DC Hazards, Arc Faults

Options
zoneblue
zoneblue Solar Expert Posts: 1,220 ✭✭✭✭
I'm starting to do the detailed design for connection, hookup, wiring etc.
Hence I think understanding the various risks inherent in DC electrical systems must be next on my list of things to get right. I understand that there are three enhanced risks, DC Arc faults, electric shock, and fire.

If they werent so jolly expensive here id stump for midnite epanels and combiners. Instead i need to be able to select cheaper local gear based on an understanding of its performance.

Using lots of redundant insulation, fireproof materials, and fuses sized to protect the appropriate cables, good labelling etc i know about. My plan is a broadly a combination of midnite dc breakers, HRC 120V fuses, and blue sea battery disconnects, about $30 ea. A nice steel box to keep everything safe from unsuspecting fingers.

My question is this really: is there any good reading on understanding DC arcs, what causes it and how to prevent it.
Ive tried googling DC arc fault prevention/understanding but all i get is a lot of stuff about welding etc.
There's a couple of youtube videos showing how pv arrays can arc and start roof fires. Scary stuff. An external array disconnect is starting to seem like a good idea. Theres also a photo of a combiner box fire that resulted simply from removing a string fuse.

I get the impression that the risk starts around 20v/3 amps, but that over 120V is likely to be more risky.
1.8kWp CSUN, 10kWh AGM, Midnite Classic 150, Outback VFX3024E,
http://zoneblue.org/cms/page.php?view=off-grid-solar


Comments

  • Shadowcatcher
    Shadowcatcher Solar Expert Posts: 228 ✭✭✭
    Options
    Re: DC Hazards, Arc Faults

    The one explination that I found that makes some sense is An arc-fault occurs when loose or corroded connections make intermittent contact and causes sparking or arcing between the connections. This translates into heat, which will break down the insulation of the wire and can be the trigger for an electrical fire. I use Anderson Power Poles http://www.andersonpower.com
  • rocmor
    rocmor Registered Users Posts: 11
    Options
    Re: DC Hazards, Arc Faults

    The problem with DC circuits is that we tend to impose our AC logic and experiences on to DC situations. Take the simple task of switching currents on and off. The basic principle involved is this: current stops flowing through a circuit when its value reaches zero. Yes, this sounds obvious. Hang on. So, in your run-of-the-mill 120V wall switch, when you flip the lever, contacts within the switch start to part. Current continues to flow through the widening gap (in an arc) until there is a “current zero”. In US households, our alternating current passes through zero 120 times a second. Therefore, the arcing between the parting contacts will last only about 4 ms on average, not long enough to damage the contact material, typically a tungsten-copper mix that can stand the heat. In fairness, I should add that the contact material and the final separation distance between the contacts also play roles in this process. But, in the end, we can build safe, reliable switches with minimal contact separation, for switching AC currents.

    Not so with DC. Take that same wall switch and try to switch current in a 50V DC circuit. Unless the current flow is small (less than what is called the “chop” level of the contacts), the contacts will start to part and an arc will strike, as described above. BUT, as there are no current zeroes, the arc will continue with the switch full open until… well, until something catastrophic happens, like a meltdown or fire.

    In DC switching, the current zeroes must be “forced”: either by proper switch design (contact separation, contact material, arc control), or via an external “commutation” circuit that “injects” a current through the switch of equal amperage but opposite direction compared to the circuit current.

    The key to controlling DC arcing is not to allow it to happen in the first place: use components made by companies that specialize in DC applications (per Shadowcatcher’s link), and use generous spacing and/or physical barriers between DC conductors between which there may be a high voltage drop.

    As for arc-fault breakers, if you google” DC arc fault circuit breakers”, you’ll find plenty of info. Eaton (parent company of Cutler-Hammer which invented the AFCI for household use), has a nice two-page app note.