Can I mix vertically wall mounted panels with 35 degrees mounted panles on one string? How bad? Scotland UK.

[Guy from Glasgow]

I think i do not have enough panels (4 panels with max working voltage of 29.7V) on my south facing roof for my invertor.
Invertor is Growatt SPH3600TL BL-UP - and invertors working voltage is 120V-450V.
So ATM my installer connected 4 panles from straight East roof with 4 panels from straight south roof.
Power produced seems to drop badly (by factor of 3-4 probably) in the afternoon when E panels are not getting any direct sunshine.

Invertor has 2 MPPTs (120V min voltage).
Question:
I have plenty of space on the south and East facing walls.
So i am thinking of installing a panel or two on each of those walls vertically (flat on the wall) and have 2 strings of 5 or 6 panels each connectd to separate E and S MPPTs?
To be clear: straight S string will have 4 panels mounted at 35 dgree angle and 1 mounted vertically on South wall.
E string will have 4 panels at 35 degree angle roof + 1 string mounted vertically on E facing Wall.

I am in Scotland 56 degrees North so in winter sun is very low at 10 degrees and in summer not higher than 57 degrees on 22 of June.
A coludy day may be a problem - when most of the diffused light is coming down from the scky (almost vertically?)???

Wonder if anyone had to do this before? How bad compared with say 35 degrees south facing roof?
Probably better than 4 straight E + 4 straight S in one string?

Thanks!

 

[Guy from Glasgow]

And another thought - would it be better if i have 2 vertical panels on South wall in parallel connected to 4 on South roof in series?
This might help with degradation when sun is high and vertical panels are about 30-40 degrees to sun direction (as opposed to ideal angle of 90 degress)?
Might work better for me?

 

[MisterSandals]

So ATM my installer connected 4 panles from straight East roof with 4 panels from straight south roof.

How are they connected? 4S2P?

And another thought - would it be better if i have 2 vertical panels on South wall in parallel connected to 4 on South roof in series?

If all the panels are the same, connecting a 4S string in parallel with a 2S string in parallel renders this the same as 2S2P (2 panel output loss).

 

[Guy from Glasgow]

How are they connected? 4S2P?

If all the panels are the same, connecting a 4S string in parallel with a 2S string in parallel renders this the same as 2S2P (2 panel output loss).

Installer connected 4 S + 4 E in series to make one string of 8 on 1 MPPT. Second MPPT is not used ATM. Thats pretty bad i know

 

[MisterSandals]

Installer connected 4 S + 4 E in series to make one string of 8 on 1 MPPT.

So all in series? 8x 29.7Vmp = 237.6V

This seems like an unusual way to configure the array. Strings pointed in different directions (or shaded differently) should be connected in parallel. 4S2P.

What is the voltage range and MPPT range of your SCC?

 

[Guy from Glasgow]

How are they connected? 4S2P?

If all the panels are the same, connecting a 4S string in parallel with a 2S string in parallel renders this the same as 2S2P (2 panel output loss).

sorry - i meant to have 1 string which would consist in 4 panels in series which will be connected in series with with (2 vertical panels in parallel).
So the total voltage for my panels producing arounfd 30 V will be around 150 V.
Te idea is that 2 vertical pannels connected in parallels will allow more current for main 4 panels on the roof - especially when usun is high at 57 degrees up.
Sorry for the quality of the picture:

 

[Guy from Glasgow]

So all in series? 8x 29.7Vmp = 237.6V

This seems like an unusual way to configure the array. Strings pointed in different directions (or shaded differently) should be connected in parallel. 4S2P.

What is the voltage range and MPPT range of your SCC?

