Seasonal effect on testing solar panels

tkc100

When testing the output of a solar panel I assume it is necessary to account for seasonal and latitude differences.
I cruised the internet and found some incredibly complicated and stuffy reports but nothing I could put into practical use.
Today I was testing 6 panels very near noon. I pointed the panels as best I could so that they were at a 90 degree angle to the sun.
I adjusted them until I got the best voltage reading.
It's January the 30th and my latitude is 29 16' 5" North Clear skies
Should I expect lower output reading than I would say on July 4th?
I am also assuming that the rated output of a panel is done under near optimal conditions.
Any advise, will be greatly appreciated.
Thanks!

niel

Re: Seasonal effect on testing solar panels

usually our tests are approximate as we normally don't control the temperature or solar insolation during such testing. for someone as far north as you the sun will output a bit less at this time of year due to more atmospheric interference. on the plus side the lower temps will bump the wattage ever so slightly upward this time of the year (accounting for the amount of solar insolation) and you may even get an increase in output from snow and ice reflections.
you can still check the open circuit voltage and short circuit current to see if the pvs are working in a rough ballpark area for you, but without controlled conditions such as a lab would give it would be difficult to say if a pv is off or not unless it is somewhat significant.

tkc100

Re: Seasonal effect on testing solar panels

Thanks for the response.
I normally do the quick multimeter test but for reason I can't explain I decided to dig a little deeper today.
I tested the panels by placing a calculated resistance across the load and then metering the voltage.
The procedure is covered in detail at this URL
http://www.mtmscientific.com/solarpanel.html
It's time consuming but seems to be far more accurate.

Perhaps I typed in the wrong latitude because I don't live in the north at all.
I'm down on the border of Mexico.

My question is does the sunlight here in January produce the same amount of electricity as it would in the middle of the summer?

I understand I will not be able to reproduce the environment these panel were calibrated at but just looking for some guide lines.

What does experience tell you? + or - say 10 %
If a solar panel is within 20% of its rated value is that acceptable or within the range of unavoidable error given two different testing environments.

RWB

Re: Seasonal effect on testing solar panels

I have been asking and wondering the same question about testing solar panels this time of year vs Summer.

Here is a Device that allows you to measure insolation which in turn lets you know what your panels are working with from the sun.

Check it out: http://daystarpv.com/solarmeter.html

BB.

Re: Seasonal effect on testing solar panels

You can use the hourly output from the PV Watts program to show the "average" output for xxx watts of solar panels on a GT system.

However, you really need to know the exact solar radiation right at that moment in time.

Use a "known good" solar cell either with a shunt resister or a DC Amp Meter / DVM... The "shorted" output current of a silicon solar panel is pretty much proportional the amount of solar power (watts per sq.unit).

Use the known good/calibrated cell to "calibrate" the amount of sun on your test panels.

Note that solar cell current increases a bit as temperature rises, but panel Vmp falls quite a bit as temperature increases. You can use this nice set of generic crystalline solar panel charts and specs. to see the various relationships.

Of course, for "real life solar" you need to run the panels under actual conditions of use... Either a PWM type load (really solar panel connected to a 1/2 charged lead acid battery--so fixed panel output voltage) or MPPT type (controller working to maintain Pmp=Imp*Vmp).

On a real system without some sort of pilot cell to account for sun and temperature--I would guess on my GT system that I could not even see a 10% difference (looking at power meter on GT system and looking at sky/sun shine).

-Bill

RWB

Re: Seasonal effect on testing solar panels

BB - would guess on my GT system that I could not even see a 10% difference (looking at power meter on GT system and looking at sky/sun shine).

Are you saying that you can have not noticed a 10% difference between your panels output during Summer vs Winter months?

BB.

Re: Seasonal effect on testing solar panels

No, for my 3.5kW (3kW CEC rated) on average, a good clear winter day I get ~10kWhr (of course, those are rare days). On a hot summer day perhaps 16-17 kWhrs per day. My peak was ~20 kWhrs maybe once a year in September.

I am just saying that I could not see a 10% change in power output due to problems or other changes (such as cleaning my dusty array with cold water on a hot summer day)... The changes from cleaning cooling just did not stand out from normal daily output drifting due to sun, position, weather, etc.

Average winter peak output is around 2kW and average summer peak is around 2.5 kWatts. This is just random watching the power meter as I walk by it during the day--I have not bothered to hook my inverter to a computer. On a rare really good day--I will see nearly 3,000 watts peak.

