Bi polar array and inverter troubleshooting

[electro7]

Hi,

We are troubleshooting an Advanced Energy 500NX inverter. It has a PV positive hot, PV positive neutral, PV negative hot and PV negative neutral- Bi polar array. The inverter is throwing an intermittent fault. We thought originally it was a ground fault in an array since one of the ground fault fuses in the inverter was blown. I was told to measure voltage with the ground fault fuses out to float the array from all four conductors in free air to ground. There are 5 "zones" for each "side" of the bi polar arrays. So there were actually 10 conductors I took readings on, only the PV positive side. My readings were around 60 volts to ground on the neutrals and 320 volts to ground from the hots to ground. I was told they should be equal. A co-worker went out today because another fault was thrown but neither ground fault fuse was blown. He tested by turning the inverter off, pulling the ground fault fuses, turning off all the combiner boxes associated with the inverter and measuring voltage to ground from the hots to ground and the neutrals to ground at the combiner boxes. He found the same thing 60V to ground on grounded, neutral conductors and 170V from ungrounded, hots to ground and bleeds off. Advanced energy is thinking there is a ground fault. I am thinking there is a malfunction in the inverter. The files downloaded from the inverter show array voltage at 750 volts then when the fault is thrown it shows a spike up into the 950 volts for the array. Any thoughts on what it could be?

 

[jaggedben]

If you have real voltage between conductors and ground with the ground-fault fuses removed, then you probably have a ground fault. You say it 'bleeds off' and I'm not sure if you're talking about a phantom voltage that's an artifact of (I believe) capacative effects between the frame and cells. But if you're getting a sustained reading of a real voltage then it sounds to me like you still have a ground fault.

When testing for a faulted string in systems with grounded conductors, it's essential to isolate all the strings from each other on both sides. That is, you can't just open disconnects and/or fuse holders that only open the ungrounded conductors, if all the grounded string conductors are still tied to a common bus. You're going to still get readings to ground on all strings because of they way they are all connected to each other on the neutral/grounded bus. Ideally all the strings have MC4 connectors at the combiner, and you can unplug all those and test each string individually at the connectors for a fault. If you do that, and there's a fault on only one string, you'll see readings to ground only on that string and none of the others. It's not clear to me from your description that you've been able to do this. But I would not consider a ground fault to be ruled out if you haven't.

Hope that helps.

 

[electro7]

We did find it. It was a loose wire in a surge protector device in a combiner box.

So I was trying to grasp it with the bi-polar array and the negative conductor grounded on one half of the inverter and ungrounded on the other half of the inverter and same with the positives. And them sharing a common bus and taking voltage readings to ground to find if there was a ground fault.

I see how you have to isolate the strings at the combiners, or "zones" in the inverter recombiner to take readings to ground because the ground fault fuses bond the neutral (grounded) conductor to ground at the inverter. What the AE tech support guy told me was the voltage to ground from each combiner should be a quater of the system voltage, and bleed off. So if the system voltage is 800 volts then from each of the PV positive hot, PV positive neutral, PV negative hot and PV negative neutral to ground should start at 200 volts and bleed off to 0 volts when measured from conductor to ground.

When I pulled all of the conductors that were going to each combiner into free air I measured 60 volts from the neutrals to ground and it would bleed off to 0 volts. And the ungrounded conductors were around 320 volts to ground and would bleed off to 0 volts.

Anyways, I think that loose wire is what was causing it and that it was intermittent. Thanks for the help again Jaggedben!

 

[jaggedben]

:thumbsup:

Thanks for the update. So was the fault an arc-fault instead of a ground fault?

 

[electro7]

I think it was, yes. The first time it happened it blew the ground fault fuse in the inverter. The other times it did not. It would throw the inverter into an "array balance" fault. Which I guess says the array is imbalanced?? On the inverter logs it said "z seq" under the fault.