Wiring gridtie panels in series and parallel together

[simnonharper1971]

I have 7 panels on an outbuilding roof that i installed a year or so ago. The panels are wired in series and run into a sunnboy 2500 inverter .The system is grid-tied
I have a number of panels (at least 10) left over and would like to use some more .
There is no room left on the roof but i am planning to mount a few more panels on the walls of the outbuilding .
These new locations will be partly shaded during the day, unlike the roof which gets full sun but these panels are left over and id like to put them to use. Because of the partial day shading I dont want to wire them in series with the 7 roof panels as i dont want to reduce the output of the roof when the new panels are in shade which is why i would like to add them in paralle
Does anyone have any experience of connecting up one or two panels in parallel to an existing series installation? Any reason why i can or cannot add panels which would effectively be outputting 35volts to a series string that is outputting over 200volts?

 

[MichaelK]

edited

 

[Hedges]

Series connection is a good way to deal with partial shading. If one panel in series produces near zero current due to shade, the others push current through its bypass diodes. Under those conditions, the PV string acts as if the shaded panel wasn't part of the string.
But all should be same (or almost the same) orientation.

Sunny Boy 2500U is limited to 600Voc (PV Voc adjusted for coldest temperature.) Mine were 480Voc on a typical day.

Where it can be a problem is an extended period of time with full sun striking the other panels orthogonally (maximum current) while the other panel is shaded. Many panels don't have sufficient heatsinking of the diode. But it is OK for a shorter period of time, or when sun is at an oblique angle.

Assuming all panels are same voltage/current, you can make multiple series strings of same length, then connect in parallel. e.g. 7s2p.
It is OK if one string has different orientation from the other.

Sunny Boy 2500US (is that the one you have? is OK with panels of excess current attached, up to the max Isc rating

Hmm, data sheet says, "Max. input current: IPV 13 A DC" but doesn't clarify Isc limit.
Sunny Boy has a diode to protect against reverse polarity; that is where current limit comes from.

What are the specs on your panels? Temperature coefficient of Voc (on data sheet, not panel label)?

 

[Hedges]

Since this a grid-tie system to feed to the power company, your provider could potentially raise criminal charges against you for attempting something like this.

If this is the older Sunny Boy 2500US inverter, it is transformer isolated and nothing he does on PV side will adversely affect the grid.
If it is a newer model and transformerless, there can be some issues.

 

[simnonharper1971]

Thanks Hedges
My Sunnyboy looks as though it can take dual strings which is why i wondered if it was possible to add additional panels in parallel ( ie high voltage /low current on the original string input and low voltage/ high current on the new string input), but it looks as though the inverter wants both strings balanced and the panels of the same orientation so im wondering if maybe i should get individual micro-inverters for the additional panels. i like your thinking regarding turning the panel installation into multiple but balanced series strings and then parallel together. I suspect i wont generate the minimum voltage required to wake the inverter up .
panels are all solarworld sunmodule sw235 poly

 

[simnonharper1971]

Just had a look at the Sunnboy and the databadge just says Sunnyboy 2500 . Its around 8-10years old

 

[Hedges]

Need to know the exact model number. Or a picture (you can link to a picture on line)
My 2500 are 18 years old, yours may be a different model.

If data sheet says 1 mppt, 2 string inputs, the two strings have to be same voltage.
If it says 2 MPPT, 1 (or 2) strings per MPPT, then you can have two strings of different voltages on separate MPPT.
Likely the 2500 has only 1 MPPT. Larger inverters have 2 or 3.

 

[simnonharper1971]

thanks for reply .
apologies .ive given incorrect info. turns out the inverter i'm using is the sunnyboy 1200. I have another sunnyboy2500 that came with the solar package that i dont use. It all came off a twin aspect roof install. I assumed they were both 2500
ive enclosed pics of the 1200 thats in use

 

[Hedges]

The SB 1200 has reduced Voc and Vmp range, so design string/array for that (including adjustment of PV panel Voc spec according to your location's record cold temperature.)

These are the older transformer type.

You have two MPPT! (one per GT PV inverter). So each can have its own array, size optimized for it.

My system (at one point in the past)

 

[simnonharper1971]

So i take it that means i can have a second unbalanced array?
What are your thoughts on shaded panels in series . Is it such an issue as it was? (assuming my panels have diodes?)
Oh and thanks for your time in following up my post with your constructive comments . It is much appreciated.
Judging by other comments on here I thought i was in danger of blowing up the national grid. Dont tell Putin its that easy.

Ive been a gas engineer for 30 years but did have a previous life studying electronic engineering. I dropped out but it does mean i can work my way around tube amps comfortably. But the whole solar thing is a bit of a new and black art to me.
Thankyou again

 

[Hedges]

"Unbalanced" meaning different voltage from first array? If so, feed to a separate inverter so it has its own MPPT.

Some panels are known to burn up in shade, if connected to other panels in full sun. Should only be a problem if other panels aimed at the sun during that time, producing full current. If sun at 45 degree angle, less current.

USSR blew up their grid with the world's largest nuke.
Internet connected Nest thermostats and GT PV inverters should be able to destabilize it.
PV panels feeding DC into the AC line could do bad things to utility transformers. SMA had to convince utilities they could do transformerless grid-tie inverters safely.

Always double-check PV voltage before making connections. Don't touch the PV wires (or meter probe contacts), especially with a high voltage series string. I check voltage between mating connectors before connecting to home-run wire. It may read voltage at first but decay as parasitic capacitance charges up. But if connected to a short, the voltage reading remains so I know to fix it. Don't make/break connections under load.