Author 
Message 
Matthew Mullins

Post subject: Incident Energy dependence on voltage level Posted: Fri Aug 09, 2013 7:46 am 

Joined: Fri Aug 09, 2013 7:39 am Posts: 3

I've always understood that Incident Energy was a function of the arcing fault current and how long that arcing fault current takes to clear based on the protective device (assuming a fixed distance). I'm curious thogh what role the voltage level plays into the IE calculation? I dont believe the equations used in 1584 take voltage into account? I've seen several cases where the IE at the 480V level is far worse than at the 4.16kV level. Sometimes this may be due to a faster clearing time for the 480V location but I have seen incoming service entrances that had no upstream protective device to relay on and therefore capped the calculation at 2seconds. For these cases with both the 480V and 4.16kV calculations capped at 2 seconds I've seen that the 480V locations have provided worst case results. Ist this strictly a function of the fault current (more fault current at 480V) than any voltage dependence? Thanks


Top 


PaulEngr

Post subject: Posted: Fri Aug 09, 2013 9:15 pm 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2174 Location: North Carolina

Roughly speaking, energy = power times time. Power = volts times amps. Now you have to take into account things like power factor and many other little nuances but the basic equation is there. Incident energy is no different except that there are some exponents involved. One of the major factors is that an arc has a voltage drop which is usually around 100150 volts. So at low voltages this has a significant impact on the available arc power. At higher voltages, the voltage drop really doesn't play a major role. In a similar way at lower voltages due to the voltage drop, arcing current is often less than bolted fault current while at higher voltages, this difference disappears. The differences you are describing are typical and indeed are the result of current and voltage differences. For instance it is extremely difficult to get reasonable incident energy values downstream of a 2500 kVA, 480 V transformer. The situation is very different with 4160 V. Don't forget that transformer impedance and thus available fault current is also quite different.


Top 


Beo

Post subject: Posted: Tue Aug 13, 2013 2:37 am 

Joined: Fri Jun 08, 2012 10:53 am Posts: 39


Top 


Matthew Mullins

Post subject: Posted: Fri Aug 16, 2013 10:10 am 

Joined: Fri Aug 09, 2013 7:39 am Posts: 3

Thank you both for your responses, they were both very informative. After reading the paper Beo posted above I now understand that the Incident Energy has much more to do with the actual voltage drop across the Arc as it does the fault current. I've always wonderd how to reationalize the fact that the IE in higher voltage systems is typicall lower then that of a LV system. I've always understood that for a given system size in kW yes the fault current decreases but at the same time the voltage is also decreasing so I always thought that keeping power in the system the same the Incident Energy should be roughly the same when comparing the two systems. I now understand that although the fault current is decreasing the voltage drop across the arc is not increasing in proportion to the reduced fault current which results in a lower Incident Energy calculation. After reading this article I opened up my software and played out the 2500kVA example that PaulEngr suggested above. I esentially modeled (2) 2500kVA transformers both having infinite primary sources, and gave one transformer a secondary voltage of 4.16kV and the other 480V kV. I gave both transformers an %Z of 5%. After running the fault current on the two secondary buses I found that the 60kA fault current on the 480V bus was aprox 8.6x the 6.9kA fault current on the 4.16kV bus which is what I would expect given these voltage differences on the same size power source. After running the Arc Flash calculation I noticed that the 480V bus had an IE of 129 cal/cm2 while the 4.16kV bus had an IE of 15 cal/cm2. When looking at this from a new perspective after reading the paper I see not that these results are a direct result of the gaps being 104 for the 4.16kV bus and 32 for the 480V bus. It is this gap distance that I am assuming is controlling the voltage drop across the arc gap as described in the paper. I also noticed that the acring fault current for the 4.16kV bus as aprox 97% of the bolted fault current, while the arcing current at the 480V bus was only 47% of the bolted fault current. I'm assuming that this difference in arcing fault current percentage at the two different voltage levels is again due to the gap and voltage drop across the gap. I'd like to get my head more around the fact that the voltage drop across the arc does not increase that much for higher voltages. Although the paper briefly touch on this topic I'm curious to understand a more detailed explananion or furhter references that I can read. Are there any that come to mind? How are the gap distances used by IEEE 1584 related to the arc gap voltage drop? Thanks for the help!