Yes one string of 8 in series connected to 1 mptt.
MPTTs voltage range is 120-450V. Starting voltage is also 120V.
Thats my problem. I can't have 4 panels producing 30v each (29.7 at full sun) with my mptt.
But i do not have space for 5th panel on the roof :-(

And panels already installed and invertor installed too. This 8 in series configuration is wrong i know - but invertor would be below minimum working voltage and will probably only work in full sun. Which is very rare in cloudy Scotland :-(

 

[Hedges]

Don't put panels of significantly different orientation in series.
Small differences, go ahead. Optimizers may correct for moderate differences.

Fine, even preferred in my opinion, to parallel strings of different orientation. Must be similar Vmp & Voc, like within maybe 10%, ideally identical.

You could select different panels such that smaller array of different orientation comes out similar voltage. Or of course use separate MPPT if available.

I've got some panels (SunPower 427W) in the 50 Vmp range. There are some from First Solar that are in the 200 Vmp range.

 

[MisterSandals]

one string of 8

i'm only seeing 6 panels in your sketch (its helpful!)

i meant to have 1 string which would consist in 4 panels in series which will be connected in series with with (2 vertical panels in parallel).

If all the panels are the same, that configuration is pretty wasteful.

You have a 4S string in series with a 2P string. If either is shaded it will drag all output down to very low levels.

voltage range is 120-450V.

This is a bad match for your array.

What SCC do you have?

It "looks" like you need a different SCC or array. I cannot see how they can work together based on your descriptions of direction and angles.

 

[Guy from Glasgow]

Don't put panels of significantly different orientation in series.
Small differences, go ahead. Optimizers may correct for moderate differences.

Fine, even preferred in my opinion, to parallel strings of different orientation. Must be similar Vmp & Voc, like within maybe 10%, ideally identical.

You could select different panels such that smaller array of different orientation comes out similar voltage. Or of course use separate MPPT if available.

I've got some panels (SunPower 427W) in the 50 Vmp range. There are some from First Solar that are in the 200 Vmp range.

Thanks Hedges.
1)Did i get it right that mixing a vertical panel (facing straight south) in 5-in-series string with 4 panels in the roof facing straight south) is a 'small difference' and should work just fine?

2)Do you think that connected 2 verical wall panels in parallel and THEN connecting it in series with 4 roof panels will work better by allowing full potential of roof panels to be used? (see picture above if not clear).
(Obviously at the expense of the 2 vertical panels underperforming when they are in good light....)
Or do you think that i do not even need that.

 

[Guy from Glasgow]

i'm only seeing 6 panels in your sketch (its helpful!)

If all the panels are the same, that configuration is pretty wasteful.

You have a 4S string in series with a 2P string. If either is shaded it will drag all output down to very low levels.


This is a bad match for your array.

What SCC do you have?

It "looks" like you need a different SCC or array. I cannot see how they can work together based on your descriptions of direction and angles.

Sorry - i am not clear again:
1)I already have system working as 8 panels in series connected to 1 mptt at 240 V BUT 4 on E and 4 on South roof.
I am stuck with inverter and 8 panels already there.

2) I suggest 2 possible improvements (which would require 2 or 4 extra panels) to my badly configured system:

Improvement variant A:
- On South side I can add 1 panel on vertical wall facing south and have 5 S panels in series on MPPT1 (should work now as 150 V > minimal range of 120V)
- On East side add 1 panel on verical wall facing East and have 5 in series connected to MPTT2
So I will have total of 10 panels this way (2 strings of 5)

Improvement variant B:
Reason/problem with A above:
It is possible that vertical panel as per improvement variant A may be partially 'shaded' when sun is high. This might drag the whole string of 5 4 roof panels down - as this panel will not allow enough current to be generated by 4 roof panels.

So to increase current produced by the vertical panel at high sun i want/consider:
replace this single vertical panel with 2 vertical penels connected in parallel. Then connect the result to 4 roof panels connected in series. With total voltage still 150 V. As per my picture above
Then do the same for East side (2 vertical panels in parallel connected to 4 in series to produce 150V)
The followinf picture is for variant B showing only South side (the same picture can be used for East side)

Questions
Would mprovement A (5 S one of them vertical, the same E)) work well?
If not - would improvement B work well?

