DC Arc Fault Detection Theory

[yesterlectric]

Hello,

I am wondering where a good article may be discussing the theory behind DC arc fault detection for PV systems. Seems most articles are based on detection for AC circuits.

Trying to have a better understanding for troubleshooting purposes.

 

[ggunn]

Hello,

I am wondering where a good article may be discussing the theory behind DC arc fault detection for PV systems. Seems most articles are based on detection for AC circuits.

Trying to have a better understanding for troubleshooting purposes.

Look for an SMA white paper. They were one of the first to build DC AFD into their inverters.

 

[GoldDigger]

In any PV system, except for any battery contribution, the DC current is strictly limited and known. The arc energy can only be high if the time is long because OCPD is slow to operate with such a small difference between normal current and fault current.
This puts a premium on GF and AF detection instead.

But I get the impression that the OP is more interested in local DC generators rather than inverters and in high power DC distribution systems. For those situations he needs to go all the way back to the generating equipment and the transmission lines instead of just looking at the last transformer in the chain.

 

[ggunn]

My understanding of the way AFD works on the DC side of SMA inverters (from talking to a rep at an SPI show) is that it looks for a characteristic noise signature on the DC lines. And that's all I know about AFD right there.

 

[ron]

Now that large UPS's have been converting into transformerless (input) designs, DC ground fault has become a popular discussion point in the data center industry.

A UPS has an ungrounded battery similar to an ungrounded PV circuit.

This white paper covers a few methods.

http://www.battcon.com/PapersFinal2006/RafterPaper2006.pdf