I live 20 miles south of San Francisco on a couple miles from San Francisco Bay (sunny side of the coastal ranges).

-Bill

Kamala

Re: Seasonal effect on testing solar panels

niel wrote: »

... for someone as far north as you...

Poster is roughly 3 degrees north of the Tropic of Cancer. Sun is nearly directly overhead at both the vernal and autumnal equinoxes.

niel

Re: Seasonal effect on testing solar panels

good catch on my bad kamala.

RWB

Re: Seasonal effect on testing solar panels

Thanks BB. So do u think that that daystar pv meter would show the actual suns output and you could just do the math on your panel to get a pretty accurate estimate of what your panel or panels could produce.

BB.

Re: Seasonal effect on testing solar panels

Somebody like Solar Guppy or boB or Robin who works with characterizing solar panels can probably give you a better idea.

Are you trying to see if you are purchasing/running "good panels" or are you trying to really deeply characterize the panels themselves?

Also, are you trying to setup a lab accurate system, or just using the meter readings from a random solar charge controller / GT inverter? Some vendors are known to have better than +/-2% accuracy on their production equipment and other vendors are known to "average high" of near +5% or more (nobody ever complains if their solar system "generates" more power than they thought it should. :roll:

There are so many possible sources of error--you could spend $1,000's to $10,000's of dollars to get better than 2-3% accuracy.

If you are just doing comparisons of panel output to make sure they are good panels, that Daystar meter is probably on the expensive side.

From another thread, Jeff Fedison posted a link to his small company's web site to show how you can build or buy his version of a 5% accuracy solar irradiance meter--basically a small shunt resistor and DVM panel instrument--for much less (I have never done anything with this website--but somebody else here said that they are good to work with).

If you are looking for longer term automated measurement--They also have a battery powered data logger for, what seems to be, a nice price.

Also, you will need an IR thermometer of some sort to measure panel/cell temperature for accurate temperature compensation.

Also, accurately measuring the panel's output power needs to accurately know your output loads. How the device behaves (PWM, bad MPPT, or Good MPPT, fixed voltage, variable voltage, etc.) will greatly affect the "panel output energy".

If you are trying to compare brands of panels--having a repeatable load is going to be very important (MPPT based load is a good start).

Getting repeatable measurements in the +/-10% range can probably be done with "cheap" off the shelf and home made equipment. Anything more accurate is going to need much more $$$$.

-Bill

tkc100

Re: Seasonal effect on testing solar panels

Thank you all for all the input.
I started this thread after testing 6 Solartex panels outside.
The output of the panels was a bit below the stated values on the tag but then I'm not sure on the age of these panels. After doing all my testing they were consistently coming in about 10% below the figures stated on the tag.
It crossed my mind could part or all of this loss be due to the fact that I was testing them in January.
Since testing the panel I've done a little research. It answered my question by provides little practical usable information.
From the attached chart it is easy to see that there is considerable more solar radiation available in the summer than in the winter. This is the direct effect of the angle of the sun or the amount of atmosphere the sun must penetrate to reach a PV. I hope the chart is readable as I had to down size it quite a bit.
Another chart I found, I love pictures, is from a PV array outside San Fransisco I will attach it also.
So in conclusion, Is there a seasonal differences in the solar radiation that reaches a given spot on the earth surface. Yes. The differences are greater the further you are away from the equator.
So a nice to know but can I put it to any practical use? No.
Could it account for the lack of performance in the panels I was testing? Perhaps
Without some sort of nifty conversion calculator, where I could plug in the time of the year and my latitude and it would give me a correction factor, the best I can do is make an "educated" guess.

dwh

Re: Seasonal effect on testing solar panels

tkc100 wrote: »

I started this thread after testing 6 Solartex panels outside.The output of the panels was a bit below the stated values on the tag but then I'm not sure on the age of these panels. After doing all my testing they were consistently coming in about 10% below the figures stated on the tag.

That might not be unusual for older panels. From what I understand, output does go down over time.

niel

Re: Seasonal effect on testing solar panels

know that the tag lists for max power under ideal conditions or stc as we know it. many go with ptc or cec values as being more realistic, but i think given that solarex has not been around for quite some time that your getting within 10% of the stc ratings given to the pvs is good.

now solarex pvs are too old to be listed under the cec listings and bp solar now is the owner, but you can see in the link how other panels really have done.
http://www.gosolarcalifornia.ca.gov/equipment/pvmodule.html

it's still true that the temperature and insolation levels are not accounted for as most tests by us are still rough being we lack lab test parameters.

tkc100

Re: Seasonal effect on testing solar panels

Once again thanks for all the input.

