Hi,

I am looking for information regarding what comprises a "DC arc fault detection device" for DC solar panels connected at the main switch for industrial plants. I have been told the incident energy will be high as the existing protections do not trip on overload. I am interested in recommendations or suggestions for completing an arc fault analysis using SKM to correctly include the DC solar system.

I'm not sure what the person talking to you meant by high. I have not really seen any DC system that had a cal/cm2 over 4, but in all our DC systems we use fuses with very short interrupt ratings. As a matter of fact, the DC arc Flash ratings for the 2 MW inverter were ten times lower than the AC Arc Flash ratings where the Inverter tied back to the grid. I could not get the cal/cm2 of the DC portion over 4 cal/cm2 without pushing the tA up to 10 times what the fuse interrupt rating is.

I have not used the software for my calculations (I have an Excel spreadsheet with all the equations programmed into it) so I really couldn't advise you on SKM.

- See more at: http://arcflashforum.brainfiller.com/threads/40-cal-suit-not-to-be-used-all-day.3173/#post-14715

Not even sure how SKM does it. I always spreadsheet it since the DC equipment is such a small part of the overall system. SKM has a tendency to make certan assumptions like reverting to Lee calculation on the AC side if you don't pay attention to the notes on the output.

You can use a fiber optic detector for ANY arc. It doesn't matter if it is DC or not. The high di/dt (HFCT's) obviously won't work though. I like using SEL relays in the first place and they have relatively inexpensive arc detectors available.

I agree with Larry though....I have yet to find any properly protected system where you can even reach above a few cal/cm^2 in a DC system. Make absolutely sure that you are properly modelling your resistance. Usually resistance in a DC system is what limits the short circuit current to relatively small values. Then one needs to decide whether the IEEE 1584 "2 second" assumption applies (it's a guess but it's anyone's guess and hey, it's a standard). If you do these two things then you won't get more than a few cal's.

People freak out when DC equipment frequently operates into thousands of amps and think that AC=DC when it comes to arcing. It's not. It's a totally different world. For one thing there's no such thing as zero crossing in DC.

If you are looking for DC AFCI, then I would recommend Senseta for the detection of DC arc faults in PV systems, this device can then be tied into a shunt trip switch, that will turn the PV system off, there is also a system from Mersen known as Green-Eye, and Green - Brain that can be tied into the DC AFCI that will shut the PV system down at the individual PV panel, it will also control multiple panel for complete shut down.

If the panel is the only source, a fault at high noon on a sunny day might approach the panel's rated output. If storage batteries exist, they are what will determine the IE. You would consider a fault at the battery terminals using the 2s rule.