After the next edition of IEEE 1584 is finally published (probably a few months yet if all goes well) what would be on your wish list for future work? NOTE: This is purely hypothetical – just seeing where the interests are. I'm sure you might want to see many of these but please select your number one pick.

Future work for the IEEE 1584 Working Group should include:

Applicable to fault currents > 106 kA (the 2002 and next edition's upper limit)
DC arc flash
Voltages above 15 kV
Single Phase
Low Voltage/Low Current Cutoff
Blast Pressure Model
Light Model
Something else – post it here!

I vote for low voltage/current arc sustainability.

Thanks! I added that one!

I concur with all of the above and would like to see this included [my 2 cents for what it’s worth];

Low Voltage Dry Type Transformers [ex. intermingled voltages 208 & 480 in the same enclosure]
And maybe
Transfer Switches where both sources are active in a common enclosure

What I would really also like to see not directly related to 1584 but In a way have a standard published [?? IEEE maybe a 1584 appendix??] for manufacturers to use to produce TCC’s in a single common file format that would include all the specific breaker information for a specific breaker or family of common breakers the library file could then be downloaded and used in any software across all platforms that can read the file format. The way I see it currently is that for curves that are not available [either older or newly manufactured breakers, protective relays, etc.], you have to send them to the software manufacturer [unless you are savvy enough to create them yourselves] they create the curve library and send it back for inclusion in your library and later include it in service pack updates. My questions is why are we burdening the software manufactures to create the curves when they are already created by the manufacturer, let them focus on the software? Also, I have seen discrepancies [although slight] between software package TCC’s when compared side by side with identical settings. Create a format whatever it may be a asci text file, csv, give it a .tcc extension and let all manufactures read from the same file [other industries already do this] and plot the same way.

There's not much to show here. You can run the analysis on either the primary or secondary side but frankly the secondary side is all that matters, for three phase systems. Since the primary side typically uses very, very long times on overcurrent protection to avoid inrush and then adding the impedance of the transformer and/or splitting the fault if it does a phase angle shift (delta-wye), arcing faults are extremely high relative to the primary side so not much to do here. On larger transformers typically the spacings are so large that the actual incident energy from testing actual equipment is much less, which is why NESC gives 4 cal/cm2 as the incident energy.



Uhh, two sources. You add them. Superposition. Done.



How about we all agree to just stick with the ANSI and IEC curve families and not come up with some strange exotic stuff? Then I can specify a curve with just 3 numbers (curve, tap, time dial). There are exact formulas for the shape of the curve so I don't need any more information than that. Then we have published standards to follow instead of every manufacturer reinventing the wheel. This is how we do it in medium voltage gear.

Jim,

What's the light model?

Good question!
It was more of a hypothetical question. There is some data from light measurements and I was curious if anyone had an interest in modeling light intensity but I don’t think it really matters. It is damaging and UV filtering is necessary.

Jim,

Just watched your video. Thanks, great info. Any thought to updating the IEEE spreadsheet for 2018? Perhaps it's not possible?

Based on the draft equations a spreadsheet is easily done. If anything there is less room for interpretation.

My biggest pet peeve is the gap between what IEEE 1584 covers and what NFPA 70, and NFPA 70E requires for labeling. The biggest example is 120/240 V 1ph/3W panels and other panels that fall below the minimums. So IEEE 1584 says these are outside the boundary conditions, yet NFPA 70 says they need a label, and NFPA 70E says that label shall have arc flash boundary and either arc flash incident energy or PPE. Many people have said they aren't covered by IEEE 1584, so they don't even get modelled or labels. No labels does not agree with NFPA 70. A paragraph that simply states for single phase panels of 240 V and below can use xx inches for an arc flash boundary and an incident energy below yy cal/cm2 or something to that effect. Give us some guidance so everybody is doing the same thing.

Other small things:
-- Gaps for pad mounted transformer secondary's below 1kV (and if there's anything like the distance exponent factor). Usually larger than the 32 mm for switchgear.

Big changes:
-- Do enough research on 15 kV overhead distribution lines to include them within the boundary conditions.
-- Add a standard that covers transmission level arc flash, i.e. what are the underlying principals that ArcPro uses and how can we bring those into a standard with reproducible results.

I'd like to see a study that correlates the NFPA70E edition changes to the actual accident reduction value in terms of severity or number of occurrences.
The current process is an evolution into a theoretical method of calculating energy, which may not actually achieve its desired benefit.
To elaborate, what if we just used the Table method in the first edition, which included the 0 and -1 PPE levels and stopped there. What was the resulting number of accident frequency and severity? Calculating arc energy is empirical at best, and subject to the variance of many field conditions. Does labelling an IE to the second decimal really provide value. NFPA is all other the place with Art 130, starting with PPE 0 and -1, then dropping them, accepting the 125KVA threshold, then dropping it, adding/deleting PPE. What started as a well intention exercise now has evolved into an owner's nightmare, and for what purpose. Its a safety recommendation, when it reaches a satisfactory reduction in accidents, it should just stop. I don't believe we will ever arrive at a universal equation that works under 'all' conditions. Does air density make a difference, how about humidity, dust accumulation, carbon on the contacts, etc?. Just too many variables. Achieve the objective, keep it as simple as possible, make it easy and economical to employ - then stop and analyse the results. I've never seen how the changes with every edition has made the real world safer. If the studies are there, please direct me. If not, lets study that!

As pope Julius II said to Michelangelo when he was painting the Sistine Chapel, "When will it end"?

the existing IEEE1584 year 2002 advocates the theoretically derived Lee Lee method for cases where voltage is over 15kV. Would the new upcoming IEEE 1584 foster the Lee method or would it promote a different method?

