I have 22 - 220w panels that power my RE system. This is a theoretical output of 4840w.  The system is less than a year old, so I am not quite experienced with it's ins and outs.

When I got the system in December, I understood that the winter output would be pretty small. It was winter, with the short days, frequently overcast skies, etc.

In the long spring days of May and early June, it was not uncommon to get 3.8+ kW out of the system, bringing in +/-30kWH/day.

During the height of the summer, I had hoped to get closer to 4.0-4.5kW out of the panels. Instead, I rarely have gotten over 2.8kW, or ~20kWH/day.


I know that some panels experience some decrease in performance due to heat, but it seems like I am only getting half the output I should be getting, and more distressingly, the output has dropped rather than increased.


The system was professionally installed, but like many contractors communication (especially 7 months after installation) is somewhat lacking.  


Any thoughts?

Sounds like something is wrong.

Do you have the ability to measure the output of the panels and relate that to what the converter is showing?

Sounds like a clip-on DC ammeter would be nice to have to check each panel string...  

Also, what's the system say the panel  voltage and current output is [are] and compare that to Springtime values.

Very possible a bad panel or bad connection.

Feel each connection for high temperature, but I expect if one were bad, it would have burned up by now...

We'd need more information to be sure, e.g. direction the panels are pointed at and angles of the panels.

I usually get peak output in the winter on bright, cloudless days due to low temps and a panel angle that was optimized for
winter output (fewer hours in a day.) No question on the BOL setup at least that total power produced peaks in May due
to longer hours, no clouds, and the sun isn't at its highest at noon, and temps are still low.

I'd still go with the previous suggestion and have each string you have tested for output in case you've got a dead panel
or something like a vent shading something so its output is off.

Thanks for the input, guys.

The panels face due south. My latitude is about N 44*. I don't know the angle of the panels, but I will try to find out.

There are two strings of panels, each with an Outback charge controller. The numbers I am giving are from the charge controller's display. Do I need to put a meter on the wires coming in as well?

Here's a photo of the panels as they were being installed. Sorry for the poor quality.








Edited to add photo

Is it possible to read the power out per string?

Also not sure on your breaker setup, but if the strings have individual breakers, try popping both
off, confirm you're getting zero output, and then switch one on/off and then the other and see if
the power levels are comparable.

In the south, we have more cloud cover in the Summer than the winter. Could that be an issue?

Each of the two strings has it's own charge controller. That's where I am getting the numbers.

I guess I have not kept track the cloudy days vs. sunny days, but the days where it IS sunny the day's total output is lower.

 
Mine does the same thing.  March-may and September are my peak months, a so generating the highest peak power for a given day.  I'm at 39*N, and my panels are oriented almost due south at 45*.  During the summer months the sun is much higher in the sky relative to the angle of the panels.  I've got 14 each 250w sharp panels in a single string, so it's easy to measure.  My controller logs each day's peak power.  Back in may I hit 4024w out of my 3.5 kW array.  

OP, you're in luck.

Compare each of the 2 Outback readings for current and voltage ---and see if they are essentially identical.


Or maybe the issue is as H-S suggests...

All good answers above. If no joy after those, try disconnecting each panel and check it's output individually at full sun.

About 7 years after we installed our first KC120 panels, I started noticing we just didn't have the power we used to. Went through all the check, scratched my head, then started disconnecting panels and checking them in full sun.

Panels that should have read 17-18 volts in full sun were reading nine.

3 out of 14 of the KC120's were not producing. Called Kyocera and they replaced THE ENTIRE ARRAY. Best part also was they had discontinued the 120's and went to a 130 so they sent 130's in replacement. It was like getting an extra panel free.

Thanks guys.

I recall the installer telling me that (for some reason I can't exactly recall) the strings were not even, that one string had a couple more panels on it than the other. That makes it tougher to compare strings.

And to compound that, here's my confession: In the interests of being cheap, I initially bought 10 of the 220V panels from Solar Blvd. Great price. They sat a year in storage in my garage for me to save enough to get the rest of the system installed. My installer discovered that they were LG panels that had apparently been sold as seconds and been relabeled. My installer found something wrong with each panel: On one the glass was very slightly hazy. On another there was a cracked cell, etc.

I was penny wise and pound foolish.

These 10 panels my well be the cause of my problems, but their output had been within specs a couple months ago.


