What is a dangerous cell temperature?

[Paco Solar]

I recently took an Infra-Red photo of a PV module, on a commercial roof, with a few cells that are overheating.
The IR photo shows the cell temperature to be about 186* F, which is about 80*F above the adjacent cells.

My question: Is this a dangerous cell temperature such that this system should immediately be shut down,
or is this in the category of a clear problem that should be scheduled for replacement.


Thank you,
Paco

 

[jaggedben]

That's almost certainly way too high a temp, assuming the IR camera is calibrated correctly, and these are some typical type of silicon modules.

I think it's ultimately a judgement call, but that it clearly needs to be addressed,

If it's only one module showing the issue, and assuming this system has numerous strings, I would be looking for a way to disconnect that string from the system before leaving the site. i.e. shut down the system, disconnect the string, turn the system back on. Then contact manufacturer and/or schedule module for replacement later. If it's numerous modules all over the system, call the boss. Also if it's entire cells vs small hot spots within a cell might affect my opinion. Obviously turning the system off is the safest course of action but what is prudent with regards to money is a consideration depending on how severe and widespread the problem is. If this is a large system you probably don't want to be responsible for shutting down the whole thing down if you can isolate a small problem.

 

[Hv&Lv]

I recently took an Infra-Red photo of a PV module, on a commercial roof, with a few cells that are overheating.
The IR photo shows the cell temperature to be about 186* F, which is about 80*F above the adjacent cells.

My question: Is this a dangerous cell temperature such that this system should immediately be shut down,
or is this in the category of a clear problem that should be scheduled for replacement.


Thank you,
Paco

Are you sure its not reflectivity?
while the panels are good black bodies the glass tops offer a lot of reflectivity.
take a picture from two or three sides, see if it goes away

 

[ggunn]

Are you sure its not reflectivity?
while the panels are good black bodies the glass tops offer a lot of reflectivity.
take a picture from two or three sides, see if it goes away

Yeah, you might be trying to measure the temperature of the sun.

 

[Paco Solar]

That's almost certainly way too high a temp, assuming the IR camera is calibrated correctly, and these are some typical type of silicon modules.

I think it's ultimately a judgement call, but that it clearly needs to be addressed,

If it's only one module showing the issue, and assuming this system has numerous strings, I would be looking for a way to disconnect that string from the system before leaving the site. i.e. shut down the system, disconnect the string, turn the system back on. Then contact manufacturer and/or schedule module for replacement later. If it's numerous modules all over the system, call the boss. Also if it's entire cells vs small hot spots within a cell might affect my opinion. Obviously turning the system off is the safest course of action but what is prudent with regards to money is a consideration depending on how severe and widespread the problem is. If this is a large system you probably don't want to be responsible for shutting down the whole thing down if you can isolate a small problem.

Thank you for the excellent ideas. It is indeed a typical silicon module, and only a single module out of the full (300kW) system. I will pursue isolating that string for now. thanks again.

 

[Paco Solar]

Yeah, you might be trying to measure the temperature of the sun.

Yes, quite sure. Multiple angles show the same 8 cells glowing "hot" in the IR photo. Thanks for the reply. How does one post a photo to this thread?

 

[winnie]

A set of 8 cells in a single module all too hot suggests something like a failed bypass diode.

-Jon

 

[Paco Solar]

A set of 8 cells in a single module all too hot suggests something like a failed bypass diode.

-Jon

Jon,
thanks for the reply. oddly, in this case, the hot cells form an "L" shape going across the short end of the module ( only 5 of 6 cells) and then going down 4 more cells along the long edge of module.

 

[Hv&Lv]

Jon,
thanks for the reply. oddly, in this case, the hot cells form an "L" shape going across the short end of the module ( only 5 of 6 cells) and then going down 4 more cells along the long edge of module.
View attachment 2562032 View attachment 2562033

I’m going to assume the panels are clean.
No poop or other things on them..

Glass cracked?

Would love to see a picture from the right shooting left

 

[Paco Solar]

Photos of the same module... slightly different angle. I don't have a right to left shot. Glass is not cracked, but Module is NOT clean (notice the residue that generally follows the hot cells) the other shot is from below the module showing stark contrast of the "hot" cell.


Could long term "soiling" of those cells cause the diodes to fail?

 

[wwhitney]

As a non-expert, I see 7 (or maybe 8) cells that seem to be generating heat, with the rest showing elevated temperatures due to conduction. The panel is an array of 6x12 cells, and if I understand correctly, there should be one bypass diode for each 2x12 subarray. Since there are heater cells in each of the 3 2x12 subarrays, that would mean that all 3 bypass diodes have failed? Along with 7 (or 8) damaged or failed cells?

