1. It has been very difficult to predict LV arcing in order to make the tests repeatable. [I was told maybe only 1 out of 10 tests passed, guess which got published...].

2. Most arc faults happen with the doors open so all tests are invalid anyway. [This is also my experience and I heard of one yesterday].

3. The approach in the rest of Europe has been to deal with the arc after it happens, very pragmatic, whereas UK tries to prevent it happening at all (greater emphasis on insulation/segregation).

4. If you buy "arc protected" or "arc certified" switchgear you can tick a box and go back to sleep.

I don't know whether any/all of these are true, but the tick box attitude to purchasing strikes a chord with me (speaking as a switchgear manufacturer based in UK). One of the drivers for change in many areas of the UK are the insurance companies so if people (specifiers) think they can reduce their risk by ticking a box they will, even if they don't really understand what they are asking for.

Statistics: Posted by rogerg — Thu May 13, 2021 2:55 am

]]>

This point could perhaps be taken into account in a future version of IEEE 1584 ......

Best regards

Statistics: Posted by Ph. Aupetit — Tue Apr 27, 2021 5:55 am

]]>

The IEEE 1584 incident energy equations are based on a three phase arc flash so the direct effect of impedance grounding would not be seen. Here are a few points regarding impedance grounding and arc flash.

- Most faults originate as L-G so with impedance grounding it would limit the current and incident energy and not escalate into a three phase event. i.e. it would stop a large number of faults from becoming a three phase arc flash.

- Since IEEE 1584 is based on three phase arc flash, modeling single phase is not addressed and impedance grounding is not an issue in the calculations.

- The 2002 Edition of IEEE 1584 equations gave slightly different results for the three phase incident energy depending on whether the system was grounded or ungrounded/impedance grounded

- The 2018 IEEE 1584 equations did away with that difference. It had to do with the erratic nature of the arc during initiation which was filtered out be removing 2 cycles.

Hope that helps!

Statistics: Posted by Jim Phillips (brainfiller) — Sat Apr 17, 2021 3:12 pm

]]>

Thanks Jim for your response, your opinions and comments are always valuable.

Unfortunately, I do not know enough about the ANSI standard to appreciate precisely the differences in the calculation of short-circuit currents with the IEC standard.

But, notwithstanding the standard for calculating short-circuit currents, my question is that if the source feeding the short-circuit is a current source (for example power station unit with full size converter), I would tend to think that the arcing current should be taken equal to the bolted short-circuit current as the short-circuit current is then independent of the arc impedance.

But what, in that case , of the reduced arcing current ?

Philippe Aupetit

Trace software International

Got it!

As far as the current, that makes sense. But if the current remains constant and there is arc impedance, that would result in a larger arcing voltage and I'm not sure how that is reflected in the model. I believe you hit upon one of the areas that was not addressed in IEEE 1584.

btw, this is just an informal opinion and may or may not not represent the views of IEEE.

Statistics: Posted by Jim Phillips (brainfiller) — Sat Apr 17, 2021 2:56 pm

]]>

Unfortunately, I do not know enough about the ANSI standard to appreciate precisely the differences in the calculation of short-circuit currents with the IEC standard.

But, notwithstanding the standard for calculating short-circuit currents, my question is that if the source feeding the short-circuit is a current source (for example power station unit with full size converter), I would tend to think that the arcing current should be taken equal to the bolted short-circuit current as the short-circuit current is then independent of the arc impedance.

But what, in that case , of the reduced arcing current ?

Philippe Aupetit

Trace software International

Statistics: Posted by Ph. Aupetit — Wed Apr 14, 2021 6:09 am

]]>

Does anyone have any ideas on how to show the affects of impedance grounding in an Arc Flash study, or using SKM software, or other software?

Thanks,

Gabe Paoletti, P.E.

Paoletti Engineering PLLC

Statistics: Posted by gabepaoletti — Tue Apr 13, 2021 8:24 pm

]]>

I co-authored an IEEE paper several years ago comparing ANSI and IEC short circuit calculation methods and how it might affect IEEE 1584 results. Although most results tracked closely, as you might expect if the calculated arcing current was close to a transition from instantaneous to time delay, different methods could have an impact on the arc duration and incident energy.

