SunPower SPR-E20-327-C-AC Power Testing

[ajonestx]

Not a rule, just something that happens often. I noticed this with my home system, lots of small off-grid systems, and a couple of big systems as well, sometimes we'll get close (or over) STC power but most of the time it's around 70-90%. I guess it has to do with less than optimum angle, sometimes temperature, system losses, and/or the sun's energy not quite matching the source used in the testing labs. Since it occurs so often we mention it so that folks won't assume there's a problem if they see the same thing.

Good info and thank you for providing a comparison point from your own system. I am good with 70 to 90% "most of the time", but just have not seen "close (or over) STC power" since my micro inverters were replaced and wondered why. I am fully aware of the impacts of weather, shade, temperature and sun energy (Watts/Sq.Meter) and just trying to stage a conversation using reasonable terms.

Up until this conversation, I too have been hearing the "80% rule" as a reason for not doing a little math to better understand module and system performance. Knowing what I have learned over the last few years (hindsight 20/20), I would have asked that "system commissioning" look a little different than just mapping micro inverters, flipping the lever and telling the client, "Look, the meter is spinning backwards!" If this comment insults anyone, I apologize. I know there is a lot that goes into system design and installation that I truly respect, but I had hoped that the commissioning exercise would have been more in depth with some type of "Functional Performance Test" as the Cx industry call it. So... I am trying to learn how to conduct a post system commissioning exercise to confirm if everything is working as it should.

Thank you again for sharing your personal system performance information for comparison.

[sdold]

Where is this 80% rule coming from

Not a rule, just something that happens often. I noticed this with my home system, lots of small off-grid systems, and a couple of big systems as well, sometimes we'll get close (or over) STC power but most of the time it's around 70-90%. I guess it has to do with less than optimum angle, sometimes temperature, system losses, and/or the sun's energy not quite matching the source used in the testing labs. Since it occurs so often we mention it so that folks won't assume there's a problem if they see the same thing.

[ajonestx]

Expecting something near peak output is not spinning wheels, I've demonstrated that is entirely feasible and is not out of the norm already.
Where is this 80% rule coming from, my panels routinely hit at or above 327w. In the 16 months that I have been tracking, the lowest peak output in day for any month is 298w in Jan of this year. Heck, one of my panels hit 328 watts last December. I just pulled all my data for May of '21 from 11:30 am to 1:30 PM - 48,997 observations - and close to half (23,392) are at or above 261w, with the average of those 23,392 observations being 303 watts.

I REALLY appreciate both of your input.

Considering Mike's max power output back in 2017 of 4.696 kW (294 S.T.C W/panel), this sounds very reasonable considering the panel temperature was around 50.8 C. I found somewhere that listed a temperature loss of about 0.34% loss per degree above 25C for SP panels which would equate to about a 9% loss in power (or about 29 to 30 kw per panel). Add a fudge factor, some inverter losses and I could see where could be seeing around 294kW. I would be happy with 294 on days like you describe, but having a hard time accepting 260 to 280kW without seeing just a little data like V, A, temp, watts per Sq. Meter, etc.

AjwWilli also has some very good points that PV systems do not just push out exactly name plate ratings, given certain conditions, they obviously do hit STC values from time to time.

I may also be spinning my wheels, but I prefer to see some type of detail to explain why I have been unable to see anything at or around (and a little above) 300 watts since the original installation. I did also clean my panels in March of 2021 and plan to clean them annually (on the list for the holiday break).

Again, appreciate your input, ideas and detailed numbers behind the scenes as real world data explains some of the generic expectations that others have given me.

[ajzwilli]

I also think both of you are spinning your wheels chasing after peak output.

> The panels are 327 watt

A 327w panel, will, generally produce only 261w, under GOOD conditions. Maybe a bit more if it's a cold day. Panels are flash tested at about 75F, in the noontime sun, they are much hotter and therefor, produce about 20% lower than nameplate, not likely to be in the optimum orientation, and unknown haze / cloud conditions.

