LG300 real world power output

[sensij]

The value I like to key in on is peak kW / kW installed, on days that are clear, and PVouput makes this number easy to see. Yesterday's peaks look like cloud effect.

Today, Poway's system maxed at 86.7%, mine was at 86.2% using Solaredge as the source (not my revenue grade meter). Makarowski's system was up at 90.0%... for reasons I don't understand, his system is routinely among the best. In most cases, this is down a couple percent from a couple months ago, when it was cooler out and the sun position was different.

[gvl]

Makarowski's system was up at 90.0%... for reasons I don't understand, his system is routinely among the best.

He must be using one of those liquid-nitrogen filled super-conductivity kits for his wires.

[J.P.M.]

FWIW my 280s that are pointed south only reached 235W (highest reported on solaredge out of 14 south-pointed panels - and most of them only reached high 220's).

The high for my five 280s pointed mostly west was 238W.

I wonder if there's a way to calculate when the sun should be almost perfectly normal to the face of the panel...
(Or what day/time it'll be closest to it)
If my house faced solar noon, I think there are tables I could look at to find out what days the sun would be at X degrees.
Since I haven't got much pitch to my roof, and I don't live far enough south, I'm guessing it'll never be exactly "overhead" for my panels, but it'd be interesting to me to know when it'd be the closest.

There is. You will need location of the array, lat./long., and the array tilt and az.

But without knowing the P.O.A. irradiance, or a way to est. wind vector or roof amb. air temp. to thus est. panels temps. and thus some representative average array temp., it won't be of much use.

Most of the time arrays will not produce S.T.C output.

An example of why:

One of two minimum incidence angle dates and times for this year for my array occurred on 04/30/20105 at ~ 1305 hrs. P.D.T. That angle was ~ 0.13 deg., but the sky was not clear so I took no readings.

On another day, 05/02/2015, at 1305hrs. P.D.T., there were no clouds. The min. angle of incidence that day was also at ~ 1305 hrs. That incidence angle was approx. 0.48 deg., assuming and using an atmospheric parallax angle of 0.005 deg. from precip. H2O estimates using dew point temp. - all very approx.

The global horiz. irrad. was measured as 936 W/m^2. The P.O.A irrad. was calc'd as 980 W/m^2. The 6 min. ave. wind vector, measured and recorded at 1 min. intervals was 1.7 m/sec w/a dominant. dir. from the WSW. The amb. air temp. at the array was measured as 29.3 deg. C. All that stuff puked out an ave. calculated array temp. of about 56.8 deg. C. The temp. est. I got by measuring the panel temps. was about 58.4 deg. C.

The leading edge vs. the trailing edge panel temps. can vary as much as ~ 10 deg. C. by my measurements for my array, making the array ave. temp. an estimate whether I measure all 16 panel temps. w/ an IR thermometer or use the Sandia empirical method. Because I can get under the array, I do both to see what I can learn. I've also got a somewhat empirical (3d) method of my own that gets #'s about the same temp. as what I measure. I've found my measurements and the Sandia method usually yield results w/in an ave. of about +/-1.5 deg. C. or so to one another. That 1.5 deg. C. spread seems somewhat wind vector dependent.

Anyway, on 05/02/15 at 1305 hrs. the total P.O.A. array incident power was about 25,563 Watts. The monitor on the inverter claimed the inverter was putting out about 4447 Watts. That gave an ESTIMATED system eff. of ~ 4447/25,563 ~ .1740.

From all this, and a couple of things like est. wire losses and a calc'd inverter eff. etc., I estimated the array fouling on that day to be about 2.7 %. That # is in reasonable agreement with other fouling estimates from similar measurements taken before and after 05/02.

Bottom line: Crunching a couple of #'s, a clean S.P. 327 system could have put out about (.1740*1.027) ~ 17.87 % of P.O.A. power as an approx. steady state instantaneous rate at min. incidence angle on what I'd call an "average" clear day.