 

[MisterSandals]

Sorry - i am not clear again:
1)I already have system working as 8 panels in series connected to 1 mptt at 240 V BUT 4 on E and 4 on South roof.

Not sure why you sent a diagram of something other than how your system is configured???

I have no idea what you have.

 

[Guy from Glasgow]

Not sure why you sent a diagram of something other than how your system is configured???

I have no idea what you have.

What i have is simply 8 panels inseries to 1 mppt.

Problem is 4 of them are on S roof and other 4 are on E roof.

Hope this helps

 

[Hedges]

Thanks Hedges.
1)Did i get it right that mixing a vertical panel (facing straight south) in 5-in-series string with 4 panels in the roof facing straight south) is a 'small difference' and should work just fine?

Differences in azimuth or orientation matter. If sun strikes panels at different angles, current is limited to the lower of the two.

On panel 90 degrees vertical, and the other 35 degrees off horizontal, is a large difference. Approximately 2:1 different if sun is on the horizon, infinity to one when directly overhead.

Put each orientation (or azimuth) on a separate string.

I'm an electrical guy, not so much a sailor or whatever, was too loose with my use of the word "orientation".
When I sail, I don't use a compass (windsurfer.)

 

[Guy from Glasgow]

Differences in azimuth or orientation matter. If sun strikes panels at different angles, current is limited to the lower of the two.

On panel 90 degrees vertical, and the other 35 degrees off horizontal, is a large difference. Approximately 2:1 different if sun is on the horizon, infinity to one when directly overhead.

Put each orientation (or azimuth) on a separate string.

I'm an electrical guy, not so much a sailor or whatever, was too loose with my use of the word "orientation".
When I sail, I don't use a compass (windsurfer.)

Thanks Hedges.
I absolutley agree that simply replacing inverter for one with working voltage below say 100V (for my 120V 4 panels) and using 2 MPPTs is cheapest and most productive thing to do. It will give me max possible power from both sides etc.

But just out of interest i did some calcs for my location (Glasgow) and using wall mounted panel(s) in series with roof panels does not look too bad - certailnly a lot better thatn mixing S facing roof panels with E facing in one string!

You wrote:
<<On panel 90 degrees vertical, and the other 35 degrees off horizontal, is a large difference>
IMHO it depends on where you are. On equtor - certainly.
I am in Glasgow and sun is relatively low here. So a simple calculation gives me the following numbers for power ratios:

Ratio Roof/Wall	Season	Sun angle at midday	Roof panel sun angle	Wall sun angle	Panel Power Roof	panel Power Wall
0.7	Winter	10	45	80	sin(45)=0.7	sin(80)=0.99
1.11	Average (spring/autum or summer morning/evening)	33	68	57	0.87	0.78
2.02	Summer (June)	57	91	33	1.0	0.49

Notes:
1)compared to what I have now (serial 4East = 4 South) it is not too bad.
Obviously with serial 8 (S+E) - when sun is past 12 noon - the East is completely shaded.
On a bright day it is easily factor of 4 difference in illumination between S and E. :-(
So currently my South panels are dragged down by x4 :-(

2)If I use 2 panles on a South Wall in parallel and connect them (as a parallel unit) in series with my 4 panles on the roof:
Winter:
- the wall panels will be dragged down a lot by roof 4 panels (but roof will be working fine) - so i will get a bit more than jsut 4 panles on the roof
Spring/Autumn/average:
- roof at 100%, 2 wall panels at 45% capability (equivalent of 1 panel on 90%)
Summer midday:
Roof 100%; 2 wall panels at 50% each - equivalent of 100% of one panel.

Interesting:
During the whole year with 2 extra panles on the wall (in parallel) i should get basically power very close to the power of 5 panels on the roof.

 

[zanydroid]

What i have is simply 8 panels inseries to 1 mppt.

Problem is 4 of them are on S roof and other 4 are on E roof.

