DC Arc Flash calculation

[Magic Gorge]

When calculating the arc flash hazard at a DC switchboard, being fed from an AC switchboard via rectifier circuit, the SKM software does not recognize the upstream AC device, so the arcing time at the DC swbd runs to the max 2 seconds. SKM support staff seem to agree with this, as the AC and DC systems are calculated separately. However, would the clearing time of the upstream AC feeder breaker relay have any bearing on the DC system?

 

[steve66]

If there are capacitors or inductors or energy storage devices in the circuit, it seems like the arc flash could continue after the upstream AC breaker is open.

 

[zbang]

If there are capacitors or inductors or energy storage devices in the circuit, it seems like the arc flash could continue after the upstream AC breaker is open.

(I really hate writing a response then fumble-fingering the keyboard and losing it all.)

Surely, but for how long? I'd expect any L's and C's to discharge rather quickly (<1 second?) unless it's a rather large bank of caps. I'd also expect the voltage to drop below that to maintain the arc fairly fast. OTOH, if the cap's bank is current-limited, that could form a ballast and maintain the arc without also destroying the caps, which is also a possibility. Still probably happens fairly quickly. (Also consider an energy storage flywheel, if present.)

Lots of "expect"s and "probably"s :happyyes:

A stack of batteries is the bigger problem- depending on the discharge rates and internal & wiring impedance, you could have a nice long-term event. That should be easier to model.

 

[wbdvt]

did you look to see the specs on the rectifier circuit? The electronics may have a max short circuit rating and time before the internal protective circuits operate to protect the electronics. you may want to use those values for the arc flash calc.

 

[Besoeker3]

I had a similar question asked of me recently about fault current contribution from a fairly large variable speed drive. Fairly large in this case is 6,600 kW. The DC link has 384 2,200uF capacitors. Enough to make quite a bit of a mess you might think. It's a slip energy recovery drive so with a mains commutated inverter. Commutation failure would result in a capacitor discharge across the DC, not fault current into the 11kV supply. Interesting case study.
And pays some of my bills.

 

[Sahib]

To the fault current, major contributor is upstream AC source, other parallel sources like motors, capacitors also contribute and so to be factored suitably into the magnitude of the fault current for arc flash calculation. A DC side protective device may operate faster than upstream AC side protective device and so DC fault current magnitude may be considered for arc flash calculation.