Arc Flash Incident - Calculation Help 1

[coloradog1]

Hello guys. I posted this on the arc flash forum website but that site is not frequented near as much as this site.

We had an incident where a 480V cable was run overhead for temporary power. A large vehicle was passing underneath the cable and the cable got caught on the vehicle. The insulation on the cable was damaged and caused arcing to the vehicle.

I have been asked to analyze the arc flash potential from this incident and I am a little confused on how to go about this for this situation.

The cable is fed from a delta-delta connected transformer so there should be no single-line-to-ground fault current. We have both SKM and ArcPro software programs. I was planning on using ArcPro for the analysis since it occured in open air but ArcPro is based on a single-phase model and since there is no SLG fault current I was not sure if using the three-phase fault current would give an accurate value. I also had questions as to the arc gap since it is a cable and not a bare wire installed overhead. Typically I would use a 2 inch arc gap for an overhead 480 V line.

Any help would be appreciated.

 

[rbulsara]

Talk about misunderstanding and misapplication of arc flash hazard calculations!

What is the purpose of this calculation? Or better yet read the preface to NFPA 70E-2009.



Rafiq Bulsara

http://www.srengineersct.com

 

[coloradog1]

I don't understand this comment by rbulsara. I understand that arc flash calculations are done to determine the hazard when performing energized work.

In this case, the question is what if a worker were standing close to the location that the cable faulted? Why wouldn't there be an arc flash event? What kind of energies would this worker be exposed to?

 

[coloradog1]

Of course we would not expect the people to be wearing arc flash PPE. The question is merely what cal/cm^2 would the person be exposed to.

I don't understand why your answers are so hostile.

Regarding what method we would use in other instances, that is my question. We typically use IEEE 1584 for arc flash events inside of equipment. We use Arc Pro for open air arc flash events such as substations and overhead lines. Since this is an insulated cable installed overhead, it would be an open air event. However, arc pro uses a single-line to ground fault current and since it is a delta-delta source, I wasn't sure what method would be appropriate.

 

[dpc]

You need to assume a three-phase fault. That will be the worst case for a 480 V cable. Most analysis software can deal with open-air faults - this is covered in IEEE 1584. You can use ArcPro if you want - it has multipliers for three-phase fault. Bear in mind that all these calculations are basically educated guesses. When I see results reported out to four decimal places I tend to question the qualifications of whoever did the work.

As for the gap, I would do a sensitivity analysis - you'll find it won't make a large difference in the results until the gap get pretty big. You should start with something like 1" or so.

You will also find that the assumed working distance will have a major impact.

The key factors will be determining a realistic bolted fault current and determining the approximate clearing time based on the calculated arcing current.

David Castor

http://www.cvoes.com

 

[HughHoagland]

I basically agree with David on HOW to do the calc. ArcPro is OK for 480V but most valuable where it has more data than IEEE 1584 and that is in medium voltage >600. Sounds like this wasn't a 3 phase fault since you only had one line. I would make this a single phase in ArcPro since much of the data in IEEE 1584 is based on 3-phase faults.

The question of arc gap is tough here but most of our research says 480V will tend to blow out at about 2-6 inch arc gap. This does depend on the fault current and equipment configuration. Do you have a definitive protection device clearing time from a SCADA system?

If not, you will have to assume a lot.

I have often used effects on clothing, equipment etc to approximate the energy levels to make sure a calculation is reasonable.

Sounds like you are going the right direction.

On the WHY do this calc, I'm a little unclear. Temporary lines do however cause a lot of injuries and fatalities because they are not often modeled at all. Modeling could help you design temporary set ups with a little extra protection so they are less likely to cause as much injury.

Best of luck,

Hugh Hoagland

http://www.e-hazard.com

http://www.arcwear.com

 

[dpc]

Welcome to Eng-Tips, Hugh.

David Castor

http://www.cvoes.com

 

[coloradog1]

Just wanted to add some clarity to the situation. The cable is 3-conductor cable so all three phases are contained within the jacket.

We recently purchased the ArcPro software since the software is based on single-phase faults it seems many utilities use this for overhead line arc flash calculations. Does anyone know how you are supposed to calculate the arcing current to input into ArcPro? I have no issues calculating the single-line-to-ground fault current but how is the arcing current determined from this for input into ArcPro? Does anyone have experience with this software? For IEEE 1584 we would take the three-phase fault current and use the equations to determine the arcing current.

Some of the other questions I had:
1. Anyone have any thoughts over whether I should be using ArcPro or IEEE 1584 (SKM) for this situation? My understanding is the results from ArcPro are much more accepted for open air single-line-to-ground faults. IEEE 1584 obviously always assumes three-phase faults.
2. I believe that the arc flash event in this example could have only been caused by a three-phase or line-to-line fault due to the upstream transformer being connected delta-delta. Since the source is delta connected on the secondary we would not get any fault current due to a line-to-ground current and hence there would be no arc. Does anyone agree with this?

Regarding why we are doing this calculation, it was requested by management to determine how serious this issue was and how bad someone could have been injured.

Thanks very much for the help.

 

[dpc]

ArcPro does not have a means of computing the arcing current from the bolted fault current. It comes from the utility industry and was used on higher voltage systems where the arcing current is generally taken as the bolted fault current (for line-to-ground fault).

I have done 480 V fault calcs in ArcPro by using IEEE 1584 equations to develop a ratio of arcing to bolted fault currents and entering that value into ArcPro.

But if you are trying to determine possible hazard, I would still suggest doing a three-phase fault calc, since this could have occurred and will be the worst case.

With an ungrounded system, I agree that the initiating event must have been a phase-phase or three-phase fault. But keep in mind that a fault that starts as a line-line fault will often escalate to a three-phase fault and this can happen in a fraction of a cycle.

The other possibility is that your "ungrounded" system already had a solid ground on one phase, then another phase went to ground creating a really nasty line-line fault. This used to happen often back in the days of ungrounded 480 V systems. Thankfully, I've outlived most of those.

Regards,

Dave

David Castor

http://www.cvoes.com

 

[jghrist]

ArcPro does output an arc voltage. This could be used to determine an arc resistance (R = V/I) which could be used in the fault calculations iteratively to get the arc fault.