Double-Glass PV panels

[sensij]

If JPM wanted to complain double-glass were higher priced, or that certain outfits were hawking bifacial panels Sunpower-style, he should simply have said so.

Please don't take my comments as speaking for J.P.M... I'm just offering my interpretation of what I read in this thread before jumping in.

Also, just to be clear, the JPL paper I linked is not meant to be treated as a definitive answer on panel thermal behavior. It is just a model, and like most, includes some assumptions, omissions, and simplifications. Relative to where the discussion had been headed before J.P.M. jumped back in, I thought it was a worthwhile contribution to show how an energy balance might be set up and solved. Its great that JohanVanR implemented it programmatically so quickly, but without more information on what he set for the material properties, and how well those choices reflect actual panels of interest, I would be cautious about running too far with those results. For a better handle on the big picture, the Duffie and Beckman reference was a good one.

With regard to your IR measurements, it is a good idea to consider the difference in panel sizes (if any) and what the panels are generating (if any) to get a better sense of what those temps are telling you about what's going on in the panel. Again, think in terms of total energy balance. Also... those IR devices are sensitive to emissivity... very good for comparing one surface to itself, but not as good at comparing the temps of different surfaces without further calibration. Comparing the reading between two different types of panels *might* be ok, but just keep in mind that it might not be, too.

Once you get connected to the SolarEdge portal, I think you will have some really nice data to add to this discussion.

Let's see what happens now that a mainstream-ish vendor, SolarWorld, is starting to make and sell 'em. Maybe it'll become just another standard part of high-efficiency panels.

The SolarWorld Protect model has been out for a couple years; they elected to keep the aluminum frame but replace the backsheet with glass. It is considered a "premium" panel, and priced as such... nice to have on the roof but not really competitive (in my opinion) with the current generation of LG panels out there.

(But crazy marketing will always be with us, alas; it's not unique to bifacial panels.)

Crazy marketing may always be with us, but I'm glad the forum has a few members who are willing to cut through the BS, in whatever form it takes. (I'm also glad the forum has members who are willing to challenge convention and seek new ideas and sources of information, although the interactions are sometimes hard to watch).

[DanKegel]

With regard to your IR measurements, it is a good idea to consider the difference in panel sizes (if any) and what the panels are generating (if any) to get a better sense of what those temps are telling you about what's going on in the panel. Again, think in terms of total energy balance. Also... those IR devices are sensitive to emissivity... very good for comparing one surface to itself, but not as good at comparing the temps of different surfaces without further calibration. Comparing the reading between two different types of panels *might* be ok, but just keep in mind that it might not be, too.

Yup. FWIW, the panels seem about the same size, and the measurements were taken with no electricity being generated. The cells are quite possibly different enough to change emissivity, I think the GxB300's have an amorphous silicon layer, but the LG310's don't? So the IR measurements may be pants. Maybe I should bond a temperature sensor to each panel; I have a handful I could calibrate and use.

The SolarWorld Protect model has been out for a couple years; they elected to keep the aluminum frame but replace the backsheet with glass. It is considered a "premium" panel, and priced as such... nice to have on the roof but not really competitive (in my opinion) with the current generation of LG panels out there.

True, but I was thinking of the bifacial module they're just now introducing: http://www.solarworld.de/en/products...nmodule-bisun/
Don't know yet whether they're frameless, but I imagine they'd be just like the Protect, with different cells.
I think this one *is* intended to compete with current LG panels, but I don't think pricing has been announced, so I could be wrong.

[JohanVanR]

The Italians in [1] experimentally show* that the temperature** of their glass-glass modules increases 20% more than that of glass-tedlar a.f.o. irradiation.

* after some filtering and fitting
** measured at the module back surface, not at the cell

[1] dx.doi.org/10.1016/j.egypro.2014.02.148

[J.P.M.]

The Italians in [1] experimentally show* that the temperature** of their glass-glass modules increases 20% more than that of glass-tedlar a.f.o. irradiation.

* after some filtering and fitting
** measured at the module back surface, not at the cell

[1] dx.doi.org/10.1016/j.egypro.2014.02.148

If I an correctly understanding what you are writing, I believe that reported result is close or similar to what I've been writing. A literature search can save a lot of time when done first.

[DanKegel]

That's an interesting paper, thanks for posting the link.
It even mentions correcting for the emissivity of the panel type when measuring temperatures using IR, and shows how they filtered out cloudy days from the data.
It also mentions that frameless modules appear to be cooler away from the center,
and that (if I read it right) black backsheets make more of a difference than a glass back panel.

Anyone who wants to use that paper to draw conclusions should read the whole thing.

[sensij]

The Italians in [1] experimentally show* that the temperature** of their glass-glass modules increases 20% more than that of glass-tedlar a.f.o. irradiation.

* after some filtering and fitting
** measured at the module back surface, not at the cell

[1] dx.doi.org/10.1016/j.egypro.2014.02.148

That is not really what they show. The comparison that appears to best isolate the effect of glass vs plastic is between numbers 3 and 4. Between those two, the biggest difference in thermal behavior is that the glass has a higher gradient between the center and the edges; the framed panels have a flatter gradient. Those two panels flip positions in the temperature list depending on whether you are looking at the peak or average temperature, and the differences are very small, perhaps insignificantly so, in either case. How the temperature *gradient* effects panel efficiency is not clearly communicated in the temperature coefficient of the panel, and is not among the results they present here.

Showing that CIGS panels have different thermal characteristics than m-Si and concluding the glass construction is responsible is not sound logic.
result.JPG

[J.P.M.]

That's an interesting paper, thanks for posting the link.
It even mentions correcting for the emissivity of the panel type when measuring temperatures using IR, and shows how they filtered out cloudy days from the data.
It also mentions that frameless modules appear to be cooler away from the center,
and that (if I read it right) black backsheets make more of a difference than a glass back panel.

Anyone who wants to use that paper to draw conclusions should read the whole thing.

Not knocking it and it seems interesting for further reading. I'd suggest keeping in mind that for findings from the type of research such as reported in the reference is usually considered most applicable and reliable to the particular situation described, less so, but perhaps more or less applicable to other scenarios. One experiment under one set of conditions may be representative of other situations, but it does not constitute fact or law without further confirmation and review.

I've seen that paper and have used it and others and other information based on training and experience in trying to come up with a representative set of temperatures for my array through temperature readings and experiment for close to the last 2 years. I've lost count of the number of times I've been under my array measuring temps. I've described the procedure in prior posts.

[DanKegel]

Thanks, Sensij, I missed that 3 and 4 differed in framedness.