Expecting something near peak output is not spinning wheels, I've demonstrated that is entirely feasible and is not out of the norm already.
Where is this 80% rule coming from, my panels routinely hit at or above 327w. In the 16 months that I have been tracking, the lowest peak output in day for any month is 298w in Jan of this year. Heck, one of my panels hit 328 watts last December. I just pulled all my data for May of '21 from 11:30 am to 1:30 PM - 48,997 observations - and close to half (23,392) are at or above 261w, with the average of those 23,392 observations being 303 watts.

[ajonestx]

Yea, and I just read it again. What a convoluted mess of confused communication.

.

back and forth text messages trying to explain a seemingly simple issue can turn complicated quick! Yes, convoluted mess of confused communication. A couple of tid bits - I'll try to just list the facts and avoid too much commentary.

• 2016 December - original install
• 2017 February - Installer sent me a screen shot that showed individual AC power output(s) at 318, 318, 303, 294, 266, 254, 251, 283, 300, 295, 298, 300, 305, 302, 298, 258, 223, and 191 watts [higher numbers at each end appear to be good. lower numbers are low and central in the array affected by shade].
• 2018 May - MI replacement
• 2021 March - Received a report from SP that listed max AC power output over a one-week period at 270, 270, 270, 270, 280, 270, 270, 280, 270, 270, 280, 260, 280, 270, 270, 280, 270, and 280 watts [average 270 watts]
• 2021 April - An installer removed and tested one panel. Results: 61 (voc) X 5.9 = 359 watts. After the panel was placed back into position on the roof and connected, the SP portal reported that the panel was producing 268 watts. He did NOT test the MI on this date. Live AC power output readings of the remaining panels on this date were, 272, 222, 268, 268, 268, 268, 270, 270, 270, 155, 263, 262, 240, 268, 265, 265, 265, and 268. The three lowest readings were due to shade and I have a picture of the shade condition.
• 2021 November - an installer visited the site and gave me some screen shots, but those values were much lower. I think the weather conditions were not good.

I really appreciate the detailed instructions for testing a panel in situ. It sounds like a need to get some tools: calibrated pyranometer, accurate way to measure (and record) average cell temperature and voltage / current meter (if I cannot find a vendor to test for me).

[Mike90250]

> The panels are 327 watt

A 327w panel, will, generally produce only 261w, under GOOD conditions. Maybe a bit more if it's a cold day. Panels are flash tested at about 75F, in the noontime sun, they are much hotter and therefor, produce about 20% lower than nameplate, not likely to be in the optimum orientation, and unknown haze / cloud conditions.

[J.P.M.]

@JPM - did you really read the thread? His micros are 320 watt - he even gave the model number in an early post. The panels are 327 watt. He is claiming he is not seeing anywhere near 320 AC ever since the new micros have been installed - he has only hit 270. At this point, it is not know if it is the panel or the micro inverter as he has only been looking at AC output. How did you draw your conclusion that the array is performing nominally well but the micros are undersized?

Yea, and I just read it again. What a convoluted mess of confused communication.

did see the micros are labeled 320 W and the OP is only hitting 270 W (or 280 W depending on which post(s) I read since the new micros went in (April 2021 ?). I also read that the OP hasn't seen anything above 280 W since the inverters were replaced (April 2021 ?).

Seems to me a 320 W replacement inverter would, over the course of 10 months (Feb. 2021 - Dec. 2021) produce a peak > 280 W, if for nothing more than a cloud reflection or two.

I believe there's some chance that the micros are mislabeled, how or why I know not.

I also think both of you are spinning your wheels chasing after peak output. But if that's still important, wait for a sunny day when it's not too hot and when the A.O.I. on the array is close to 90 deg. and see what the output looks like. If it's close to 320 W or so, the micros are probably 320 W. If it's closer to 280 W, they ain't.

Take what you want of the above and my other posts to this thread. Scrap the rest.

