Arc flash calculations 1

[budhiman]

Hello All,

I am performing arc flash analysis on small MV system, 12.47kV is the service voltage with some 12.47 equipment and goes down to 208V. I have couple of questions on the calculations:

1. For arc flash, IEEE 1584 equations are used which are typically for 3-phase faults. Why is line-ground fault not considered for arc flash? Is the reason that line-ground fault quickly turns into a three phase fault in 15kV or small systems?

2. Now the code requires to install a arc flash mitigation system for systems rated for 1200A or more. These arc flash mitigation is typically designed for a 3-phase arcing fault. For example a maintenance setting is provided for a 3-pole 2000A breaker. The maintenance setting reduces the instantaneous to 5000A to lower the arc flash energy, but the ground fault is set for a instantaneous of 2000A, which is still lower than the maintenance setting. Why is still the L-G not considered for arc flash?

 

[davidbeach]

In switchgear there are no 1LG arcing faults that remain 1LG for even a whole cycle. Everything becomes a 3LG fault.

 

[budhiman]

Thankyou David. Is there any white paper or document or test data discussing this which I can read more on.

 

[davidbeach]

The IEEE standard might have something to say about it; what I know is what I've seen in the relay event files after an arc flash event. The initiating phase is clear but the plasma cloud expands so rapidly that all three phases are involved in a few milliseconds.

 

[ZeroSeq]

If you want to read about arcing ground fault dynamics, refer to chapter 7 of "Industrial Power System Grounding Design Handbook" by J.R. Dunki-Jacobs, et. al.

It is probably the best resource on the specifics in question.

 

[DanDel]

For the first question, 3-phase arcing faults are typically the worst-case scenario, and as already mentioned, L-G faults turn into 3-phase faults quickly. GF protection is not considered in AF calculations because the low trip setting would provide an erroneously low amount of AF energy. In other words, using a GF trip setting (which will not pick up on a much higher 3-phase fault) to calculate AF energy would not give you the worst-case energy value.

 

[stevenal]

I'll disagree with Davidbeach on this one. Maybe most, but not all. I have an event record from a switchgear incident. Line to ground 13.8 kV fault was limited by a grounding resistor for several cycles until it went line to line. 6 cycles later the circuit switcher tripped with no involvement of A phase at all. Three phase will be the worst case, however.

 

[dpc]

As mentioned, three-phase faults will be the worst case in terms of incident energy, so that is the basis of the IEEE 1584 equation. For open air substations and overhead lines, the NESC requirements are based on line-to-ground faults, the (dubious) logic being that three-phase faults are uncommon because of the large spacings involved. If resistance-grounded, then the situation is different, but three-phase is still possible and will still be the worst case.

 

[krisys]

budhiman (Electrical),

Arc flash study is identifying the worst case energy, which is for three phase fault and the trip settings are corresponding to that current. Now instead of three phase fault if there is a ground fault, the arc energy for the same fault current would be reduced to 1/sqrt3. So when the protective device operates, the arc energy would be restricted to much lower level. Hence there should not be any concern.