GFP and Arc Fault Protection with multiple inverters

[Kent0]

When using the Classic's built-in GFP and Arc Fault Protection systems in an installation with multiple inverters, do the inverters have to be networked together so that all of the PV arrays are disabled?  What about operation with older charge controllers - is that possible?

[boB]

"When using the Classic's built-in GFP and Arc Fault Protection systems in an installation with multiple inverters, do the inverters have to be networked together so that all of the PV arrays are disabled?  What about operation with older charge controllers - is that possible?"


Hi Kent and a merry new year to you and all of  eastern Oregon  !!

As far as we know, the Classic should not have a problem with multiple inverters and charge controllers just as they are.
If you already have a  GFP for your PV, then you would probably want to disable the GFP on the Classic
by removing the jumper that connects Ground to negative.  (otherwise it would be a double-negative !  [har ! har !] ) 

As for the arc fault detection, it only looks at the array  the particular Classic is connected to, so
other units in the system "shouldn't"   have a problem.  If, for some reason the rest of the system
wants to play havock with the Classic's AF detection, then you can adjust its sensitivity, or, if necessary,
turn it off altogether.  So far it's behaving though.

boB

[Robin]

Hi Kent. Multiple inverters should have no affect on the DC-GFP circuit. A DC-GFP is connected up to the PV inputs, right? If there is one or ten inverters in the system, that won't affect the PV circuit. Since the DC-GFP is looking for current flowing between battery/PV negative and earth ground, a fault in the inverter circuit might trip the DC-GFP. It will not prevent a fault from continuing though. When a DC-GFP breaker assembly trips, it disconnects the PV positive circuit. Any issue that develops in the invereter circuit won't be helped by turning off the PV plus. The Classic DC-GFP works a bit different. It detects a fault between battery/PV negative and earth ground just like the breaker DC-GFP system. The difference with the Classic is that it simply turns off when a ground fault is detected. This is different than disconnecting the PV plus circuit. This trick of turning off was first pioneered by Xantrex. It is a much less expensive solution for the DC-GFP requirement. I am personnaly not a big fan of the DC ground fault requirement in the NEC. It does make sense for grid tie systems due to the high voltages involved. Now that we are getting into relatively high voltage for off-grid, I suppose you can make a case for the C-GFP requirement. When you think about these types of faults and compare ground fault vs. arc fault, the AF wins hands down. We have seen very dangerous arc happening at fairly low voltages. The Classic will stop these if they are a series arc fault. (most are series) We are also developing a way to stop a parallel arc fault. Those are much more involved and are not even a requirement in the 2011 NEC. So, multiple inverters are not a problem for the GFP, but multiple PV arrays can be. If you have ten PV arrays in a system and ten DC-GFP's, then it will take ten times the fault current to trip anything. That may not be good enough. If you disable or eliminate all but one DC-GFP, then only the PV array that has the GFP will be protected. An adjacent PV array can have a fault. The fault will possibly trip the one and only DC-GFP. That will not stop the fault though. This gets complicated huh? When you have an arc fault detector, you do not disable any of them. An arc will shut off the offending PV array. Problem solved. We will eventually be able to send out an e-mail telling you there has been an arc event. The Classic now just squaks at you and displays the arc fault message. Voiuce is being added soon. Gee, I wonder what the Classic will say when an arc fault is detected? That should be fun.

[Kent0]

Thanks boB and Robin.  From what you have described, the Classic's arc fault protection is looking for the arc signature on the current from the PV array and only the affected array should be interrupted, so multiple charge controllers presents no difficulty.  And systems that are expanded under the 2011 code should work fine aside older systems that didn't (don't) have arc fault protection.

For ground faults, an external ground fault assembly that interrupts all the PV arrays as well as the ground connection is probably better.  If GFP is active in several charge controllers it may make the GFP protection less sensitive but all of the GFP devices will eventually open.  And the fact that the GFP is in the charge controller opened doesn't mean that the associated PV array is implicated in the ground fault.

Another question: how much fault current can the Classic's GFP interrupt?  Obviously these are intended for battery systems and a ground fault in the battery can produce a large current.

