diy solar

diy solar

Used Solar Panels

‘Cracked vinyl’ I understand and agree they are likely to lead to premature death of the panel, but what exactly are ‘snail trails’?

I have to admit I thought they were merely a cosmetic issue when I saw the description on Santan’s website, but you are making it sound as ‘snail trails’ may actually as big of a reliability concern as cracked vinyl…

Would like to learn more if you’ve got any pointers…
This is one of the better explanation I've seen on the issue of snail trails. Some manu claim this does not hurt performance but IMO, any crack is eventually going to cause an issue. Maybe not today, or tomorrow, but I have to wonder about several years into the panel's lifespan. This article is a bit more in-depth, with bonus pictures. avi snail trails

YMMV
... and shipping a pallet of panels was nearly $400. Envious of those who can just stroll down to the local solar store.
 
Last edited:
I tried testing panels, after I determined some must be underperforming.
Voc/Isc only flagged some bad panels. Vmp/Imp (or rather Vload/I load with a resistive load) was needed to see others.
Using an IR camera showed individual hot cells while an array was operating with an MPPT (Haven't determined if short circuit testing would produce such an image.)


PV panels are relatively cheap in comparison to balance of system, so I pay extra for brands known for quality.
Wow, very cool. So that array was 16 or 17 years old and at least some of the panels with hotspots were poly, correct?

Were they still under warranty and were you able to make a claim with the manufacturer?

From your experience with this, what level of with peak production drop of annual production drop would you advise should trigger analysis for defective panels?

I got peak production out of my Microinverters one day last summer corresponding to ~90% of rated output levels (corrected for temperature) which is at the level that my Microinverters will start to throttle (output was 106% of my Microinverters sustained rating and 101.3% of peak rating (though there was some wind that day that kept everything cooler than ambient temps which I did not know how to correct for).

I’m any case, doesn’t sound like I have anything to worry about with this 5+ year-old array of new mini panels yet, but I’m just curious what level of drop in peak output levels would cause you to break out your fancy thermal imager ;)?
 
These were poly, from Sharp. My other panels of the same vintage are mono, from AstroPower.
Sharp's warranty is 90% for 10 years, 80% for 15 more years. I haven't tried the warranty service yet, but another forum member indicated this model was being replaced by Sharp. So I'm hoping that is entire array replaced with panels that don't have this degradation mechanism.

Previously I thought the Sharps were holding out well, producing near 100%, but AstroPower showed early degradation.
I then ran both (and SunPower) on a transformerless GT PV inverter for a year, which biases half the panels negative and half positive.
Previously all were positive biased (negative ground array.)
Then, I moved all panels to another model inverter which can be jumpered for either positive or negative ground. I positive grounded the SunPower panels, negative grounded the others.

Once I bought a clamp DC ammeter, I checked all strings. One SunPower was open circuit due to wire nut, which I fixed. One Sharp string was half the current of the other, which is why I did that testing. Problem was 5 panels under producing, all in one string. IR images seemed to show some hot cells but not as severe in other panels.

What I'm not sure of is if this degradation was occurring over time, or if it happened during the year of living negative biased. PID (potential induced degradation) has some reversible and some irreversible mechanisms. Maybe that happened.

You can't necessarily count on young panels being in good shape. It varies by brand/model, and environment. Hot and wet is one of the conditions which causes some kinds of damage. Some panels are more susceptible than others.


That was a colleague's fancy $600 thermal imager, so a one-time thing. That, a clamp ammeter, remote data from micro inverters, historical data from SCC or string inverters would high differences from neighbors or previous performance. I finally have a logger connected, but haven't bothered to look at it. I just look at monthly net-metering bills.
 
These were poly, from Sharp. My other panels of the same vintage are mono, from AstroPower.
Sharp's warranty is 90% for 10 years, 80% for 15 more years. I haven't tried the warranty service yet, but another forum member indicated this model was being replaced by Sharp. So I'm hoping that is entire array replaced with panels that don't have this degradation mechanism.
That would be sweet - please let us know how that all shakes out…
Previously I thought the Sharps were holding out well, producing near 100%, but AstroPower showed early degradation.
I then ran both (and SunPower) on a transformerless GT PV inverter for a year, which biases half the panels negative and half positive.
Previously all were positive biased (negative ground array.)
Then, I moved all panels to another model inverter which can be jumpered for either positive or negative ground. I positive grounded the SunPower panels, negative grounded the others.

Once I bought a clamp DC ammeter, I checked all strings. One SunPower was open circuit due to wire nut, which I fixed. [b{One Sharp string was half the current of the other[/b], which is why I did that testing. Problem was 5 panels under producing, all in one string. IR images seemed to show some hot cells but not as severe in other panels.
Yeah 50% output from a single panel or an entire string would definitely get my attention…
What I'm not sure of is if this degradation was occurring over time, or if it happened during the year of living negative biased. PID (potential induced degradation) has some reversible and some irreversible mechanisms. Maybe that happened.

You can't necessarily count on young panels being in good shape. It varies by brand/model, and environment. Hot and wet is one of the conditions which causes some kinds of damage. Some panels are more susceptible than others.


That was a colleague's fancy $600 thermal imager, so a one-time thing. That, a clamp ammeter, remote data from micro inverters, historical data from SCC or string inverters would high differences from neighbors or previous performance. I finally have a logger connected, but haven't bothered to look at it. I just look at monthly net-metering bills.
I’ve got panel-level monitoring on my Microinverter-based array, so I can have a look at the day’s production curve whenever the fancy strikes me and look at panel by panel status through the web-browser link on a laptop anytime I see something fishy…

Caught 3 Microinverter issues in 5 years so far (all based on grid parameters that needed to be better tuned by the supplier - amazing how they can do that remotely from Japan overnight!).

And I’m still getting performance data on my new DC-coupled array but once I have daily production curves during the various seasons trough the year, I think that should allow me to know whenever a panel starts acting up (certainly by 50%, since that would be a 17% loss in total daily production).

I’m all 1S now and have been mulling a move to 5S for my final DC-coupled build in ~18 months.

But if panel reliability issues are markedly reduced at lower string voltage, I may decide to just stick with my 1S full-parallel array (an easy option because I’ve already invested in running 3 6AWG home runs…).
 
'm not sure how having more panels in series changes the voltage across any single panel.
The higher resistance/open circuit across a defective panel when it is at or close to system Voc?
I see your point tho....the damage I referred to would have been caused at a low voltage - less than 1/3 single panel V and the damage missed so got put on the ok used heap since the by pass diodes were possibly still working.
 
Back
Top