Good to be here!

[Fourdoor]

Hey everyone,

Just had my 9 KW solar array installed late last month, and it was approved for use and activated early this month (4th of April). My setup is on a perfect south facing roof with some trees to the East and West that have almost no impact on production. I had my first "perfect" day today with near zero clouds, making a perfect curve from dawn till dusk... with just a bit of clipping in the middle of the day. I have 20 450 watt DC panels on Enphase microinverters rated at 380 watts AC. At first when I was researching prior to purchasing my setup I was concerned about the difference in output between the panels and the inverters, but since most days other than spring and fall when it is cooler and still sunny I will not be peaking above 7.5 KW anyway, I went ahead with this array. From what my research has shown, for my area days like today are about the absolute max I will see. I made a graph of actual production today vs what I would have seen without clipping, then I did the math on how much I "missed out on" from the clipping. I missed out on a whopping 0.6 kWh of production over a 2 hour and 15 min timespan. From what I understand, the additional production "off peak" from the larger panels more than makes up for the small amount I loss from the clipping.

I have included the graph of todays production vs a theoretical setup without any clipping.

Later,

Keith Best vs no clip.jpg

[J.P.M.]

Welcome to the forum of few(er) illusions.

For a lot of reasons you may and will probably fairly soon discover on your own, modern silicon PV panels will rarely produce more than about 0.85 +/- about .03 or so of their STC rated output under ideal, real-life conditions. That's about the best you'll get.
For example, I believe I've got a well-designed and well-constructed array that I watch like a hawk and the best quasi steady state rating (inverter output/STC rating) I've been able to achieve is 0.886 * 5.232 kW = 4.635 kW on a very clear, very cool and very windy April day with the array very clean and relatively new (about 11+ years ago).
Backing out the (many times measured) inverter efficiency of 0.9703, means that the average panel output was (4,635W/(16panels*.9703) = 299 W/panel. (Note: 299/327 = 0.914).

While some of that straight line you're seeing on your graph may well be due to clipping, depending on your lat./long. and array orientation, it may be also/partly be the due to the usually quite steady irradiance input to the array as well as the other local environmental conditions which may also be somewhat constant as is usual about that time of day under clear skies.
Happens to my array all the time.

For example, my array has an STC rating of (327W/panel *16 panels) = 5.232 kW.
The string inverter's maximum measured output has been 5.032 kW under non-steady-state operation (partly cloudy sky conditions with added irradiance from cloud albedo), so there may have been some clipping under those conditions.
However, under quasi-steady-state (clear skies around solar noon for an equator facing array) input conditions, my inverter is probably a bit oversized and the system output under those conditions has always been quite steady for +/- about 2 hours or so (1 hour either side of min. beam irradiance angle), pretty symmetric and pretty flat even though the inverter hasn't maxed out yet.

So, it may be some clipping you're seeing but I wouldn't be too surprised if pretty constant input conditions also had something to do with it.
For most well-oriented and mostly equator facing arrays without much shading, output can be surprisingly constant around the time of minimum incidence angle.
If a graph shows a very wide band (time-wise) of constant output, that may be, and often is, a smoking gun for clipping, but short(er) straight line periods up to maybe a bit more than 2 hours or so, on a graph may just be normal operation of a well-designed system.
Sometimes more granularity in the output for power output (maybe 3 significant figures) and time (smaller increments) can show a difference.
If tighter granularity shows a very constant output, then it's probably clipping.

[azdave]

Wait a few years for your panel output to degrade a little and then you'll probably see a nice, perfect curve with no clipping.