I really hope Home Slice's answer is the correct one, but I will do some checking. It's gonna be a bear to climb up there, disconnect and check each panel!

OP, the Outback's will show the wattage being produced by each of the two strings at any instant.

Simply calculate the expected wattage of what one string, and the other, should be producing. By adding up the panel wattage considering how each string is wired with respect to panel arrangement.

Then look at the ratio of the expected wattage of the 2 strings and the ratio of the wattage outputs as displayed on the two Outbacks, and compare.



Incidentally, are you using The FM80 model Outback charge controllers?

Maximum input voltage to them is about 150 volts. Many of the 220 watt panels output about 30 vdc so the maximum quantity of panels in series are about 4. So this implies that there are a couple series strings in parallel -if this makes any sense.

Just speculating on the above, it seems you need to determine what the panels wiring arrangement and the contractor should have left you with some installation/system design info. Required to submit for your .gov and power company credits.


Outback FM80 link

Been running a handful of the panels from Solarblvd, although not that exact type- and never had a problem with them.

Keep in mind the installer was probably one of those guys that asks people to do a "energy worksheet" wherein you list everything in a normal house then he would show you a $60K system and tell you that's the minimum you need to get it done..... I can't tell you how many times I've heard that from people.

So basically he makes his money from selling you the components and some from installing them. You took away some of his profit margin (the markup he would have received on the panels) so of course he isn't going to be too happy about that.

Check the panels yourself individually the way I mentioned, that will give you a per panel idea of what is working and what isn't. My question to him would be why he installed them if he thought they were FUBAR?

Also, your far up north IIRC. Some friends in VT put in a similar system to mine to begin with, then wondered why mine situated in the deep south did better production wise than theirs.

These are just some of the things you have to work through with an energy system sometimes.

Heat might have something to do with it, but it's really the angle of the dangle, i.e. if your panels are sloped to give you more output during the winter, then you will have less during the summer, some of us in the South have two struts to change from winter to summer.

Rancher

I set the panel slope of the 2 systems here for winter sun -and leave them there, and have great harvesting during the summer.

Plus the summer days are longer.

Exactly what I did, my BOL house has the roof over the garage built to the winter sun angle, so everything is just surface
mounted to that. Longer days makes up for the off-angle in the summer. Simpler mounts, no moving parts = high reliability.

Just an update for anyone interested:

As time progresses, it seems my production is trending up again. Yesterday I made almost 27kwh, which is a definite improvement. I am definitely going to keep an eye on it.

Thanks again for all the ideas regarding what might be going on.

Less output in hot weather is normal. the derating data is   printed, back side of panel, or freely available from module manufacturer.
 If your predominant   Solar  power load is recharging batteries....    then you may  be  just USING less power per day in the summer.



. example , my batteries are  usually  at 100% by  noon or 1PM, after that very little current flows  from the   offgrid arrays.

.

 My Gridtie Arrays make a LOT more power in  spring and fall months,   peak power outputs are usually on cold sunny days with snow on the ground.

With 22 modules, you've probably got 10 going to one charge controller and 12 going to the other.  Strings of 4 is probably a bit high once module voltage is corrected for temperature.  Down here, a 60 cell module has a corrected Voc of about 42V...  So 4 modules in a string would be a problem for the FlexMax charge controllers.  In Wisconsin, your correct Voc would be even higher.  

I'm betting 4 strings of 3 on the one CC and 5 strings of 2 on the other.  All strings connected in parallel to a single CC better be the same length - if not, it hurts production pretty badly.  

If you can measure the temperature on the back of the array, you can apply some rough math.  Power drops about 0.45 - 0.5% for each 1C that the module is over 25C.  So if you measured 65C on the back of the array (not unheard of), you'd lose about 18 - 20% of your name plate power...  assuming "full sunlight".  

You also need to know "how much" sun is shining before you can figure out if there is an array problem.  Array output current is tied to irradiance.  The Imp listed on the label only happens at "full sunlight" - 1000W/sq m.  We use a pyranometer to measure it.  If it reads 1000, then I check for the output current of each string to match the module's nameplate Imp (maximum powerpoint current).  If it reads 500, look for 1/2 of Imp.

I use this doodad, an IR thermometer, a couple of clamp meters and a DVM when performance checking a new PV system  -