Hope this helps

Can’t swap to a low string voltage inverter?

Grid tie inverters are a dime a dozen.

Other idea is to buy voltage boosting optimizers direct from China so 4S is boosted to enough to activate your MPPT.

I believe there are also higher voltage panels available in same form factor as regular panels. EG half cut wired for +50 to 100% voltage in same size.

 

[SeaGal]

Who configured your system initially? You refer to an installer, but it seems they have not designed it correctly and the workarounds you mention are sub-optimal.

My approach (with or without co-operation of the installer, depending on whose fault the original design was) would be to replace the 8 panels on your roof with 8 panels of higher Vmp than your existing ones (as @zanydroid mentioned). The 29Vmp you mention is too low for your inverter with only 4 in the string, as you noted.

Something like a Trina S+ Vertex (430W) has Vmp of 43V and VoC of 51V. Is similar size to most standard size panels and available in UK for <£100.

Maybe the installer can swap the panels and re-use the others for another project?

 

[Hedges]

using wall mounted panel(s) in series with roof panels does not look too bad - certailnly a lot better thatn mixing S facing roof panels with E facing in one string!

One string of panels on roof in series with panels on wall may not be as bad as one string panels on S roof in series with panels on E roof. But I don't see how your calculation addressed current limited to the panel with less illumination. Can't add power, have to use the lower of two orientations for panels in both conditions, because it sets the current.

I think your table only covers power (current being similar) at mid day. I think their current differs more at some other times.

I would use the daily or monthly power calculations to evaluate parallel strings (doesn't reveal peak power, which may or may not be clipped by inverter.)

 

[Guy from Glasgow]

have to use the lower of two orientations for panels in both conditions, because it sets the current.

Yes. That's what I did if u look in the table
Sun 10 degrees :
limited by 4 roof panels
2 bottom panels produce the same as 1 roof panel;
Total the same as 5 roof panels at 10 degrees.
Sun 33 degrees:
Current Limited by roof panels. The bottom 2 panels produce the same as 1 roof panel. Total power the same as 5 roof panels at 33 degrees.

Sun at 57 degrees.
Current of 2 wall panels together matches Current of 1 roof panel.
Total power = power of 5 roof panels.
All panels at optimum - neither roof panels or wall panels are limiters. Perfect balance.
‐-------------------
So design effectively means:
At 10 degrees losing power of 1.8 of wall panel. They have 1/0.7= 1.4 times more illumination each than roof panels. So could produce 2.8 times more than 1 roof panel in this condition.
Instead the 2 wall panels produce the same as 1 roof as Current is limited.
--
At Sun at 33 degrees the 2 wall panels receive 1/1.1= 0.9 light of roof panels.
So together they could produce 1.8 times more than 1 roof panel. But are limited by them.
---
Sun at 57 degrees is perfect balance.
2 wall panels at 0.5 illumination produce the same Current as roof panels.
‐-----------
What I do not know is:
how badly limited/limiting panels will be affected in terms of overheating at 10 and 33 degrees?
Basically 4 roof panels are on the situation of shaded compared to output of 2 wall panel. This is probably not too healthy for them?
How bad?

 

[zanydroid]

TBH I originally thought you skipped this step but few days back when I was reviewing a response to you, I figured you might have designed this to avoid the bottleneck. If you design things to parallel extra panels with the bottleneck panels (or just all panels) then you should get more output.

That said, this is so complex compared to the other solutions it's hard to want to review it in depth (lots of scratchbook math and looking at squiggly diagrams for something that feels worse at the outset = not motivating. if it was formatted better then it would be less cost, so less resistance to look at it. but spending the time to reformat something that smells like a loser -- weird + opportunity cost. why not spend the reformating time evaluating something that feels more like a winner).

Shading is unhealthy if bypass diodes get activated and there is a high bypass current going through them. If you avoid that with your crazy layering of complexity, then it is fine.