This week’s question is another in the line of questions a few at an “unspecified committee� were pondering.

Have you/company/client ever had a DC arc flash? Was there an injury? (something larger than a few sparks)
Yes – No Injury
Yes – Injury
No/Doesn’t apply

As always, stories and comments are welcome and encouraged!

GE MD824 DC motor flash over about 12-24" from an employee. Those have a 6 kA stall current off the generator (Ward Leonard loop). There is overcurrent protection but even after removing excitation off the generator there is plenty of residual flux to keep it going for a few seconds. Very common in mining excavators and old elevators. No arc flash PPE. No injury.

I've long been arguing that 70E tables are all but useless for DC because they are not representative. DC systems industrially are controls (too low voltage for sustained arc), batteries, UPS, drives, motors, and electrostatic precipitation. Of those the Duke test takes most but not all DC battery systems off the table. In general 70E is way off here because the spacing is enormous to avoid problems with flashovers and tracking from salts and leaks, and cell voltages are 1-2 V per cell so its only at the ends of strings where voltages even approach power arc potential. Due to the inherent difficulty of breaking DC arcs those areas have typical 600 VAC (1" or more) spacing. Common utility and UPS voltages are 48, 125, 250, and 480. Really only the 250 and 480 should be a concern and those are rare.

In DC busses in drives and UPS (not battery bus) we have to look at two factors. The semiconductors are well not conductors and very short lived (much less than 1 second) in an arcing fault. Second there is some capacitor storage. Enough to be lethal for shock but not arc flash for even large drives where I've run the calculation.

Electrostatic precipitators use low currents (milliamps) with very high voltages (25-100 kVDC). They often arc under normal operation but the arcs are contained under lock and key with no reason to get that close. Hot sticks are used to apply grounding since gloves are 1pp% useless. The highest class glove, class 4, goes to barely 50 kVDC. Even then it's either basically a big capacitor or a milliamps, high voltage source. Tough to model. The protection system uses a resistor bridge to measure output voltage and a CT on the low voltage side of the rectifier-transformer to measure current. Again scary arcs but I've never calculated over 1.2 cal/cm2 with Lee.

I would suggest adjusting the 70E tables for DC to be representative of actual application ranges, not contrived theoretical crap based on working the equations backwards. If nothing else the table should have lower limits to where no PPE is needed, not just upper bounding at a silly value.

At a previous employer we had an electrician inadvertently short a freshly equalized 250V lead acid battery bank via a connecting cable he was holding in his bare hand (had taken off his glove). The result was blistering on parts of the hand, but stopping at the sleeve cuff (was wearing AR long sleeve shirt). The connector on the end of the cable and a portion of the battery post were melted.

I voted No, because we've never had such an event, although I feel we have the possibility. By my calculation, our DC arc-flash boundary is greater than 10 feet.

All the best, Doug