Any thoughts on when we should be implementing the new 1584? It will have a fairly large impact upon data collection requirements and also NFPA 70E makes some specific reference to the 2002 version of 1584. Should we wait to implement 1584 until 2021 when 70E catches up?

I'm trying to develop our company's position on this which has brought me here.
Annex D in 2018 70E, D.4.1 states that Users are encouraged to consult the latest version...
And the results I've looked at with SKM seem to be more conservative.
I will look at the differences in some more Studies before making a final determination, but right now I'm leaning towards using the new version from this point on.
John M

That is a question that I frequently receive. This week's "Question of the week" is to address this. Should be interesting to see what people think about the implementation time frame for this question.

Ethically, I think that we should be using the new standard immediately. However, in the real world, many clients will not want to pay the extra cost of data collection and entry.

We are now quoting any new projects with using the new IEEE-1584 2018 version. We have had to increase our cost a bit to cover the extra data collection and entry.

For jobs that are in progress, we are sending out a letter telling them that a new version of IEEE-1584 is available. We state that if they would like us to use the new version, then there will be a change order with a higher amount to cover the extra costs.

We are busy modifying our reports to incorporate the new standard.

I like your approach of notifying clients about the change and leaving it up to them to decide. Fair and square. Personally, I don't really see much problems collecting extra information required by new IEEE 1584 - 2018 (enclosure size if any and electrode configuration).

I went throguh this when NFPA 70E-2004 Annex "competed" against IEEE 1584-2002. The Annex was essentially an earlier prototype of the IEEE 1584-2002 equations. Eventually some comparison graphs came out and other details that put NFPA 70E-2004 to bed as a viable calculation method.

I am also struggling to determine how to implement this for my company. The thing I am struggling with is what electride configuration to use. From the descriptions it sounds like some equipment can be considered multiple configurations based on what task is being performed. For example draw out breakers. If the breaker is removed it’s different than if the breaker is in place when a fault happens. Is the horizontal always the worst case? Is always using that config overly conservative?

Yes and yes. Also compartments that are large or shallow can significantly reduce the resulting incident energy. Assuming a worst-case for either electrode configuration or compartment size is not a good practice, data collection should include these factors at least in some form.

That said, I think the electrode configuration for low voltage MCCs and panels will typically be reduced to a mere "does it have horizontally placed conductors aimed at a worker" with a calculation for HCB if yes, the rest being VCBB.

I'm not sure when VCB is a valid calculation for an enclosure. The sample photo for VCBB shows a breaker counting as "insulation plates". The standard doesn't offer guidance but it can be argued that fuses are insulation plates as well. That would mean any enclosure with a protective device needs to be classified as VCBB.

jvrielink thanks for your response! You confirmed what I was thinking.

We do field gather all information possible which will now include configuration and size. The problem (for me) with the electrode configuration is to come up with which to use when different activities in the enclosure could result in different configurations. Annex C.1 describes 3 cases for drawout switchgear. In this case do I have to use HCB (as worst case, Case 2, for any activity in this enclosure) because this is possible if a breaker is removed, even though this is unlikely when energized. Or do you think this is overly conservative?

I agree with you about low voltage MCCs and Panels. So I guess a panel with main lugs only would be VCB but any protective device would make it VCBB.

I also find it odd that the SKM default is VCB when that seems to be least likely to me..... any idea why they would do that?

Even if you do not have a true horizontal electrode, a large percentage of equipment has an "insulating barrier" configuration which calculates similar to a horizontal electrode. Example: Any feeder into (line side) a molded case breaker would also be an "insulating barrier" configuration, a good example is a panelboard with a main breaker.

Opinion - A basic fusible switch might not be considered a horizontal configuration unless they use a molded case switch as the line-side opening mechanism, then again, there are many line side configurations. I did a forensics on a horizontal bus fault. It was an ITE Bulldog bus plug, which was a pure horizontal configuration, the arc burned the wall 12-15 ft from the bus plug.

I know of many companies that rack breakers off of energized bus, for LOTO purposes. It seems like these would need to be treated as HCB.

VCB and VOA are the original (IEEE 1584-2002) configuration. So SKM defaults to the 2002 standard value. I believe somewhere in the standard it kind of hints at this being sort of a default.

Not sure if you really would assume VCBB for MCC for instance though. Most arc flash cases I've heard of initiate in the back at the bus which is VCB, not VCBB. There's really nothing exposed at the breaker so this doesn't seem like a likely scenario and this is the case in many circumstances. The classic case with HCB by the way is that some meter sockets have horizontal bus bars which have been known to have caused much higher incident energy burns than IEEE 1584-2002 predicted.

That is correct. There may be two configurations within the same enclosure . ETAP has introduced a concept called the enclosure editor which allows the selection of a maximum of two electrode configurations (ECs). When more two ECs are selected, the arc-flash program automatically runs an arc-flash calculation for each. The EC which produces the highest incident energy is reported as the final solution. So one can simulate VCB and VCBB in the same enclosure and get the worst-case incident energy results.

Is there a source or case history for this incident? And was it 480 V or 240 (or 208) V?

PG&E did extensive testing of meter sockets and other metering equipment and never had anything over about 4 cal/cm2, IIRC. So I'm interested in actual incidents where 1584-2002 calculated IE is thought to have been non-conservative.

Thanks,

Dave

EPRI did the testing for meter sockets. In the initial testing incident energy never got very high. But then a couple major incidents happened which were published by OSHA I think somewhere in the 2009-2014 time span but it's been a while since I saw it. In response EPRI set up some test cases with 480 V meter sockets with horizontal bus bars and confirmed that the incident energy can get very high. It's in their public (free) documents so not too hard to find.

i have seen a couple of 480V meter incidents, one of them is a well documented forensic evaluation, it is what i would call a typical horizontal arc, almost rifle like. The arc burned through a 40 cal coat, however, the pattern was confined to a 12" diameter area.

Approximately ten years ago, several major utilities started requiring a line side meter disconnect (480V only) , due to faults occurring when removing the meter.