Flat maximum energy production

[scrambler]

Can you describe what instrument you used, how you set it up, how the data looked and what it cost ?

In my case it was part of an Ambient WS2000 weather station that has a small solar panel on top of it. Rough but still matches the Solar production pretty well.
Since then because I have two of my 20 panels on Enphase micro inverters, I setup monitoring of these two Enphase panel with an Emporia Vue 2 Energy monitor, and it gives me the perfect baseline for what the other panel should be producing

Add: Given the cost and the likely fact that the problem is not the irradiance but clipping, seems to me that before the OP embarks on what could be a costly in terms of time/treasure and probably error prone quest to measure irradiance, it might be better to figure out where the clipping is coming from.

I am definitely not suggesting to invest any significant amount of money on that. Just mentioned it in case he hits a wall with the installer.
The argument in itself should be sufficient. And at worst he can wait for summer to show his installer that his explanation does not hold...

[J.P.M.]

One thing you could do to support your argument is use a Solar radiation monitor beside the Array, record that and put it in correspondence with the inverter solar production graph.

I had a bug in my inverter last year that the Provider refused to consider, but I have a small weather station that has a solar panel and records solar radiation during the day.
By graphing both Solar production and solar radiation together, it was clear the Solar production was not following solar radiation and so the inverter was up to something..

They finally agreed to look it up and did find a bug that was later fixed.

In your case I don't think it is a bug, but if they do not listen to a rational argument such as the one I mentioned previously, this could add some concrete element.

FWIW, I agree that irradiance monitoring is a good idea - I've got 1 minute monitoring records for almost 8 years continuous - but without care about how it's done in terms of instrument set up and then data interpretation and instrument calibration error it can create confusion, not to mention the lack of credibility if an installer knows anything about irradiance measurement/monitoring and it's not done right.

Can you describe what instrument you used, how you set it up, how the data looked and what it cost ?

Add: Given the cost and the likely fact that the problem is not the irradiance but clipping, seems to me that before the OP embarks on what could be a costly in terms of time/treasure and probably error prone quest to measure irradiance, it might be better to figure out where the clipping is coming from.

[scrambler]

One thing you could do to support your argument is use a Solar radiation monitor beside the Array, record that and put it in correspondence with the inverter solar production graph.

I had a bug in my inverter last year that the Provider refused to consider, but I have a small weather station that has a solar panel and records solar radiation during the day.
By graphing both Solar production and solar radiation together, it was clear the Solar production was not following solar radiation and so the inverter was up to something..

They finally agreed to look it up and did find a bug that was later fixed.

In your case I don't think it is a bug, but if they do not listen to a rational argument such as the one I mentioned previously, this could add some concrete element.

[smbunn]

Awesome, thanks Scrambler

[scrambler]

This seems to confirm that the MPPT raising voltage and dropping amps to do the clipping which is consistent with what we sam on the 3500W data.

If you did the test of shutting down one inverter and that did not affect the clipping on the other one, then the clipping is not at the grid level but at the inverter level.

So you are back to figuring out what is making the inverters do that...

As for your installer remark. You can tell him, that the sun makes an arc in the sky, and regardless of the season, on a clear sky day, this translate to a Bell curve for the solar radiation.
If all panels have the same exposure (which I understood is your case), they all have the same irradiation, and therefore the production curve will mirror the Radiation bell curve.
Depending on the Orientation, that bell curve can be skewed left or right, and depending on the season, the maximum reach will be higher or lower, BUT the curve will be a bell curve.

There is no natural phenomena that can make the solar radiation curve flatten (and therefore the production in your case). So something is purposely clipping your production and should not be any different when summer comes.

Good luck with them.....

[smbunn]

Here it is

[scrambler]

Can you share the data file of the 2500W clipping so we can check the MPPT values

[smbunn]

So I have had quite a long run. Unit one with the 2,500 limit is clearly clipping at that limit regardless of what is going to or coming from the grid

solar2a.png

And unit 2 which has no limit is happily clipping along at 3,500 watts as we saw before

solar2b.png

Would you now agree that the clipping is not associated with my energy export?

[smbunn]

its winter so cannot guarantee a long stretch of sunshine, not this week anyway

[scrambler]

For the shutdown of one inverter, I would do a few hours test rather than a few minutes
For the 2500W limit, you should do a full day test during a clear sky day to have clean data, especially the MPPT values

[smbunn]

Right, the sun came out. I set unit one to the 2500 export limit. Unit two has no export limit
I ended up dropping generation from 6,800 watts to about 5,800 (i.e. it dropped 1000 watts and unit one dropped from 3,500 to 2,500)

solar1a.png
You can see that in the image above. The red line is when I changed the limit.

Unit two however looks like this

solar1b.png

As you can see it carried on generating a maximum of 3,500 watts when in full sun. So unit one is clipping at 2,500 which is not the export CT on my grid connector but the export of the unit itself

I assume this is helpful?

[smbunn]

I was at the 6,800 watts yesterday. Both units giving 3,400 each (there or less) and shut one unit down. The other remained unchanged at 3,400 watts. Waited 5 minutes, saw it updating every minute but only be a few watts, it remained at almost excatly 3.400 watts. Restarted unit 1, it took a few minutes to boot up but when it did it got back to 3,400 watts

Today I have a lot of cloud, will wait till I get a sustatined run to do the 2,500 watt max export function on the one unit with this feature

[scrambler]

Tell us what happens when you shutdown one, and share the data when you set one to clip at 2500W

[smbunn]

Thanks for the input. By the way I have a degree in Electrical Engineering and run a sizable software company that specialises in energy management and green house gas reduction...not so much of a neophyte

Its cool but sunny today, 16 degrees Celcius and strong breeze so pretty cold arund the panels. I am confident its clipping, just need to figure out what is clipping since both SolaX units are doing the same thing but only one has the export limit function

[J.P.M.]

Yes, despite the E-W thing, I have pushed the DC:AC to just going into clipping. Only just,
there is NO bell shaped curve potential rising far above that line. Loss of any element
would cause dropping below that 15kW line. This was demonstrated last year when my
weakest (oldest) panel string (of 6 strings) burned out a connector.

Have not developed the estimated clipping meter just yet, adding that info might make my
first curve look closer to my second.

There is not any point going after the 8.5kW level, since he is restricted to exporting 5kW.
UNLESS, he can daily coordinate a heavy load with noon sun.

If I could avoid shade till 5:30pm, I would not bother with things like optimizers. Bruce Roe

Bruce: I was referring to your statement: "... each of those PV systems should have a 5 kW capability without clipping". That's all.

My point was - they probably won't (exceed 5 k woutput, at least not for any sustained length of time.
We all know that a system will have little likelihood of sustaining its STC rating for very long unless it's way above sea level, and/or operating under very cold ambient conditions, and or with a lot of albedo, or some combination of those conditions - all of which are unlikely to boost output over the STC system rating as a steady state type of situation and even less likely to be found in combination with one another - except maybe by design and it doesn't look like the OP has any of those design enhancements - but with the dearth of information we have at this time it's hard to know just what we're dealing with here, which makes all this somewhat of a cluster function.

My read of the OP is that he's a neophyte and might be thinking he's got a problem because, among other things, his 10 STC kW system isn't putting out 10 kW. I'm just trying to avoid the OP getting any more confused. He's got a problem, but being out of his depth he doesn't need more confusion.

Smbunn: Your system is clipping. The goal is to find out why. I don't think we've got enough information yet to find a probable cause.