[ajzwilli]

After reading through this thread, it looks to me like your array is performing nominally well but the micros are undersized for the panels' max. output.
I think your array is running as designed but the max. ouitput is limited by the micros.

@JPM - did you really read the thread? His micros are 320 watt - he even gave the model number in an early post. The panels are 327 watt. He is claiming he is not seeing anywhere near 320 AC ever since the new micros have been installed - he has only hit 270. At this point, it is not know if it is the panel or the micro inverter as he has only been looking at AC output. How did you draw your conclusion that the array is performing nominally well but the micros are undersized?

[J.P.M.]

will ask for peak AC and DC power. Also, good find on the warranty test option - will ask for this too. I have asked two different installers if they know where / how to test... I find it interesting that of the 3 or 4 places that I have asked about testing, NO ONE seems to know. Testing V and A at DC level is pretty easy in the field, but an STC test or an inverter test is foreign to most for some reason. Agree, I should be able to see something in the 97% range (DC / sometime during the year at module level). I will have to look into isolating the issue between AC and DC more. Most of the installers that I have spoken with are pretty good at plugging the parts together, but forensic investigations at this level must be reserved for NABCEP certified folks. I sent a note to NABCEP member last week, but did not get a response. Also sent a note to NREL, but no response there either. I am also taking PV classes and learning a lot about the inner workings of the system.

Thank you for all of your ideas - every day, with more information, I get closer. I am happy to hear your system is working and it brings you joy to see it perform like it should.

If you're going to field test a single panel, or multiple panels (that is, in situ), a set of instrumenets will be required including a calibrated pyranometer and an accurate way to measure (and record) average cell temp. over a panel as well as voltage and current.

I'd suggest cleaning the panels thoroughly on the day of the test which ought to be done on a day and at a time when the angle of incidence on the array is as close to 90 deg. as possible, that is perpendicular to the panels, and the sky is as clear as possible with little or no coincident wind. All that is a PITA and all the necessary conditions are unlikely to happen coincidentally, even if you did know what to do. For example, among other things, what do you know about measuring cell temps.? Or, what orientation is best for the pyranometer? Or, how does the wind vector vary across an array ? Without care and precision and knowledge of what and how to measure what's going on, any results you get are likely to have such a wide range of uncertainty so as to be useless.

The other way to test is, as you've written, to remove the panels and send them (most likely) to a manufacturing facility and getting them flash tested. I kind of doubt that will happen but there may be some independant testing lab/service. I'd also SWAG it that the cost of such removal, testing, and shipment to and from the test facility might be prohibitive and might approach the cost of swapping out the micros.

After reading through this thread, it looks to me like your array is performing nominally well but the micros are undersized for the panels' max. output.

BTW, If you test in situ, some of the type of data you'll need is the kind of stuff I've been measuring at my home array for about 7 years or so.

As an example only: I've had 16 ea. S.P. 327's wired to a string inverter (2 strings of 8 ea.) for 8 years and watch the array like a hawk.
The highest specific daily output to the inverter I've ever measured was on 06/12/2017. That daily specific array output to the inverter on that date was 36.56 kWh/5.232 kW = 6.99 kWh/(day*STC kW) = 2.29 kWh/(panel*day).

The highest instantaneous power generated and sent to the inverter for that day was 4.696 kW (294 S.T.C W/panel) at 1309 hrs. P.D.T.
The beam irradiance angle on the array at that time was 8.16 deg.
The average array string voltage (2 strings) at that time was 386.8 V/string.
The average string current at that time was 6.07 A/string.
The average cell temp. for the array at that time was 50.8 C.
The ambient air temp. on the roof at that time was 24.5 C.
The ambient air ground level temp. at that time was 20.2 C.
The P.O.A. irradiance on the array at that minute was 1002 W/m^2.
The average wind vector magnitude across the array at that time was 2.1 m/sec.
The sky was cloudless as is a requirement for all such measurements.
The array had been rinsed, brushed and rinsed again at 0700 hrs. every day since 06/01/2017.