Sooo...Each 327 in my array actually managed to put out ~ 4447/16 = 278 Watts at min. incidence angle for that day. That's about 278/327 = 0.85 of S.T.C. output. (note (?): 235/280 =~ 0.84)

Most of the time, and for a lot of reasons more numerous and more varied than accounted for here, and maybe sort of Q.E.D., arrays or panels will not produce S.T.C. output - that output being something like but not completely analogous to EPA mileage.

My guess is you're doin' fine.

Take what you want of the above. Scrap the rest.

[J.P.M.]

The value I like to key in on is peak kW / kW installed, on days that are clear, and PVouput makes this number easy to see. Yesterday's peaks look like cloud effect.

Today, Poway's system maxed at 86.7%, mine was at 86.2% using Solaredge as the source (not my revenue grade meter). Makarowski's system was up at 90.0%... for reasons I don't understand, his system is routinely among the best. In most cases, this is down a couple percent from a couple months ago, when it was cooler out and the sun position was different.

And perhaps when the array was cleaner.

[sensij]

And perhaps when the array was cleaner.

Could be, but the arrays I compare against each other were installed many months apart, and unlikely to have been in the same state of cleanliness. Also, I definitely see peak conversion efficiency move up and down over short time periods, with no rain to which the increase could be attributed. Temp looks more dominant.

[J.P.M.]

Could be, but the arrays I compare against each other were installed many months apart, and unlikely to have been in the same state of cleanliness. Also, I definitely see peak conversion efficiency move up and down over short time periods, with no rain to which the increase could be attributed. Temp looks more dominant.

I'd probably agree with that. Panel/array temp. is, in my experience anyway, much more noticeable and probably more dominant in very warm/cold environments, not only for the half % or so each deg. C. decreases/increases relative performance, but also because it is close to instantaneous in its observable effect, especially in windy/gusty environments. Even if that temp. effect is not always accurately quantifiable, it's rate of effect tends to make it makes it more noticeable, and probably greater than the effect of fouling until the fouling gets => ~ 6-10% or so.

[Poway]

The value I like to key in on is peak kW / kW installed, on days that are clear, and PVouput makes this number easy to see. Yesterday's peaks look like cloud effect.

Today, Poway's system maxed at 86.7%, mine was at 86.2% using Solaredge as the source (not my revenue grade meter). Makarowski's system was up at 90.0%... for reasons I don't understand, his system is routinely among the best. In most cases, this is down a couple percent from a couple months ago, when it was cooler out and the sun position was different.

PVoutput is awesome! The bottom line is KWh/KW and PVoutput summarizes this nicely. By the way I see PVoutput showing my array at 6.695 KWh/KW but I don't quite know how you translate this to 86.7%?


But to be fair to the panels I think it is best to look at the PVoutput performance vs insolation data as it removes array pointing effects.

In this first plot we see the KWh/KW winner (6.969 KWh/KW ) of Team San Diego for the day.KWh_KW_winner.JPG Which is providing 96% of predicted (insolation based). Array has a relatively small tilt which is good this time of year.

In the second plot we see my array (6.695 KWh/KW) which is providing KWh_KW_mine.JPG which is providing 105% of predicted (insolation based)

The final plot shows a low performer (5.637 kWh/KW) for the day KWh_KW_low.JPG which is providing 108% (insolation based)


The first two are directly comparable (since both use SolarEdge data) my conclusion is that my LG300 panels are slightly better than the SW280 used in the highest performing array. Why they are better may be due to age or how clean they are vs. quality of the panel

If the Xantrex inverter is not overstating the output of the Kyocera panels of the "low performer" then these panels are the best of all (performance vs. nameplate) which is especially impressive given age of panels.

[sensij]

I'm a little bit cautious about using the insolation view in that manner, because I don't know what clear sky model PVO is using to generate their estimate. Also, some systems have temp coefficients loaded and others don't... I'm not sure yours does, and that definitely makes a difference. Most people load temp data from some nearby weather station, but depending on how accurate it is and how well it matches the ambient conditions at the array, more error is introduced.

