Troubleshooting my solar strings

[Vectron]

My ~ 6kw rooftop system has been performing strangely ever since it was commissioned by our installer in late 2020. Can you all please share your wisdom about what might be wrong?

We have two identical strings of 9 panels (details below). One is oriented due east, and one is oriented due west. Absolutely no shading. Theoretically, on a clear day, their output ought to be near mirror images around solar noon.

Problem: the west string's power outputs slowly falls to nearly zero by mid-afternoon (~ 3-4 pm), and then instantaneously surges back to nearly full power (often 1-2 kW) and then follows a normal curve until sunset.

My installer blamed this problem on the MPPT design of our original inverter (Darfon H5001), but when I replaced it this week with a new Sol-Ark 15K, the problem persists. See the data below:

1. The two strings' currents rise in lockstep from ~6 am to 10 am, and then fall off in lockstep from ~2-6 pm. That shouldn't be happening, right?
2. The voltage on the east string (i.e., the one whose curves look ~ normal) is actually only about 75-80% of the west string. That would be consistent with missing about 2 panels (7/9 = 77%).

What could explain this? Are the strings perhaps mis-wired?

My theory: perhaps some of the East panels might be mis-wired into the West string. What else could explain this?



I would appreciate your advice. Any explanations or things I should look into?

Thanks so much!

Details:
Panels: Two arrays, oriented due W and due E, each with 9 LG350N1C-V5 panels. (V_mp = 35.3 V). ~18 degree tilt.
Strings: Supposed to be an east string and west string, but maybe this was misconfigured by installer?
RSD: Tigo TS4-A-F
Inverter: Previously Darfon H5001, now replaced with SolArk 15k. Problem persists with both inverters.

 

[BentleyJ]

This is a South facing 2.8 kW array this is a typical afternoon power curve. Not sure exactly why, I always figured its the MPPT algorithm. Been consistent like this for 15 years.

 

[Vectron]

Two things I find especially puzzling from my data:

1. Why do the West and East strings produce such different voltages nearly all day long, despite (purportedly) having the same number of identical panels wired into them?
   
   
2. Why do the currents track so closely between the two strings in the afternoon?

 

[outsider]

The bump in current looks to be about a factor of 10, so it almost looks like it's changing the scale on you. That is, it may be "helpfully" magnifying the shown current by a factor of 10. Which would of course be a lie, because the scale doesn't change on the left axis.

 

[outsider]

Incidentally this could explain why the trouble is seen with both inverters - the problem is coming from the central reporting thing, not from the inverter. What is the agent that is reporting this data?

 

[BentleyJ]

1. Why do the West and East strings produce such different voltages nearly all day long, despite (purportedly) having the same number of identical panels wired into them?
   
   
2. Why do the currents track so closely between the two strings in the afternoon?

Sounds like a great question for @RCinFLA

 

[Vectron]

Incidentally this could explain why the trouble is seen with both inverters - the problem is coming from the central reporting thing, not from the inverter. What is the agent that is reporting this data?

I downloaded the data from the Sol-Ark monitoring system and plotted them myself.

 

[outsider]

Does your neighbor to the west just happen to have a really, really, large magnifying glass set up in his yard that just happens to focus the setting sun onto your west-facing array?