Yes I know that Solarex is no longer around but I haven't been successful in obtaining any sort of date code from BP. I e-mailed them twice with on response.
With a serial number like RK89J170194TF I would "guess" 1989, or is that just too easy?
I attached the tag information just for grins. I haven't read any that was so confusing since the last time I negotiated a cell phone contract.
I tested my 6 panels and with a 5 ohm load they produced between 54 and 52 watts on a clear January day in south Texas. I didn't record the temperature but I would guess it was in the upper 60's perhaps low 70's.
If in fact these panels were made in 89, 21 years ago it says a lot about the panels Solarex made. As far as I know these panel have been on the roof out in the desert sun for the whole time.
I will be putting these panels back to work soon.

niel

Re: Seasonal effect on testing solar panels

try the short circuit test (no resistor) and tell us what you get. watch that you connect the meter properly to avoid damage to the meter or blow an internal fuse in the meter.

tkc100

Re: Seasonal effect on testing solar panels

niel
I attempted to do the short circuit test but the Multimeter I have wasn't up to the task. It has a 0 to 10 amp scale but seems to read only in full digits. In other words 2 and 2.9999 all read as 2. So all the panels read 2 amps. That is the main reason I took the time to chart the panels using fixed resistance.
The procedure I followed is detail in the document that can be found at:
http://www.mtmscientific.com/solarpanel.html
I would attach the actual file but it's is bigger than this forum will allow.
What do you think about this procedure?
The resistance I chose to use was 50, 25, 20.7, 10, 5 and 2
I entered the data in a Excel spread sheet and was able to plot a curve very similar to the one shown in the article.
I could post the chart within a zip file if your interested.
I have ordered a new multimeter but as it turns out it is on back order and quite literally on a boat from China. The vendor is telling me some time in the middle of the month.

niel

Re: Seasonal effect on testing solar panels

you are graphing the approximate power curve for the pv. the exact knee of the max power point you are kind of guessing at as more resistance plots would show it more accurately. we usually go with the open citcuit voltage and short circuit current in our measurements as they are the easiest to measure and compare to the original. it is still rough because of variables we don't account for that a lab can. those factors are exacting insolation in watts/m^2 being set to 1000w and the panel temperature being at 77 degrees f (25 degrees c).

tkc100

Re: Seasonal effect on testing solar panels

niel
You know I thought of that but the figures were interesting.
Voc 21 volts
At 10 ohms resistance 18.7 volts 1.87 amps or 34.97 watts
At 5 ohms resistance 16.3 volts 3.26 amps or 53.14 watts
At 2 ohms resistance 7.7 volts 3.85 amps or 29.65 watts
The knee in the graph is really steep and the difference in resistance is small.
With just three resistors and subsequent tests it's possible to get a good idea of the panel's output. Perhaps there's a bit more to be had one side or the other of 5 ohms but when I looked at it all graphed out I don't think it could have been much.
Is this time consuming, without a doubt and if all you are interested in is a relative go or no go test then I would certainly go with throwing an amp meter across the panels output terminals. If you've got a good meter that you can trust.
As I mentioned in my previous post I really questioned the credentials of my meter so this offered an alternative until the next boat from China gets in. In the end no matter which method of testing is done as you have pointed out numerous times thorough out this thread the results are only relative. Without the exacting controls found only in a lab the output figures are going to be variable and not likely to match what's printed on the tag. I have a suspicion that even in the lab there is a likely margin of error.
Thanks for taking the time to kick this back and forth with me.
I have learned a lot.

niel

Re: Seasonal effect on testing solar panels

most likely the max power point will occur about 4.4 ohms. that is the approximate internal resistance of the pv calculated by their specs. this may not apply now as the pvs are older now.

tkc100

Re: Seasonal effect on testing solar panels

I think you may very well be right.
I don't have a high wattage rheostat that is likely to maintain that fine of tolerance but if you look at the attached graph it is easy to project.
So perhaps the old PV is capable of a couple of more watts.
You know I'm pretty happy with the results. 6 panels out there doing their thing in the sun for 21 years and still working. Wish I could say the same for my dishwasher.