So, IEEE 1584 is the model for incident energy etc. but it does does not address methods for determining the bolted fault current or clearing time.

Hope that helps clarify!

Statistics: Posted by Jim Phillips (brainfiller) — Fri Apr 09, 2021 8:14 am

]]>

In IEEE 1584, the arcing current and the reduced arcing current are calculated from the bolted short-circuit current to take into account the arc impedance.

This method is relevant when the different sources supplying the short-circuit are voltage sources.

But what about when the one or one of the sources feeding the short-circuit is a power station unit with full size converter (for instance large UPS or PV or wind power station unit with full converter)?

In this case, for the calculation of the short-circuit current (see IEC 60909-0_2016), the source is modeled in positive-sequence system by a current source. The value of the short-circuit current is therefore independent of the downstream impedance, in particular of the arc impedance.

My question is then the following: is it relevant to use the IEEE 1584 method when one of the sources feeding the short-circuit is a power station unit with full size converter?

Thank you for your contributions on this topic.

Best regards

Philippe Aupetit

Trace software International

Statistics: Posted by Ph. Aupetit — Thu Apr 08, 2021 7:46 am

]]>

The benefits of arc ignition protection zones are fairly limited as the testing conditions (such as closed panel doors) typically fall under the "no likelihood of occurence" in NFPA 70E Table 130.5(C). Work that happens where the testing conditions are invalidated, for example after opening a door or removing a barrier to do voltage testing, still require a risk assessment like for any other type of switchgear. So the hazard is the same, the likelihood of occurrence is lower for conditions that already had a low likelihood. You would need a very mature and detailed electrical safety program for this to make a meaningful difference in working procedures.

There are useful exceptions like LV switchgear that maintains the arcing class while racking breakers but I've never seen those in the wild.

Statistics: Posted by jvrielink — Mon Mar 22, 2021 8:03 am

]]>

I'm posting this here as it is really this new question of IEC 61641 in my North American world of NFPA 70E/IEEE 1584. I've done various studies based on the latter for my US clients.

To the best of my knowledge, North American standards like NFPA 70E and IEEE 1584 regarding arc flash have no equivalent to IEC 61641 for "arc ignition protected zones". I understand the IEC used to refer to this as "arc free".

In Section 1 (Scope) of the IEC 61641, it ends with the sentence, "This is a voluntary test made at the discretion of the manufacturer." At the very end of the standard in Annex B, it has this very curious note (that is suggested to need inserted right after the Section 1 last sentence):

"NOTE 2 The recommendations of this report are not acceptable in the USA, nor in Canada, nor in the United Kingdom."

Can anyone speak to why this was inserted. I have seen some misgivings in posts elsewhere regarding IEC 61641. This note appears to have actually written into the standard some sort of significant split. I can't help wonder if it is the "arc ignition protected zones" versus the rest of the standard's testing to something similar to IEEE C37.20-7.

Thanks in advance for any potential help.

Statistics: Posted by HenryOhm — Sat Mar 20, 2021 5:15 pm

]]>

Hello Jim,

Thanks a lot for your answer.

It's now quite clear and consistent.

We'll be able to finalize our arc flash calculation module in our software.

Best regards

Philippe Aupetit

Trace Software International

Great! Glad to help. Check back here if you have any other questions.

Statistics: Posted by Jim Phillips (brainfiller) — Tue Feb 02, 2021 9:39 am

]]>

It's now quite clear and consistent.

We'll be able to finalize our arc flash calculation module in our software.

Best regards

Philippe Aupetit

Trace Software International

Statistics: Posted by Ph. Aupetit — Tue Feb 02, 2021 4:14 am

]]>

After discussing this with various people, the reference to the examples in Annex D that use Iarc-min in the numerator of equations D53, D55 and D57 should be Iarc without the “min” This is consistent with the equations in Clause 4 of the body of 2018 IEEE 1584.

Have a great day!

Statistics: Posted by Jim Phillips (brainfiller) — Sat Jan 30, 2021 10:20 am

]]>

Thank you very much for your involvement.

Best regards

Philippe

Statistics: Posted by Ph. Aupetit — Thu Jan 28, 2021 7:25 am

]]>

Statistics: Posted by Ph. Aupetit — Tue Dec 22, 2020 10:23 am

]]>