The normalized power (kWp / kW) shows up when you drill down into a particular system on a particular day... it isn't in the team summary page. If you mouse-over where it shows you the peak power for the day just under the chart, it will tell you the percent. You can also scroll down and look for the max in the table.

A big driver of normalized energy (kWh/kW) is how much shade / cloud cover each system receives. In the examples you gave, both your system and the "winner" had some morning clouds, while the low performer had some evening shade. The winner is in Bay Park, close to Mission Bay. I was higher than normal on the kWh / kW list today, and am only a couple miles from the winner, so I'm guessing more divergence than normal between coastal and inland temps helped both of our systems. The Carlsbad systems typically outperform, but it looks like they got hit harder with marine layer this morning than those of us in the 9211x zip codes.

The low performer's peak conversion was only 76.3%, so you would expect that on a truly cloudless day, they would continue to be lower in the kWh / kW than the systems with peak output in the high 80's.

Edit: I do like the insolation view though, it really helps show the effect of clouds (both plus or minus) and shade (always minus).

[Silver_Is_Money]

There is a module rating method developed for the State of California called 'PTC' that is supposed to provide the "real world" output of just about every UL listed solar panel ever produced. The PTC rating is generally from 88% to at best about 92% of the STC output, with most panels hovering at around 89%. The various of the many iterations of the LG 300's range from extremes of 266.3 to 276.5 as to their PTC rated output, with most of the iterations hovering between 272.4 and 276.5 (and with only one at the 266 level). Call them 274.5 on average if the low rated LG 300 is excluded. That averages out to a particularly good 91.5% of STC.

Here is the PTC ratings web link:

http://www.gosolarcalifornia.ca.gov/equipment/pv_modules.php

[sensij]

Below is a better explanation of PTC rating. It was not developed by California.

http://solarprofessional.com/articles/design-installation/q-a-changes-to-the-ptc-module-ratings

[J.P.M.]

below is a better explanation of ptc rating. It was not developed by california.

http://solarprofessional.com/article...module-ratings

fwiw, +1.

[J.P.M.]

PVoutput is awesome! The bottom line is KWh/KW and PVoutput summarizes this nicely. By the way I see PVoutput showing my array at 6.695 KWh/KW but I don't quite know how you translate this to 86.7%?

The kWh/installed kW is a time integrated daylong ratio of total energy produced over a day/system size.

But to be fair to the panels I think it is best to look at the PVoutput performance vs insolation data as it removes array pointing effects.

Looking at it this way may mask orientation ("pointing") effects more than it removes them.

In this first plot we see the KWh/KW winner (6.969 KWh/KW ) of Team San Diego for the day.[ATTACH]6990[/ATTACH] Which is providing 96% of predicted (insolation based). Array has a relatively small tilt which is good this time of year.

In the second plot we see my array (6.695 KWh/KW) which is providing [ATTACH]6991[/ATTACH] which is providing 105% of predicted (insolation based)

The final plot shows a low performer (5.637 kWh/KW) for the day [ATTACH]6992[/ATTACH] which is providing 108% (insolation based)

of the "low performer" then these panels are the best of all (performance vs. nameplate) which is especially impressive given age of panels.

The peak kW system output/installed kW says much more about the max. instantaneous P.O.A. irradiance and thus probably more about clouds/cloud reflections over any period than it does about system performance.

.

[gregvet]

PVoutput is awesome! The bottom line is KWh/KW and PVoutput summarizes this nicely. By the way I see PVoutput showing my array at 6.695 KWh/KW but I don't quite know how you translate this to 86.7%?


But to be fair to the panels I think it is best to look at the PVoutput performance vs insolation data as it removes array pointing effects.

