Solar panel efficiency

[s/v Jedi]

Originally Posted by Dsanduril

A 100W panel is a 100W panel at STC (1000 W/m2, 25C). Check two panels at 45C (typical NOCT) and you may find that one loses .45%/C and the other loses .35%/C. On a 100W panel that means a difference of 2W at normal operating temperatures. That means that at normal operating conditions one of those panels is a 91W panel and the other is a 93W panel. Nitpicking? Yes, but a better panel, with smaller losses, will produce more power overall, even if both are rated 100W at STC.

You say that there is a 2% variation in efficiency among panels on just temperature correction? I did not know that but would love to see this supported by figures from manufacturers or test labs

[Dsanduril]

Originally Posted by s/v Jedi

You say that there is a 2% variation in efficiency among panels on just temperature correction? I did not know that but would love to see this supported by figures from manufacturers or test labs

Jedi - I'm posting this in response to your question, and know you know the details, but laying some of them out for others who might not.

It's the temperature coefficient of power production. Power output goes down as cell temperature goes up and panels are rated at STC, which are unrealistic operating conditions. So a 100W isn't really a 100W across the board. Here are some samples (of well known, respected, widely used brands). As my example I took 250W panels so it would be easy to make comparisons.

Sunpower X20-250-BLK -0.30%/C (Sunpower X20 Datasheet)

Sunpower SPR-245NE-WHT-D -0.38%/C (Sunpower E19 Datasheet)
(an older 245W model - couldn't find the actual datasheet for the 250)

Solarworld SW250 -0.45%/C (Solarworld SW250 Datasheet)

Kyocera KD250GX-LFB2 -0.46%/C (Kyocera KD250 Datasheet)

Just in those three manufacturers there's a range of 0.16%/C. Given that panels generally operate about 20C above the STC rating temperature (as evidenced by the NOCT standard), that means a spread of 3.2% in panel outputs under real world conditions, all for the same nominal power rating.

Another way to look at this is the NOCT/PTC power ratings of the panels. These are ratings under more typical operating conditions (including 20% less irradiance for NOCT, so an automatic 20% hit to power rating).

Sunpower X20-250 = 188W (NOCT) / ???(PTC)

Sunpower SPR-250-BLK = ??? (NOCT) /231.7W (PTC)

Solarworld SW250 = 183.3W (NOCT) / 222.3W (PTC)

Kyocera KD250 = 180W (NOCT) / 223.7 (PTC)

All of the above listed panels are "250W" panels, but if you dig around in datasheets (the information is not always easy to find, even at aggregating websites) you can see that using NOCT rating conditions the panels range from 180 to 188W. That's a 4.3% difference for panels of the same nominal rating. These are all values published by the manufacturers, reputable ones can back them up with independent test results, other maybe not.

One of the things to note in all of this is that NOCT is the measured normal operating cell temperature, that is, at 20C ambient and 800W/m2 irradiance, what is the steady-state operating temperature? If you look at the Sunpower panels it is 41-43C, while for the others it is 45-46C. So the Sunpower panels stay cooler, giving them a better power output. There are numerous factors; more efficient cells convert more sunlight to electricity and less to heat, and different panel designs can be more or less efficient in passive cooling (thickness of silicon, glass, coatings, frame,.....).

Ain't it grand that an apple isn't quite really an apple?

[noelex 77]

Originally Posted by Dsanduril

Another way to look at this is the NOCT/PTC power ratings of the panels. These are ratings under more typical operating conditions (including 20% less irradiance for NOCT, so an automatic 20% hit to power rating).

Spot on
The NOCT rating is always a much better comparison than the STC output.

If the manufacturer does not quote NOCT (many do not). You can at least calculate the real output at a realistic cell temperature from the thermal coefficients which are always listed.
This is a bit time consuming to do especially as the the units used are not always the same (% per C, or V per C), but its worth doing if you debating between a couple panels.

It is also worth looking at number of bypass diodes (more is better) this will impact the very important shade performance.

Finally look at the allowed tolerance. Some manufacturers specify the real power output as 0 to -5%, or +5 to -5% others +5 to 0%. Assuming the mean is in the middle (not necessarily a valid assumption) this is 5% difference in average performance.

Manufacturers can be very sneaky with the truth. Kyocera proudly featured "tighter specifications assuring consistent performance" as one of the advantages of the KD series to the older KC series. In practice they went (from memory) from +5 to -5% with the KC series to +3 to -5% for the newer KD series

[CAELESTIS]

Correct me if I am wrong, but I think the Sanyo beats them all on the heat issue.http://www.solarpanelstore.com/pdf/hit_240.pdf

[Dsanduril]

Originally Posted by CAELESTIS

Correct me if I am wrong, but I think the Sanyo beats them all on the heat issue. http://www.solarpanelstore.com/pdf/hit_240.pdf

Interesting. The temperature coefficient on the datasheet you linked is -0.30%/C, which matches the Sunpower X20 rating and is better than about 98% of the other manufacturers out there.

However, the NOCT is listed as 48.3C, whereas the X20 is listed 41.5-43.0 (depending on exact model) and most others are 44-46. That means about 6 degree operating temperature difference from the X20, which would result in a 1.8% difference in output at typical operating conditions. And yet the Sanyo datasheet argues that it has the highest PTC ratings in class, which doesn't match with the NOCT data.

As Noelex said, have to be careful with what spin manufacturer's put on data, but either way those are pretty nice panels (at least on paper).

[CAELESTIS]

Originally Posted by Dsanduril

Interesting. The temperature coefficient on the datasheet you linked is -0.30%/C, which matches the Sunpower X20 rating and is better than about 98% of the other manufacturers out there.

However, the NOCT is listed as 48.3C, whereas the X20 is listed 41.5-43.0 (depending on exact model) and most others are 44-46. That means about 6 degree operating temperature difference from the X20, which would result in a 1.8% difference in output at typical operating conditions. And yet the Sanyo datasheet argues that it has the highest PTC ratings in class, which doesn't match with the NOCT data.

As Noelex said, have to be careful with what spin manufacturer's put on data, but either way those are pretty nice panels (at least on paper).

Agreed (there are many good products out there). The frustration is it is nearly impossible to buy Sunpower. My neighbor put 20 on his roof, and I offered the installers cash to sell me one or two panels. They did not have them available.. Plan B was Sanyo/Panasonic.