You won't find that. First off, adoption is not a step change, just like NEC where my state (North Carolina) just recently as of June went from the 2011 edition to the 2017 edition of NEC. Many plants similarly adopt at a slower rate.

What can be shown is that over the last 10-15 years there has been a 60% reduction in the number of incidents reported to OSHA. The problem though overall is that we're talking about a few handfuls of incidents every year. In terms of overall injuries, electrical injuries account for something like 0.4% of total reported accidents combined. Arc flash according to Caulder's tabulations is about half the occurrence of shocks. So trying to develop quantitative data about particular changes to any particular standard is going to be nearly impossible to find. The only reason that electrical accidents have so much attention is that despite the incredibly low number of incidents, it is the seventh leading cause of fatalities.



Actually, it's far worse than that. As you delve into the underpinnings of the method, it flat out doesn't hold water at all. The whole "2nd degree burn threshold" idea is a purely contrived target. However what we do have is that there is some theoretical evidence to show that when using this as the basis for developing safety strategies, it works 100% of the time, contrived or not.



One of the issues with the IEEE process is that it only moves forward when more or less there is a consensus, which is really hard to achieve in practice. This is as opposed to the NFPA process that produces an incremental update on a fixed schedule at least theoretically whether there is consensus or not. So changes get made much more often and small details often get missed when you don't have multiple drafts and refinements as you do with the IEEE process.

But all that being said, OSHA is not mandating that you use a particular edition of NFPA 70E. States freeze the state electrical Codes (NEC) all the time, sometimes for 10 years or more. You can voluntarily use any edition you want to use. If you have a business with about 100,000 employees, at least theoretically you can get an arc flash injury at least once a year. That makes it easily possible to track the impact of your program to some degree. But for most entities out there, that's a pipe dream. We have no chance of collecting those kinds of statistics, except in the large aggregate, and at that point things like versions and particular issues wash out quickly.

I will say this much though. The early editions of the tables (prior to the 2015 edition) were clearly and obviously wrong. They changed the PPE requirement depending on the task but the hazard did not change, only the likelihood. Also particularly in the 2004 and 2000 editions, face protection was often minimal or not required. And many plants followed the somewhat strained wording that basically said that if you have no exposed conductors, there is no arc flash hazard. We don't need to get into the statistics if we show a few cases where an injury occurred and others where it doesn't occur and that's exactly what we see in the article, "Update of field analysis of arc flash incidents, PPE protective performance and related worker injuries". See:
https://ieeexplore.ieee.org/document/6164450/

This encompasses about 55 actual arc flash incidents covering a wide variety of incident energies. So with this limited data we can do some things. In 43% of cases PPE worn met IEEE 1584 criteria and it worked 100% of the time. In 25% face protection was worn meeting that criteria. In just 22% PPE worn was arc rated. About 17% of the time when PPE is ether not worn or under rated worn by IEEE 1584 standards. The number goes up to over 50% if the PPE is FR even if it is under-rated. The table method prior to 2015 (the changes were largely cosmetic from 2000 to 2012) works 50% of the time. In just 9% of cases where arc rated PPE was worn irrespective of the ATPV, no injury occurred. In the sole injury case the incident energy exceeded the ATPV rating of the PPE. This is despite the fact that 45% of the time the PPE was inadequate. In those cases with inadequate PPE, it only failed to provide sufficient protection 20% of the time.

Thus we can conclude (granted severe bias in the data, etc.) that based on that data set without PPE at all or adequate PPE, the chance for a burn injury is around 80% in the event of an arc flash. If arc rated PPE is worn even if it is under-rated, the odds of no injury jump to 80%. Using the NFPA 70E tables works about 50% of the time. In the two cases where injury occurred no PPE was worn but the incident energy was 26 cal/cm2 in one case and 6 cal/cm2 in the other case. In the no-injury cases it was 2-20 cal/cm2 in one case and 8 cal/cm2 in the other, PPE was worn for the 8 cal/cm2 case but not for the 2-20 cal/cm2 case. So if arc flash PPE is the "minimum" PPE as it is in the current tables it seems conclusive that it would boost the no-injury rate from 50% to at least 80%. If the updated tables also encompass "under-rated" cases (hard to tell from the report), the likelihood might be even less but we can conservatively at least claim 80% based on the current tables.

So...not using arc flash PPE should still not cause injury 20% of the time based on this extremely biased data. It jumps to 50% with the old tables and a case could be made for 80% on the new tables, or simply an FR-mandatory policy regardless of task. It jumps to 100% with the full IEEE 1584-2002 approach.

Hope this helps clarify it a little bit. But as I stated in the beginning it is really hard to come up with solid conclusions like you are seeking mostly because of the rarity of the events.

I've just noticed that in Draft 6 threshold incident energy of 1.2 cal/cm2 (5 J/cm2) is hard-coded in arc flash boundary equations. Anex B.2 makes reference to a likelyhood of incident energy of 5 J/cm2 to cause the onset of a second-degree burn. The text reads "If a butane lighter is held 1 cm away from a person’s finger for one second and the finger is in the blue flame, one square centimeter area of the finger will be exposed to about 5 J/cm2"

It is well know and widely documented that only a fraction of 5 J/cm2 is required to cause the damage when heat flux is more than the 5 J/cm2/sec flux delivered by a butane lighter to a person's finger 1 cm away. I'm sure the IEEE committee is well aware of the fact but am surprised the new edition has not addressed the issue. I surely hope there is still time to have this grave omission amended before 2nd edition of IEEE 1584 has been released.