Cheers, Wayne

 

[Hv&Lv]

It almost looks like the soiling is causing the heating.
I believe I would clean and reshoot


“Hot spots can origin, if one solar cell, or just a part of it, produces less carrier compared to the other cells connected in series. This may occur due to partially shading, dirt on the module (leaf, bird drop) or cell mismatches. The less producing part is only able to pass current corresponding to its own amount of carrier. Additional carrier, produced in the other cells, accumulate at the cell edges, which leads to a reversed bias of the affected cell. Thus, it works like a resistor and the voltage drop is transferred into heat.”

Handbook of photovoltaic science and engineering

 

[wwhitney]

It almost looks like the soiling is causing the heating.

That is a good point. In the first IR photo in the thread, it looks like the (presumably unshaded) roof next to the panel in question is about as hot as the 7-8 exceptional cells on the panel. So light absorbing soil on the face of the cells could plausibly cause those cells to heat up to a similar temperature as the roof.

Cheers, Wayne

 

[winnie]

Soiling is even worse than 'absorbing sun and heating the cell up light the surroundings'.

Hv&Lv's quote describes the physics of the problem. A solar cell generates a quantity of charge carriers that depend upon how much sunlight is hitting the cell. These charge carriers set the cell short circuit current.

The cell operating current is slightly less than the short circuit current; you can think of the cell as a current source in much the same way that LEDs require constant current drive to operate properly.

If the other cells in the PV array are clean and pushing higher current than the cell short circuit current, then the cell will reverse bias and start consuming electrical power from the array rather than adding to the output. The shaded cell will get much hotter than the rest of the panel.

-Jon

 

[wwhitney]

If the other cells in the PV array are clean and pushing higher current than the cell short circuit current, then the cell will reverse bias and start consuming electrical power from the array rather than adding to the output. The shaded cell will get much hotter than the rest of the panel.

But only if the bypass diodes have failed. Otherwise, the bypass diode dissipates power proportional to the bypass diode forward voltage drop, rather than the PV cell dissipating power proportional to the much higher PV cell reverse voltage drop.

While soiling alone, if it absorbs as much of the sun as the surrounding roof, could cause the soiled area to match the roof temperature. Without any electrical heating, I think.

Cheers, Wayne

 

[jaggedben]

I'd say soiling over time could definitely expose possible deficiencies in manufacturing.

It would be interesting to know which end of the module the junction box is at, or if it has the style with separate jboxes in the middle.

 

[Paco Solar]

As a non-expert, I see 7 (or maybe 8) cells that seem to be generating heat, with the rest showing elevated temperatures due to conduction. The panel is an array of 6x12 cells, and if I understand correctly, there should be one bypass diode for each 2x12 subarray. Since there are heater cells in each of the 3 2x12 subarrays, that would mean that all 3 bypass diodes have failed? Along with 7 (or 8) damaged or failed cells?

Cheers, Wayne

thank you Wayne!

 

[Paco Solar]

But only if the bypass diodes have failed. Otherwise, the bypass diode dissipates power proportional to the bypass diode forward voltage drop, rather than the PV cell dissipating power proportional to the much higher PV cell reverse voltage drop.

While soiling alone, if it absorbs as much of the sun as the surrounding roof, could cause the soiled area to match the roof temperature. Without any electrical heating, I think.

Cheers, Wayne

Thank you, guys, for the feedback!... a very interesting issue to be sure. I will post anymore developments.

 

[winnie]

But only if the bypass diodes have failed. Otherwise, the bypass diode dissipates power proportional to the bypass diode forward voltage drop, rather than the PV cell dissipating power proportional to the much higher PV cell reverse voltage drop.

The bypass diode snips a bit of the total string out, and limits the total voltage of the 'snipped section' to the forward voltage drop of the diode.

But I think the 'snipped section' can still have circulating current and that some cells in the 'snipped section' can generate power which goes to heat other cells in the 'snipped section'. I need to think about this point further, however.

-Jon

 

[wwhitney]

But I think the 'snipped section' can still have circulating current and that some cells in the 'snipped section' can generate power which goes to heat other cells in the 'snipped section'. I need to think about this point further, however.

My understanding is that the substring ('snipped section') will have 24 cells on a 72 cells panel. And that for the individual PV cells, 23 * Voc + bypass diode forward voltage drop < the reverse bias voltage of one cell. So I'm not seeing how that can happen.

Not sure on the numbers, but I think Voc of a single PV cell is less than 1V, and that the reverse bias voltage is 30V or more.

Cheers, Wayne