String Set-up on Sunny Boy inverter and production behaviour?

[scotsmanswan]

Not a Powerwall / Tesla solar question but I know there are plenty of smart folks on here who may have the answer...

Anyway I have noticed that sometimes I see some variable / unexpected production by string and was wondering how much the accurate set up of strings impacts MPP optimization on SMA Sunny Boy inverters?

Some more details:

• My initial system instal was 22 * 275W panels split equally over strings A and B (all SW facing) with SMA Sunny Boy 6.0 inverter and Powerwall 2; I assume this was correctly set up on inverter with initial commissioning
• After about a month 2 * 275W panels were added to string B as system likely underperforming versus what was guaranteed; no change was made to inverter set up
• After about a year we added 6 * 305W panels to string C through installer

What I have observed is that at times sting B (13 panels; originally 11 panels) underperforms sting A (11 panels) in the morning for up to a couple of hours (when all other variables are similar - i.e. good sunlight, no shading, etc) and on closer examination is seems that string B gets stuck in a lower voltage range within the power curve (i.e. a voltage range more consistent with 11 panels) than days when string B performs more normally (i.e. voltage ranges that would theoretically be in excess of what would be expected with 11 panels)

I have contacted my installer who says it's "performing as expected" but I'm not convinced and so would love any advice from the group.

 

[zƬesla]

@scotsmanswan, I have a SunnyBoy 7.0 (2 strings) and 7.7 (3 strings). Each string has 10x 350kW panels. You can see each string graphed below for a mostly sunny day this past week (I1=7.7, I2=7.0 S#=string). We are still getting morning shade from trees which is the large dip you see around 9am. The larger graph shows overall performance.

 

[scotsmanswan]

@z'tesla, thanks for posting your experience. It would seem that while you observe some variability by string (likely to be expected) the MPP voltage optimizes quite quickly for you. However I'm seeing this optimization take some more time on certain days - like you I have some shade in the morning but the attached graph shows two consecutive days where you can see that String B (the longer string in green) remains materially below String A (blue) for a couple of hours after 9am on first day but had no issue second day (as it used a higher voltage during this time).

I'm thinking this may be because the strings are not set up accurately on inverter (i.e. it thinks string B is 11 panels versus the 13 it actually is) but wondered if anyone can confirm whether Sunny Boy inverters actually use this information?

Note the strange dip in String C (grey) is clipping that is occurring during peak hours and is expected given time of year and my system.