Top 


PaulEngr

Post subject: Posted: Fri Aug 16, 2013 6:34 pm 

Plasma Level 

Joined: Tue Oct 26, 2010 9:08 am Posts: 2174 Location: North Carolina

Gap distances in IEEE 1584 have to do with equipment design. There is no standard for bus gaps. The minimum distance that still passes a hi pot test is the only "standard". But when you look around there are plenty of engineering guidelines that specify these distances anyways such as approximately 1 inch spacing at 480 V, regardless of whether there is a standard or not. So I don't know how to say this exactly except to say that there is something of a "working standard" even if there is not a consensus standard. Similarly there is something of a "working standard" for electrical equipment in terms of bus placement despite the fact that by the standards, you can place the buses anywhere you want within a cubicle. This relative uniformity is the reason that the standard working distance table in IEEE 1584 exists in the first place. As for electrical arc phenomena, the "arc flash" literature is not very enlightening. You'd be better off looking for books such as "Electric Arc Phenomena" by Rasch which is widely available in a google search. It contains a wealth of information on DC systems and is somewhat oriented towards designing and building neon lighting. This book is extremely old but if you read it carefully you will find that arcs below around 28 V DC are essentially impossible at any level of current, and extremely unlikely up to around 50 volts. You may definitely also want to find books talking about circuit breaker design and read up the first couple chapters on theory of arcs. For the AC version of the same thing, google "High Voltage Circuit Breakers Design and Applications" by Garzon. You can preview the first several pages containing the theory on google books for free which covers the major points of what I'm referring to. These sources will not really be useful for relating arc flash to voltage but are very useful for understanding how electric arcs form and behave under controlled conditions.


Top 


Jim Phillips (brainfiller)

Post subject: Posted: Sun Aug 18, 2013 3:48 pm 

Plasma Level 

Joined: Mon Sep 17, 2007 5:00 pm Posts: 1615 Location: Scottsdale, Arizona

As usual, Paul has provided a lot of good information to the question here. I have a couple of things to add. The percentage of arcing current as a function of bolted short circuit current can vary significantly at lower voltages as you point out. An additional consideration is the Thevenin equivalent impedance at the point of the fault relative to the arc impedance. For larger magnitude bolted short circuits, the Thevenin impedance is small  hence the larger fault current. When an arc flash occurs, the arc impedance is large by comparison so there is a much larger percentage drop from the bolted fault current to the arcing fault current when the arc impedance is factored in. However, the IEEE 1584 equations do not actually provide a method to calculate the arc impedance  they just give you the arcing current. At lower magnitudes of bolted short circuit currents, the Thevenin impedance would be larger. When the arc impedance is added, it does not have as large of an impact on the already larger impedance of the bolted short circuit. Therefore, the percentage drop is smaller at lower values of bolted short circuit current. This becomes pretty apparent when you run several arc flash calculations at different current levels. At higher voltages (above 1 kV), the IEEE 1584 equations result in arcing currents that are only a few percent lower than the bolted short circuit current. To add fuel to your comment about the lower voltage drop at higher voltage levels, I was recently involved with a few arc flash tests at 6.9 kV which resulted in arc voltages in the range of 700 to 900 volts  confirming your comments. This has also been one of the drawbacks in using Ralph Lee’s theoretical equations at higher voltages. They are heavily weighted towards voltage and as the voltage increases, so does the calculated incident energy with the Lee equations  to very high levels. There are of course, several sources of good information such as those that Paul and Beo pointed out. I just wanted to add these few additional thoughts.
_________________ Jim Phillips, P.E. Brainfiller.com


Top 


321Liftoff

Post subject: Posted: Wed Aug 21, 2013 10:10 am 

Joined: Mon Nov 19, 2007 5:25 am Posts: 32 Location: Titusville, Fl.

Yeah we’ve scratched our heads over here attempting to determine the arc flash energy of a MV cable splice failure resulting in ground fault. Of course the IEEE 1584 calculation wasn’t built for such. Using the waveform captures from our intelligent Feeder CB’s, and determining the incident energy (IE) using the available power (via V&I curves in the time domain) at the point of such fault led us to some very different incident energies than using the IEEE 1584 calculations, hence much higher by as much as an order of magnitude. Videos of similar faults, lead us to believe the faults are conical in geometric shape, hence higher amount of IE directly dispersed. Suffice to say, we’ll be studying these type faults, by duplicating them in a manhole at the KEMA labs to determine what application (instantaneous trips, ground resistance, …), reduces the incident energies in the event of a fault, in efforts of protecting that someone working in the man hole. Hopefully from these tests, it will allow for an applicable/credible calculation to be determined. More to follow – mid next year...


Top 



Page 1 of 1

[ 7 posts ] 


You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot post attachments in this forum