In this first plot we see the KWh/KW winner (6.969 KWh/KW ) of Team San Diego for the day.[ATTACH]6990[/ATTACH] Which is providing 96% of predicted (insolation based). Array has a relatively small tilt which is good this time of year.

In the second plot we see my array (6.695 KWh/KW) which is providing [ATTACH]6991[/ATTACH] which is providing 105% of predicted (insolation based)

The final plot shows a low performer (5.637 kWh/KW) for the day [ATTACH]6992[/ATTACH] which is providing 108% (insolation based)


The first two are directly comparable (since both use SolarEdge data) my conclusion is that my LG300 panels are slightly better than the SW280 used in the highest performing array. Why they are better may be due to age or how clean they are vs. quality of the panel

If the Xantrex inverter is not overstating the output of the Kyocera panels of the "low performer" then these panels are the best of all (performance vs. nameplate) which is especially impressive given age of panels.

The efficiencies(kWh/kW) stated in the links are the daily overall/instantaneous values. I think the lifetime efficiency is a better indicator of the long term performance. You can find this by clicking on the "Your Outputs" tab, then on your homepage, you will see the Generation ranking. Click on the "Generation" word and it will change to Efficiency ranking. This is the lifetime overall efficiency and you ranking compared to rest of the PVoutput accounts.

[sensij]

The peak kW system output/installed kW says much more about the max. instantaneous P.O.A. irradiance and thus probably more about clouds/cloud reflections over any period than it does about system performance.

.

edit: i re-read what was written about masking vs removal, and withdraw my comment.


As for kWp / kW... yes, just looking at the maximum on an intermittently cloudy day doesn't say much at all. On days like that, I scroll down the table to approx solar noon (or for systems off-south, whatever time gives best incidence angle) and look at the normalized power at the time, as long it was clear around then.

[Poway]

The normalized power (kWp / kW) shows up when you drill down into a particular system on a particular day... it isn't in the team summary page. If you mouse-over where it shows you the peak power for the day just under the chart, it will tell you the percent. You can also scroll down and look for the max in the table.

I see, this quickly gives you the peak during the day as a % of rated capacity, this will save me some math (PVoutput is filled with useful info).

I'm a little bit cautious about using the insolation view in that manner, because I don't know what clear sky model PVO is using to generate their estimate. Also, some systems have temp coefficients loaded and others don't... I'm not sure yours does, and that definitely makes a difference. Most people load temp data from some nearby weather station, but depending on how accurate it is and how well it matches the ambient conditions at the array, more error is introduced.

I was quoting the total over the day as one simple number (which can be effected by clouds etc.) but looking at the actual curves vs insolation predicted over the day. PVoutput help says the following about insolation data:


Insolation data shows the estimated live system output during a clear day. It is based on the following parameters -
Day of the year
System Latitude/Longitude
Array Tilt
Azimuth or Orientation
Timezone
Shade
Age (he just uses 1% per year, this may be why the older array looks better)


He has a few more details but basically ignores weather etc.

I still think this is the best panel metric I have found (viewing the whole perf vs insolation curve over the day and comparing with local arrays). But two things I will now consider:

1) Aging used in PVoutput may actually favor older arrays. AND
2) What you hint at and J.P.M directly addresses in later response: Looking at it this way may mask orientation ("pointing") effects...
If someone were to point their array north with a high tilt angle then the vs. insolation data could be good (if the panels worked well at low light levels) even if the panels didn't work well in full sun.

My first comparison (reviewing perf curves over the day) still appears valid.

The first two are directly comparable (since both use SolarEdge data) my conclusion is that my LG300 panels are slightly better than the SW280 used in the highest performing array. Why they are better may be due to age or how clean they are vs. quality of the panel

But this is really splitting hairs. The larger conclusion I had reached looking at this data, is that panels are truely a commodity (as a number of the senior members expouse). With no truely significant performance differences with respect to nameplate between several "good" brands.

The reason I was pointing out what I believe was a splitting hairs win was because the original poster was concerned about the performance of LG300s.

Thanks