Still, the primary fuses
for the VTs and CPTs could be located in the metering
compartment, in which case, there would be exposure to
480 V or 240 V"

And here is the first issue: the word EXPOSED. Exposed means that it is not insulated, guarded, or isolated from inadvertent contact.That's the first logic error. A lot of switchgear and MCC's for that matter contain little or nothing which is exposed. The power conductors are all recessed or covered in such a way that inadvertent contact is eliminated. That isn't true of everything. I just got through replacing three contactors and an overload relay dated February, 1947. Believe me, they didn't worry about such things back then. Many times the contacts on the door equipment (back sides of push buttons) sticks out and definitely qualifies as exposed to the point where some equipment even comes equipped with a tarp-like arrangement to eliminate this. But within the metering enclosure itself, few parts are exposed. If we assume that the word exposed does not mean what it is defined as in 70E and instead maybe extrapolate it to something like "higher voltages are in the vicinity" then the 120 V task in the task table is totally meaningless because at that point the only consideration is the maximum system voltage within a compartment.

Sorting out which equipment is "exposed" and which isn't, isn't that hard. It takes a couple minutes of visual examination. And although the Code itself is not retroactive and would only apply to new installations going forward, NEC now requires not only a "Danger--High Voltage" label on doors that cover exposed high voltage circuit parts but a new (2014) rule now requires a similar label on low voltage equipment. Although not required, it might make sense to simply do a survey of all electrical equipment. Remove the "danger -- high voltage" labels that are simply slapped on everything medium voltage whether exposed conductors exist or not, and install labels on doors for BOTH high and low voltage equipment where exposed conductors exist. Then electricians need not question going forward whether or not arc flash PPE is required. If it has a label, it does. Otherwise it falls into a determination as to whether or not the task requires PPE (as mentioned previously). Since the metering compartment would need to be opened and documented for arc flash calculation purposes anyways, the labeling can be done at that time. Most of the time relatively simple efforts to isolate or insulate can eliminate exposures in the first place, which following ANSI Z10/NFPA 70E elimination of the hazards is mandatory before PPE can be considered.

The final concern in this scenario would be of course whether or not some activity (interaction) by the electrician could cause an arcing fault. OSHA 1910.269 for instance talks about scenarios where dropped tools can cause an arcing fault. With some older CPT's and PT's I could easily envision this happening and I've actually seen cases of this sort of thing. Another concern would be probing with meter probes or a screwdriver for instance and causing a short between two phases or phase-to-ground. That's the purpose of insulated tools and the little covers on the ends of the meter probes. If the exposed metal is less than the arc gap plus the breakdown distance for air (close counts, too), then it's not possible to cause an arcing fault by taking voltage/current readings.

Moving on to consideration of incident energy in the event that conductors are exposed...

The conference publication is a little unusual in that the conductor gap was chosen as 13 mm. This is very narrow for 480 V equipment so it's not really representative, but moving on...the highest incident energy for 208 V equipment in the conference publication considering effectively "all" configurations (about 6000) of metering equipment came up with 34 cal/cm2 for the line side of a PT fuse. However it doesn't consider the likelihood of sustaining an arc which is what NESC does which gives a maximum of 4 cal/cm2 based on actual equipment tests. It gives a range of 20-157 cal/cm2 for 480 V equipment which again is based on crazy high available fault currents (up to 200 kA) and an arc gap that is not representative of 480 V equipment, and again doesn't considered sustainability when the #14 wire vaporizes leaving an arc gap of several inches or feet. On the load side it indicates that incident energy is less than 3 cal/cm2 with a 13 mm arc gap, cable size #10-#14 and under 100 feet long, and in particular that available fault current is 200 kA. Lee method is being used over 106 kA but even below that except for very low (10-15 kA) available fault currents the results don't look good.

Based on lack of documented evidence of this scenario actually happening I'm still very dubious. The conference publication shows that it is theoretically possible to cause a significant arc flash in a metering compartment where the high voltage cables are in the compartment which is often true in 600 V class equipment if the system voltage is 480-600 V. NESC gives much different results based on actual testing which shows that IEEE 1584 is not applicable here and I'm contending that it has to do with cables vaporizing. It's true that #14 or smaller wires are used as "fuses" for IEEE 1584 testing but the arc gaps are under a couple inches. In this scenario the heated air from the arc sustains the arc and there is enough mass from the busbars used that the wire is just a fuse and plays no role once the arc starts. This is very different from vaporizing a foot or conductor to where arc sustainability becomes questionable at best. However not withstanding whether or not this scenario is realistically possible, which is what I question, keying in on considerations of exposure eliminates the issue.

Tom Short and Marcia Eblen did a fantastic paper on meter sockets. I was impressed on how a small change in the design makes a huge difference in the damage done in a meter socket.

I agree with Paul that the potential energy could be large even with a small wire. stevenal is correct that the upstream is protected but there is no discussion of in this post about that protection clearing.

We have been working with utilities trying to figure out the IE at meter sockets. Of course this only applies to 480V-208V three phase meters. Any other voltage would have PT's. Single phase calculations just don't exist.

Even if you don't have the article, this is available as one of the free publications on the EPRI web site. What they were measuring is an example of the "plasma" effect where if you have horizontal bus bars that effectively behave like a rail gun, they will launch "plasma" out. That's only the tip of the iceberg though. On top of that work done in Hoagland's group shows that woven textiles such as treated cotton are roughly HALF as effective against it so the ATPV 40 suit might only be 20 ATPV in comparison. However rain suits are about twice as effective. So it seems like the only effective way to deal with these things is either to use one of the new meter socket tools that have been developed, replace them with a different design meter socket, or work de-energized only. At the very least the combination of the two reports would seem to show that there's something going on here that we don't fully understand that is far more dangerous than it would seem. There have been multiple reports of severe injuries in OSHA logs from meter sockets so this is not a theoretical concern.



I don't think you realistically can. None of the IEEE 1584 "box" models match a meter socket in any way. So the only option really is to use a table such as the one in NESC that uses EPRI data for almost all of the values and simply takes the worst case that EPRI found in testing. That way you will know that this is the worst that it can possibly get to. The meter socket designs that Eblen and Short tested are fairly easy to determine by visual inspection. I know that the desire is to do some calculation here but we don't have 300+ test data points on meter sockets. Eblen and Short might have a dozen. That's enough to upper bound the results but not to quantify them any further.