Do your homework when buying used solar panel

[Prosolar]

Well folks, I thought I'd pass along some information.

Last summer I purchased 12kW of used Trina 250W panels. The price, even with shipping, was $.28 cents a watt. These were 60 cell, mono crystal in a black frame.

Life is good! Right?

I completely re-did my 48V system using these panels. I was, and still am, extremely happy with the results!

However! When I went to install these in my soon to be done grid tie, the electrical inspector rejected them even without seeing the panels. WHY? Because the company I purchase them from removed the labels that showed the panels was UL approved. I knew this ahead of time and didn't give it much thought. However, Without the label, he would't okay my system. So I had to purchase some new panels to get the approval.

Once I did that, and the system was passed, I installed some of the used panels.

That's when I had troubles. The solaredge inverter would fault with a 'isolation error' and only when I disconnected the used panels did the inverter start up.

long story... all the used Trina modules have leakage from the wires to the frame, and thus are unusable in my gird tie system.

Solaredge trouble shooting wants you to use a megger to read the resistance between the positive and negative leads connected TOGETHER and to the frame. A good panel should have at least 80MΩ of resistance. The new panels read >4000Ω.

The trina panels? 100kΩ 1000kΩ and 438KΩ.... all panels failed this test. Now if you wait a few dry sunny days, the panels dry out and the inverter will operate.

Today, right now my solaredge inverter is telling me the isolation value is 472kΩ. The minimum is 600kΩ. Normal is 11,000KΩ

On the battery array, if you disconnect the positive lead and connect it to your meter, then place the negative lead to the frame, I can measure 28V!
My fluke 83 meter has such a high input impedance that it can read that voltage. There isn't enough current flow to trip the GFI inside the Midnite Classic charge controller.

So? Bottom line. Sometimes the sweetness of a good deal will leave a bitter taste in your mouth.

I'm mike

 

[MisterSandals]

Now if you wait a few dry sunny days, the panels dry out and the inverter will operate.

Is there something about water that was left out of the story?

 

[Prosolar]

Is there something about water that was left out of the story?

The problem seems to get worst on humid days or days when it rains. Sometimes the wet weather affects them, sometimes it won't.

 

[MisterSandals]

I've never needed to do this but I would suspect that you would see some corrosion or water signs in the distribution box (is that what its called?) or along the traces on the front. I might even remove the frame and glass on one that i thought was bad to have a closer look.

 

[Prosolar]

I've never needed to do this but I would suspect that you would see some corrosion or water signs in the distribution box (is that what its called?) or along the traces on the front. I might even remove the frame and glass on one that i thought was bad to have a closer look.

Trust me! I've looked and looked. I even went to far as to test the bypass diodes thinking maybe one was cracked and leaky. No joy!

It also appears that not all of them have this problem. I wonder if that is why they were taken out of service to begin with?

Other than the issue with the grid tie system, they work GREAT in the battery array.

 

[MisterSandals]

Other than the issue with the grid tie system, they work GREAT in the battery array.

Curious... why the difference? I am a cause and effect kind of guy and NEED to know when crap like this happens, just to learn.

 

[Prosolar]

Curious... why the difference? I am a cause and effect kind of guy and NEED to know when crap like this happens, just to learn.

I suspect, and I'm only guessing, but the Solaredge inverter is being very picky. At start up it measures the isolation value on the PV input strings. If it sees something it doesn't like, well it simply refuses to start up. I've been using a little trick. I added a fused disconnect between the inverter and the used panels. When the inverter is shut off in the morning because of isolation fault, I disconnect the used panels and that allows the inverter to start then an hour later I simply turn on the used panel array and the inverter doesn't have an issue. At night, I'll see the isolation resistance drop into the low 1000kΩ range, still high enough to start the inverter the next day.

On the battery array, you can read voltage from the panels to ground. (they do NOT have the negative lead or positive lead grounded. This system floats, just like the grid tie arrays do.)

My fluke meters have an unbelievable high input impedance and therefore don't load the circuit down. I suspect a old Simpson 260 with it 20K per Ω resistance wouldn't see the voltage simply because the meter would load the circuit.

I wonder if there are micro cracks that allow humidity into the cells? I don't know.

 

[MisterSandals]

I wonder if there are micro cracks that allow humidity into the cells? I don't know.

I think you would need to isolate individual cell(s) that RELIABLY exhibit this behavior. That in itself might be tricky. But narrowing it down is pretty much a requirement to debugging your problem.

That's a nice workaround and maybe its sufficient for your use.

 

[Prosolar]

I think you would need to isolate individual cell(s) that RELIABLY exhibit this behavior. That in itself might be tricky. But narrowing it down is pretty much a requirement to debugging your problem.

That's a nice workaround and maybe its sufficient for your use.

It's a pain in the ass!

The plan is to remove those 12 panels from the gird tie array and use them in the battery array.

then purchase 12 brand new panels to go into the gird tie.

I spent hours one day megging all the panels. It would be way too much labor to disassemble the rack to move panels around.

 

[MisterSandals]

It's a pain in the ass!

Yea, most of this is!

I spent hours one day megging all the panels. It would be way too much labor to disassemble the rack to move panels around.

Will your finicky SCC take input from a single panel? If you could run a wire out to individual panel(s) or even take your SCC and a small battery out to the panels it might be less of a PITA.

Just a thought. I can tell you've spent a LOT of time on this. Just trying to offer some simple things in case you have more time to waste.