DC Arc Flash / Maximum Voltage

What is the highest DC voltage on your (client’s) system?

No DC systems
300 Vdc or less
301-1000 Vdc
1001 – 1500 Vdc
Greater than 1500 Vdc

We don't really have DC distribution to speak of except for substation control power and some 480V UPS's with >400VDC battery banks.

Thanks! It's the >400 Vdc that I was fishing for. DC voltages are increasing and thinking ahead to the time we start a DC arc flash standard, I was curious about the present day upper limits of Vdc.

We have electric shovels in the Mine where the DC Bus is at 1800 VDC. The input voltage to the AFE is 900 VAC and the output voltage from the inverter to the AC motors is 1400 VAC.

1800 Vdc? Good to know! Thanks!

Yes Sir, 1800 VDC.

Let's see...

First we have your typical UPS which used to be pretty low but internally these days tend to match battery string voltages typically seen at utilities or based on diode/SCR bridge voltages so I'd say 48, 125, 250, and 480 VDC are becoming more common and those are also true for substation battery strings but the vast majority of substations are either 48 or 125 with the trend being 125 VDC for newer ones. In power generating stations I typically see that for the switchgear and either 125 or 250 or 480 for the motors such as an oil coolant or lubrication pump, particularly for steam valves and turbines. Since they are usually only dealing with one battery at a time, this stuff is generally NOT very well isolated. At the same time the bus gaps are vastly larger even at 125 VDC (typically an inch or more) to where the tested and calculated (table) ranges for the DC hazard are meaningless...basically a lot of time time you can't even achieve a power arc. Plus we have a dearth of information on the conditions necessary for a power arc so it is much easier to set limits on where a stable arc can be achieved (a low voltage/current limit) for arc flash unlike AC.

Then moving on into VFD's, remember that the bus voltage is 1.44 times the nominal line voltage so for your 480 VAC drive, ALL of them are going to be running at 690 VDC and on a 600 VAC system it becomes 864 VDC. Once you get into medium voltage drives, it gets a little crazy because it is possible to do a straight bus voltage conversion but the problem is the insulation won't tolerate it so in general you see multiple H-bridges, neutral point clamping, series IGBT's, and so forth so there are often multiple bus voltages but the general maximum would be again 1.44 times the nominal voltage of 2400, 4160, 7200, 9600, or 13.5 kV. What makes this particularly bad is that with small drives they are packaged to where realistically except right at the bottom of the drive and in say a multiple DC bus arrangement, most guys are never really exposed to the DC bus or if they are, it's strictly to discharge it or to do testing such as measuring for shorted antiparallel diodes (measure DC bus to AC bus with diode check on your meter, done offline). BUT with larger drives once you get up over about 250-350 HP or so, we get away from the prepackaged stuff and the drives become modular designs to where it is common to be doing all the testing with exposed equipment. With medium voltage obviously there is a lot of safeties that would have to be jumpered out to be directly exposed but on low voltage equipment that is definitely not the case.

Then we have electrostatic precipitators which can have enormously high DC bus votlages. They use a specialized transformer called a TR set which is short for transformer-rectifier that usually contains a step up transformer followed by a full wave diode bridge floating in oil along with a resistive divider for measuring what's going on. I haven't run into a lot of them but the highest voltage I've seen in practice on the HV side runs around 50-60 kVDC. Granted the current is in milliamps but it's there.

On older excavators such as the one just mentioned they use MG sets. The nominal voltage is usually 90 VDC which is name plate but the output current at the armature leads at stall is as high as 6000 A but that's at stall so it drops to under 28 VDC so an arc isn't even possible. At the peak power point you're looking closer to around 1500 A at 90 VDC as I mentioned. Modernized versions jack up the excitation votlage up to around 400-600 VDC maximum for faster acceleration response but that's on the field side so there might be a couple hundred amps there at most.

Sync motors fall in line with the excavator stuff. Generally most armatures rarely get over 250 or 500 VDC for fields on sync motors, as do most industrial DC motors. The most common DC voltages are 125, 250, and 500 VDC for name plates but there is a ton of "off name plate" stuff.

So...if my answer was multiple choice, I would have checked them all off.

BUT....even flashovers of big MG sets like in the mining industry are scary but rarely cause serious burns or fatalities and similarly DC shock hazards are vastly different from AC ones. The table in 70E is basically way, way off base particuarly with the DC arc flash hazard table because almost none of the entries reflect reality at all. Either the available fault current is vastly less or the voltage ranges are totally out of line or both because they basically just adopted common industrial AC voltage ranges rather than starting with common industrial DC voltages. We don't have a convenient ANSI standard for DC voltages like we do with AC but that doesn't mean that something of a standard doesn't exist.