Some other such clear sky measurement days had slightly higher or lower clear sky instantaneous output, but none higher than 321 W/panel to the inverter.

I think your panels are running as designed but the max. output is limited by the micros.

[ajonestx]

I forgot to add, I believe SP should be guaranteeing you to be at 98% the first year and .25% reduction each subsequent year, but never below 92%. So in year 5, you should be hitting at least 97% per STC. This is usually on the DC side, not the AC.

I am assuming you were aware you can hover over the "info" in the 1st screenshot to get both AC and DC production. The value displayed is AC production. Have you isolated the problem to being DC or AC (i.e. has DC production ever been near STC). I think the inverters you have are 96% efficient, so you should be seeing similar values between AC and DC when production is "stable" (nice cool sunny day with no clouds / shade)

will ask for peak AC and DC power. Also, good find on the warranty test option - will ask for this too. I have asked two different installers if they know where / how to test... I find it interesting that of the 3 or 4 places that I have asked about testing, NO ONE seems to know. Testing V and A at DC level is pretty easy in the field, but an STC test or an inverter test is foreign to most for some reason. Agree, I should be able to see something in the 97% range (DC / sometime during the year at module level). I will have to look into isolating the issue between AC and DC more. Most of the installers that I have spoken with are pretty good at plugging the parts together, but forensic investigations at this level must be reserved for NABCEP certified folks. I sent a note to NABCEP member last week, but did not get a response. Also sent a note to NREL, but no response there either. I am also taking PV classes and learning a lot about the inner workings of the system.

Thank you for all of your ideas - every day, with more information, I get closer. I am happy to hear your system is working and it brings you joy to see it perform like it should.

[ajzwilli]

I forgot to add, I believe SP should be guaranteeing you to be at 98% the first year and .25% reduction each subsequent year, but never below 92%. So in year 5, you should be hitting at least 97% per STC. This is usually on the DC side, not the AC.

I am assuming you were aware you can hover over the "info" in the 1st screenshot to get both AC and DC production. The value displayed is AC production. Have you isolated the problem to being DC or AC (i.e. has DC production ever been near STC). I think the inverters you have are 96% efficient, so you should be seeing similar values between AC and DC when production is "stable" (nice cool sunny day with no clouds / shade)

[ajzwilli]

The monthly email reports are a good idea, they mostly match up with my monitoring but some months are 85% of my peak, so maybe they aren't using as frequent intervals as I am (hourly maybe?) Since you are pulling panel level data occasionally, you have been able to deduce that the low system production isn't due to a couple bad panels/inverts (however on your screenshot, a few of the invers were significantly lower than the rest of the system - but since that is only 1 data point - it could be shade/clouds/soiled panel, etc.)

In terms of the the original question, according to SP warranty, they can provide a list of recognized testing agencies - since I am assuming you are going after a warranty claim, it may be best to use one of their testing agencies.

Footnote 2 of the warranty:
Minimum Peak Power” is defined as the minimum rated power, as shown on the label. Peak Power is defined as the watt peak at Standard Test Conditions (1000 W/m2 irradiance, AM 1.5, 25°C. SOMS current, LACCS FF and voltage from NREL calibration), as described in IEC 61215, measured per IEC 60904, and accounting for 3% measurement tolerance. SunPower AC Modules shall, in any event, require a sweep rate of no less than 200 ms to ensure an accurate power measurement. SunPower can provide a detailed testing procedure or a list of recognized testing agencies upon request

While you are talking to them (directly or via an installer), I would ask them to see if they can provide you (maybe verbally) what the peak AC and DC power has been this year for each panel/micro.

Good luck!

[ajonestx]

On both the new and old website, the 5 minute power increment is still there. For the apps, looks like the new app is in 15 minute increments for power versus previously 5 minutes for the old app (old app and website sunset at the end of this month).
So your system was installed in Dec '16, micros were replaced in Feb '17 and you haven't hit anywhere near STC since? Mine was installed in April '20. How many data points do you have since Feb '17? Can you share how you are reviewing the data? Are those random views from the sunpowerconsole site or are you viewing daily power from the consumer Sunpower website / app?

