Production dips at mid-day

[rmk9785e]

I'm experiencing a strange phenomenon with production from my grid-tied solar system. Almost every day, the production makes an M shaped graph is highest around 10AM and then dips to a much lower level around Noon to go back up until 3 PM and drop from there with very few exceptions as depicted in the third image. I seem to be losing out at prime time of day for solar. My installer thinks it may be because of shading but why would shading be an issue when the sun is relatively at a higher level. Is there anything I can do to fix this?Graph1.PNGGraph2.PNGGraph3.PNG

[Mike90250]

Shadow or over-temp fold back..

[rmk9785e]

It is in the 30s and 40s these days in Northern California so over-temp is probably not an issue. Wouldn't the sun be at higher level at mid-day to result in shading? Are there any on-line sites where I can analyze the impact of shading?

[bcroe]

Shading happens, but it should be obvious. A narrow streak can knock out a lot of panels.

Are you running micro inverters? If so, high line voltage could cause some of them to trip
out as production tries to rise, sort of a self limiting thing. This should be easy to check as well.
Bruce Roe

[J.P.M.]

Provided everything is running correctly:

Looking at the two days, to me it looks like cloud patterns are responsible for the output variations as f(time), but without some P.O.A, irradiance data, that'll be hard to confirm. If there was shading from some combination of obstruction and higher solar zenith angle near the winter solstice, that would show up as reduced output in a consistent pattern more than the than the 2 consecutive days' output seems to be showing.

Although it's possible, I kind of doubt that array or panel temps would change as quickly as would be needed to cause that much of a voltage change as the 12/19 graph seems to be showing.

My money's on clouds.

Any nearby PVOutput sites for comparison ?

Any Impp (current) data ?

[sensij]

That looks like SolarEdge data. At a minimum, use the "layouts" view to review what the individual panels are doing during that time. If you have full reporting access, you can get even more detail on the electrical behavior.

Do you have a public site setup for it that you can share?

[NukeEngineer]

Looks like shading. I have similar shaped graphs ever since early Nov. Got trees shading my panels as the sun got lower. Take a look at my public portal and use the playback feature. Dec 21 was my most recent bright-n-sunny day, and you can see the typical Arby's logo in my output graph.

https://monitoringpublic.solaredge.com/solaredge-web/p/site/public?name=NE_Public#/dashboard

[rmk9785e]

That looks like SolarEdge data. At a minimum, use the "layouts" view to review what the individual panels are doing during that time. If you have full reporting access, you can get even more detail on the electrical behavior.

Do you have a public site setup for it that you can share?

I'm trying to setup a public site but it is not a simple process. I only have homeowner level access. I see some mismatch reports. I don't think my installer ever checks what's going on. The layouts view shows shading on panels. There are a few large trees that my HOA wants to replace but the city is not issuing permits so it is in the courts. Meanwhile I lose much production.
Layout.JPG

Looks like shading. I have similar shaped graphs ever since early Nov. Got trees shading my panels as the sun got lower. Take a look at my public portal and use the playback feature. Dec 21 was my most recent bright-n-sunny day, and you can see the typical Arby's logo in my output graph.
https://monitoringpublic.solaredge.c...lic#/dashboard

I thought sun goes higher until Noon and then lower afterwards. Must be a different angle during winter.

[J.P.M.]

I'm trying to setup a public site but it is not a simple process. I only have homeowner level access. I see some mismatch reports. I don't think my installer ever checks what's going on. The layouts view shows shading on panels. There are a few large trees that my HOA wants to replace but the city is not issuing permits so it is in the courts. Meanwhile I lose much production.
Layout.JPG


I thought sun goes higher until Noon and then lower afterwards. Must be a different angle during winter.

You may well have shading but the two graphs included in your first post to this thread sure look like they're due to at least some cloud activity.

As for solar geometry and solar position, and how solar geometry and those angles affect array incidence angles, the solar zenith angle is almost (but not quite) entirely symmetric about solar noon on a daily basis, every day of the year. However, the daily minimum solar zenith angle (the angle between the vertical and the sun) which occurs at solar noon, changes a bit from day to day while maintaining that close symmetry about solar noon. If an array has a true south azimuth, the P.O.A. beam solar incidence angles will show the same symmetry. Arrays with off south azimuths will not show such symmetry.

[rmk9785e]

Attached please find the production since the system was installed a year ago. The difference between the panel producing the maximum and all others is about 326kWh. Is that to be expected or a result of shading over some panels?

[J.P.M.]

Attached please find the production since the system was installed a year ago. The difference between the panel producing the maximum and all others is about 326kWh. Is that to be expected or a result of shading over some panels?

Provided everything is running right and being reported correctly, yes, you likely have some shading going on. BUT, and in addition to that shading, I still maintain that most of the difference in the two graphs you've included is due to clouds and not shading.

[rmk9785e]

You may well have shading but the two graphs included in your first post to this thread sure look like they're due to at least some cloud activity.

As for solar geometry and solar position, and how solar geometry and those angles affect array incidence angles, the solar zenith angle is almost (but not quite) entirely symmetric about solar noon on a daily basis, every day of the year. However, the daily minimum solar zenith angle (the angle between the vertical and the sun) which occurs at solar noon, changes a bit from day to day while maintaining that close symmetry about solar noon. If an array has a true south azimuth, the P.O.A. beam solar incidence angles will show the same symmetry. Arrays with off south azimuths will not show such symmetry.

My installer recorded Azimuth of 160 for my installation.
The M shaped graph is consistent practically every day during the winter months so maybe the angle of sun is low enough during those hours to throw a shade from the tall Sycamore trees.

Provided everything is running right and being reported correctly, yes, you likely have some shading going on. BUT, and in addition to that shading, I still maintain that most of the difference in the two graphs you've included is due to clouds and not shading.

It is possible however according to Weather Underground, it was a clear day on the 19th, the 3rd graph that fluctuated a lot. I will visually check for the source of shade between 10 AM - 2 PM in the coming days.

[SunEagle]

My installer recorded Azimuth of 160 for my installation.
The M shaped graph is consistent practically every day during the winter months so maybe the angle of sun is low enough during those hours to throw a shade from the tall Sycamore trees.



It is possible however according to Weather Underground, it was a clear day on the 19th, the 3rd graph that fluctuated a lot. I will visually check for the source of shade between 10 AM - 2 PM in the coming days.

If it was clear on the 19th and your production fluctuated I would guess something was given you intermittent shading on the panels. Maybe the swaying of those trees keeping the sunlight from hitting the panels for those dip durations.

[J.P.M.]

My installer recorded Azimuth of 160 for my installation.
The M shaped graph is consistent practically every day during the winter months so maybe the angle of sun is low enough during those hours to throw a shade from the tall Sycamore trees.



It is possible however according to Weather Underground, it was a clear day on the 19th, the 3rd graph that fluctuated a lot. I will visually check for the source of shade between 10 AM - 2 PM in the coming days.

You've got winter shading initiated by the sun being lower in the sky in winter with longer shadows being cast and reaching your array as a result. End of story.

[foo1bar]

I thought sun goes higher until Noon and then lower afterwards. Must be a different angle during winter.

It also moves across the sky, which means any trees straight south might cast a shadow that moves across the array (or parts of the array)

The 'playback' feature on the layout view of solaredge can really give a good visual of the shadow(s) moving across the array. You can pick any 15 minute interval, or just hit play to see a "movie" of your production over the day or week.