[Robin]

Kent, Your comments are correct.
Your last question regarding how much current we can interrupt is almost irrelevant. We do not interrupt the fault. We merely shut the controller off. That stops some of the ground faults. The resettable fuse inside the Classic interrupts the bat/PV minus to earth ground fault. This will also interrupt a PV+ to ground fault by the way. None of the DC-GFP's on the market are designed to detect and interrupt a fault in the battery circuit. These devices are connected to the PV circuit, but do connect to earth ground and bat/PV minus. I am not sure what the interrupt rating of the PTC is? I will look it up and report back. If the fault was in the battery circuit, that may be an issue. The Classic GFP is not intended to interrupt battery fault currents.

[boB]

Kent, Your comments are correct.
Your last question regarding how much current we can interrupt is almost irrelevant. We do not interrupt the fault. We merely shut the controller off. That stops some of the ground faults. The resettable fuse inside the Classic interrupts the bat/PV minus to earth ground fault. This will also interrupt a PV+ to ground fault by the way. None of the DC-GFP's on the market are designed to detect and interrupt a fault in the battery circuit. These devices are connected to the PV circuit, but do connect to earth ground and bat/PV minus. I am not sure what the interrupt rating of the PTC is? I will look it up and report back. If the fault was in the battery circuit, that may be an issue. The Classic GFP is not intended to interrupt battery fault currents.

I don't think that it will matter where the current is coming from.  The only thing that limits the internal GFP resettable fuse is the voltage across it when it trips.

This fuse has a bit of series resistance in it when closed (less than 1 Ohm though) and won't allow thousands or hundreds of Amps through it instantaneously and there are other current limiting resistances as well.

It trips after about 1/2 to 3/4 Amp and then heats up to lower its series current even more.

That all being said, I will play with these parts again wit batteries and report back.

I know that they work find directly across a 240VAC line which can  have a LOT of current capability.

I think that what happens is that they just get hot real FAST.

boB

[Robin]

Kent, I found boB at the factory today so asked him to run some tests on the PTC device we use. He shorted out a 24V battery bank with the PTC. It did exactly as expected. Made a spark at the connection point and then just sat there very warm to the touch. He then re-ran the test with a 5000 watt 250 volt power supply. Same thing happened. The interrupt capacity of the device must be quite high as neither of these tests had an issue. I think we ran these tests years ago when designing the GFP system. Now we documented it so we won't have to do this again in years to come. The PTC device is called a critical component by ETL and is listed in our listing report as such. The ETL inspector can review the production parts periodically to keep us honest. They don't want us sticking some unknown part into the Classic that may not behave as expected.

[steff]

What if I have 3 charge controllers , serving 3 arrays all connected to teh same battery, only one of which is a Classic and I want to protect everything. Do I have to use mutliple Midnite breaker  DC-GFPs ? And if so do they need separate grounding rods to be safe ? I mean if I have several MNDC mini disconnects each with a DC-GFP, won't they make grounding loops if connected to same ground rod ? So I guess the question I am asking is ,  how many GFPs do I need and do they need separate ground rod each ? I am assuming here that I will ground my batterbox enclosure (MNBD) metal but that I will NOT ground the battery negatives, and that I will disable the Classic's  built-in  GFP via jumper since I have DC-GFP in the MNDC min-disconnects ?   Thanks for your expertise

[steff]

I forgoto to mention, Running 3 separate inverters onlyu one of which will be connected to future AC panel, other ones standalone. None of them support multiple AC sources. Also have a IOTA dls-45 battery charger. so 4 circuits. 3 PV circuits each with combiner,SPD,Diconnect,GFP . Each inverter protected by breaker in MNDC mini disconnect and another breaker for the IOTA power converter. How many DC-GFPs do I need and where  and where are they grounded ? Common grouding rod ? or separate for each or am I allowed only 1 ? I presume if I have only 1 it only protects 1 circuit which sucks. IF I have more than 1, how do I set it up so I don't need 4x the fault to trip them or create grounding loops ? The only way that sort of makes sense to me is if I have separate grounding rods ?