If from the consumer website/app, you have reviewed the power data in the app/website for every single day since micros were replaced in February 2017, looking for the peak 5 minute interval and the peak POWER production is 4.6kW for any interval? Did you click through daily data for over 3 years of data or did you find someway to download the data?

The reason I ask is that unless you have lots of data points (i.e. viewing each day and each 5 minute interval), you are likely to miss it. For each panel I have nearly 30,000 observations logged, and only 9% of the time is the output greater than 270.

ok. I found the "cloud" download button on the new website. Thank you for letting me know!

I used to download daily data and was very excited for a bit, but have experienced our fair share of inverter and communication problems which dulls the excitement some. I now just review the monthly reports that get emailed to me that list system monthly and lifetime maximum power levels to get a system level feel for performance which does not really tell me much since I have some shade during the Winter when the sun is low in the sky.

Correct, I have not seen a single instance of system power output near STC in three years. The few occasions that I have been able to obtain some inverter level data, I have not seen anything above 280 watts since the micro inverters were replaced.

Which leads me back to my original question - does anyone know where I can get a panel / micro inverter tested?

[J.P.M.]

On both the new and old website, the 5 minute power increment is still there. For the apps, looks like the new app is in 15 minute increments for power versus previously 5 minutes for the old app (old app and website sunset at the end of this month).
So your system was installed in Dec '16, micros were replaced in Feb '17 and you haven't hit anywhere near STC since? Mine was installed in April '20. How many data points do you have since Feb '17? Can you share how you are reviewing the data? Are those random views from the sunpowerconsole site or are you viewing daily power from the consumer Sunpower website / app?

If from the consumer website/app, you have reviewed the power data in the app/website for every single day since micros were replaced in February 2017, looking for the peak 5 minute interval and the peak POWER production is 4.6kW for any interval? Did you click through daily data for over 3 years of data or did you find someway to download the data?

The reason I ask is that unless you have lots of data points (i.e. viewing each day and each 5 minute interval), you are likely to miss it. For each panel I have nearly 30,000 observations logged, and only 9% of the time is the output greater than 270.

My Sunpower data is in excel format and mostly 5 minute continuous from 10/13/2013 onward which was nice but I quickly found dropouts of a few data in most days which made like less rosy for data analysis, especially when attempting time correlation with 1 minute data from the Davis weather station on the roof.

Also, confirm the old Sunpower website with the 5 min. data is being retired 12/31/2021. I'm looking around for another way to monitor/log output. It's now a shadow of its former self anyway with only array power and energy output data. In the past it was real good including V's, I's and lots of other stuff that's displayed on my string inverter. My guess is, among other reasons, that S.P. got tired of getting questions from users who didn't know what they were looking at.

[ajzwilli]

On both the new and old website, the 5 minute power increment is still there. For the apps, looks like the new app is in 15 minute increments for power versus previously 5 minutes for the old app (old app and website sunset at the end of this month).
So your system was installed in Dec '16, micros were replaced in Feb '17 and you haven't hit anywhere near STC since? Mine was installed in April '20. How many data points do you have since Feb '17? Can you share how you are reviewing the data? Are those random views from the sunpowerconsole site or are you viewing daily power from the consumer Sunpower website / app?

If from the consumer website/app, you have reviewed the power data in the app/website for every single day since micros were replaced in February 2017, looking for the peak 5 minute interval and the peak POWER production is 4.6kW for any interval? Did you click through daily data for over 3 years of data or did you find someway to download the data?

The reason I ask is that unless you have lots of data points (i.e. viewing each day and each 5 minute interval), you are likely to miss it. For each panel I have nearly 30,000 observations logged, and only 9% of the time is the output greater than 270.