Arc Flash Inconsistency Question

[TWD82]

Here is my system data:
V = 12.47kV
Isc = 8.338kA
Ia = 8.12kA
D = 914mm (36")
t = 1.1 seconds
G = 152mm (~6")

Using the IEEE 1584 equation in 70E I get an incident energy of 11.3 cal/sq cm

Using the theoretical equation [D.8.4] in 70E
I get an incident energy of 70 cal/sq cm

How can they be so different? I ask because my Gap is not 6", it is 8" and 70E states that the 1584 equations cannot be used with gaps larger than 152mm. So I must use the theoretical equations and state that the arc flash hazard is much much greater than it seems it should be. Using the 1584 equation you can see that I.E. vs. Gap is a linear plot. Is it a terrible assumption to think that it is also linear past 6 inches? I know I can't use the 1584 equations because they haven't been tested with larger gaps, but I have a hard time labeling the hazard as 70 cal/sq cm when it seems that the theoretical equations are grossly conservative. I am going to get the clearing time down to a little over 5 cylces, which according to 1584 equation is about .8 cal/sq cm. The theoretical equations gives over 5 cal/sq cm. Is the descrepancy between the two methods really this big or am I making a mistake somewhere?

Thanks,
Travis

 

[wareagle]

Bussman.com has a software program that you can check you figures.

 

[rbulsara]

Refer to all posts in thread238-171682.

 

[dpc]

I just checked your values using EasyPower and get the same results you get for both IEEE-1584 and NFPA-70E equations. You may want to go here and under "Technical Newsletter" look at "Putting Arc-Flash in Perspective" or something like that.

http://www.epc-website.com/

Your numbers are correct - whether either value makes any sense is another question.

 

[wbd]

dpc,

On Easypower, when doing arc flash calculations, which do you use? 1584 or 70E? and why?

 

[dpc]

I use IEEE-1584 because it's the best available recognized standard at this time. Certainly it has its own problems and inconsistencies, but because I'm a consultant, I want to use a standard calculation method that is not just my personal opinion of what arc-flash level should be.

And I put "based on IEEE 1584-2002" on all the labels as well.

I do think it's more accurate at 480V. It's probably somewhat low for medium-voltage, but the Lee equations are probably way too high.

 

[wbd]

Well, we're on the same page...I also only use the IEEE 1584 as it is a recognized standard. I'm also using Easypower 8.0

 

[Electic]

NFPA 70E calculation method is not recommended for over 600V calculations. Some software vendors (e-tap) have disabled the capabilty to calculate Arc Flash Exposure using NFPA 70E for over 600V.

 

[TWD82]

Thanks for all the input. This won't be such a big deal when we get the clearing time reduced, but until then how would you quantify the hazard? I am tempted to use the 1584 equations (even though they are not valid past 6"), increase them by some margin and then write up an explanation as to why, but I'm not sure how this would hold up from an OSHA standpoint.

Thanks Again
Travis

 

[Electic]

As the spark gap increases, I would expect the arc resistance also to increase, so there would be some cancelling of the effect on Arc Flash Exposure.

Arc resistance is a big part of these calculations and impedes quick clearing times one would normally expect of a fault in these situations (that is another gripe I have with the forumula as it exists).

But everything is emperical, and not much linear extrapolation is possible.

I would use IEEE 1584, and play with the default spark gaps length, but also see what is happening to the arc current as the gap is increased.

 

[stevenal]

What is the purpose of increasing the gap? The the gaps are standardized based on the voltage level. The gap is not the phase to phase or phase to ground distance, but the likely gap created by an incident. The 1584 spreadsheet does not have the gap as an input, but does a table lookup. The tables are based on 70E. With each standard referencing the other, it's hard to say which one rules. Meanwhile another IEEE committee, the NESC, is going with Arcpro.

 

[dpc]

What type of equipment is this and what is the basis for the assumption of an 8" gap?

As long as you have a reasonable basis for your PPE requirements and are consistent in application, I don't think OSHA will be a concern.

Of course plaintiff's lawyers are another matter...

 

[TWD82]

This is medium voltage switchgear. I got the 8" gap by measuring the seperation of the stabs on the back of the breaker. I was under the impression that the Gap was the acutal spacing between the energized parts, is this not correct? Table D.8.2 in 70E has a column labeled "Typical conductor gap" based on voltage level and type of equipment... but even their suggestion for Gap in MV swgr is out of the range of the 1584 equations (only by 1mm though). I do not have IEEE 1584 so I wasn't aware that Gap isn't used as an input into the spreadsheet, I was simply using the 1584 equations found in 70E. I've reviewed the NESC '07 Arc flash requirements and will be using their table for transmission/distribution switching, but these tables do not apply for indoor switchgear.

Travis

 

[dpc]

I think the IEEE 1584 typical gap data is based on bus spacing (actually space between the phases. That is where the 152 mm comes from. I'd just use what they use.

 

[stevenal]

The instructions in 1584 do not say to measure clearance, but to use Table 2 of the same standard. The spreadsheet uses the more comprehensive Table D.8.2 from 70E, and a whole extra millimeter for this class. I suggest you obtain the standard before getting too deep into your calculations.

 

[stevenal]

I see IEEE Table 4 is identical to 70E Table D.8.2. Maybe IEEE published it first.