Recover cells from glued panel with broken glass?

[Ontheronix]

Hello people,
I bought a 250Wp panel from wich the glass cover is broken, the glass used reminds of safety glass used in car windows, each square inch is broken:


Plan was to replace the glass .. I thought the glass would peel off easily but everything seems to be glued together. 

Has anyone experience in this? The only solution that I can think off is trying to dissolve the glue in alcohol or ether ... of course it may not damage the cells...

This is the panel: http://hjsolar.co.uk/pv-modules/monocrystalline/hjm250m-32


Thanks!

[ThomasJ]

Have you tried to contact the manufacturer for advice? I would assume that the glass might be safety glass ie it has a layer of plastic sandwiched between to layers of tempered glass or it just has a layer of plastic bonded to a single sheet of tempered glass. I can't think of a solvent that will work to dissolve it without potentially damaging the cells. Is there no way to disassemble the frame to get the glass out?

[Ontheronix]

In the sides of the panel, the glass can be removed with some force, so the plastic 'sandwich' won't be the biggest problem I think.
The problem is that the glass is glued to the cells, so removing the glass with force will breake the cells in peaces without any doubt. 

[richhagen]

This issue seems to crop up from time to time. 

Traditional panels made with mono-crystaline or poly-crystaline solar cells are typically made with low iron tempered glass with breaks into small pieces as you noted.  Typically ethylene-Vinyl acetate, (ETA) a type of plastic is layered on top of the glass, then the solar cells, then another layer of the ETA, and then a back layer of a different poly-flouro plastic such as Tedlar to give water and abrasion resistance to the back side.  The whole thing is laminated together. 

The problem is that the EVA is bonded to the glass, which is now broken into many smaller pieces.  Removing the glass is at best a very slow and tedious process by all accounts I have read to date, and usually futile.  For these types of solar panels, the glass is what gives most of the structural strength to the panel, and the embedded solar cells are easily broken within after the glass is shattered.

If the panel still outputs power, then one possible solution would be to place the panel upside down upon a flat and stable work space, and pour resin such as polyester resin on the back side to give the panel some rigidity, and then once that is set up, coating the front side with a thin coat of UV resistant clear epoxy to seal it up as best as possible.  The results would be imperfect by might allow the user to get some useful life and benefit out of the panel. 

If the Panel does not output any power, then likely one of more of the cells in series has cracked clean through, and I am afraid I have no good advice as to digging out the individual cells.  You might try tapping the panel from the back to access the bus bars at various points and measuring voltage to determine where the break is and if there is only one you might be able to simply bypass that cell. 

All in all it is a hefty amount of labor for a rather modest return in my opinion, but best of luck with project.  Rich

[Ontheronix]

Hmm, I see, your suggestion seems indeed the most doable/reliable option richhagen.
Although I'd like to try the following before doing the above:
Wikipedia tells me that EVA can be dissolved in an organic solvent (e.g., dichloromethane), but personnally, I've never heard of 'organic' solvents.
If somebody here has an idea, I'd like to try it on a corner and see if cells don't become damaged by the solvent.
Also I emailed the manufacturer with the question

[Bruce S]

As an "organic" solvent, it'll give you a nice skin burn  .
I remember using this stuff as a plastic weld. Try going to a modeler's shop and ask for it.
As a methyl=== don't breath the fumes  .

My thinking on those panels will be that you'll probably soften the stuff enough, but may very well ruin what's underneath.
Best of Luck!
Bruce S

[Ontheronix]

Hmm it seems agressive indeed.

Just got a reply from the technical desk in the UK:

We've got two here like, that, though we haven't started to take then apart (yet) that's a project for a nice warm day!

 

I think you'll find that the cells are held in place with a self adhesive backing and that the slicon per se isn't stuck to the glass, though I may be wrong. So if you try to remove with a solvent, then the backing will come off as well, - perhaps not the best idea as they help to hold it all in place dimensionally.

 

Assuming the glass has shattered like safety glass, I'd just sit down with a good pair of gloves in a warm environment and pick the pieces off one by one.

 

If you find a better way – let us know!

[Ontheronix]

Hello all,
for now I've decided to leave things like this, and first try to get some output from the thing.
It delivers 60V short-circuited and 50V on PMax, and are divided in 4 groups with bypass diodes.
So I was thinking, why not cut the path between each group, so I can wire the 4 groups in parrallel to load my 12V lead-acid battery?
50V/4 would give me 12,5V/group, minus diode drop of 0,7V and some wire resistance leaves me maybe 11,5V?
My question is: if the panels only load to 11,5V, and not to a 14-ish V, the battery won't be fully charged? Or would it just take longer?



Junction box:


All cells wired in series:


Specs:

[Mary B]

you need at least 15 volts to properly charge a 12 volt battery. Lay the panel on its back and use a clear coat on the front to seal the glass in place. It will also make it a little bit clearer and up the output. Reinforcing the back is a good idea also.

[Ontheronix]

This is a very lowbudget project, I will take further care of the panel when everything is properly tested.
I'd like to wire the cells as direct as possible to the battery, since I was not planning to spend money on a controller or dc-dc converter or such thing, only a diy dump load controller.
Maybe if I can free som glass I can make 3 groups? That would gave me 16,6V in full sun and a 15,5-16 in clouded wheater, wich we have alot here.

[oztules]

You can run three separate groups. You will need to hack into the EVA on the back and isolate the three cell units at where ever the three points are.... not so easy as they are at intermediate points.

You could hack into the 3- column points and get an 10.6 amp @18wv unit with 24 cells left over..... for a 190 watt panel

Not the best outcome but decent non-the less, and completely doable.

You could get a buck converter (mppt) and do better.

For me, the 10.5A hack would be the simplest with the best outcome for free for a 12v charger.

The 16v 15A hack would be technically more difficult, but still be done if you put your mind to it. (divide into three 32cell counts in parallel) and use all the panel.

And yes I have had to do that on a brand new 200 watt panel, feel a varmit doing it, but it did the job very well.... just seal it properly with more EVA or a good silicon .



...................oztules