What conditions make it ok to have uneven strings of paralleled panels?

DIY
, ,
1

I know the general consensus is ‘don’t do it’ but I have read a couple of posts on the diysolarforum that it can be done but is ‘advanced’ but have not seen the commenters go into it as they don’t want to confuse the newbie questioning.

Yes I know there will be a performance compromise doing it but as long as there is a decent net gain it doesn’t matter as panels are far cheaper than another SCC not to mention easier to add panels to the existing system for my personal setup.

I want to increase the panels on my existing SCC and NOT get another SCC at this time as there is loads more capacity on my SCC before I reach the VOC but I have not been able to get clear responses of how to do it as people are just quick to discourage and say ‘just get another SCC’.

I have 300W already, 3 x 100W in series. What I would like to do is add either one big panel of 5-700W in parallel with the existing 3 in series or add 2x series to give a very close Voc to the existing 3 and then parallel those. Having been running calculations yesterday I realized the latter way of 2x in series would allow much closer Voc because larger panels rarely go above 50V voc I noticed but with series you can easily pick and choose lower ones in series to get a closer match.

I am not able to get clear data on why this wouldn’t work, only words of discouragement without explanation. I have read generally the naysayers claim that one of your strings will produce 0 output but not really explaining why.

If the Voc is very close then why would they produce 0 output and not just use the lowest Voc, which would be very close anyway? I don’t get why an uneven, as in 1 and 3, or 2 and 3, number of strings is any different than even so long as the Vocs are close?

Also, as I wrote, some seem to do it ok in some situations so it can’t be impossible, so what are the criteria to make it work?

Now it is possible for me to get 3x smaller panels to make the strings equal but then the price savings, which is the main motivator for overpanelling, get less and less because strangely, it seems smaller panels cost way more watt for watt than larger, 400W+, panels. So getting 3x smaller panels would end up costing more than an SCC so pointless then. On the other hand 1 or 2 400W+ panels is ironically much cheaper.

Does string 1 pull down string 2 if string 1 is in a poorer shaded spot?
2

The rules are simple for different panels and the key values are not Voc, but Vmp and Imp, maximum power voltage and current.

Strings in parallel need to have their Vmp close to each other but can have different Imp. The string with higher Vmp will run at the lower Vmp generating less power to the ratio of the 2 Vmp. The 2 strings should have the same orientation so they follow the same trend with irradiance.

Panels in series need to have their Imp close to each other but can have different Vmp. The panel with higher Imp will run at the lower Imp generating less power to the ratio of the 2 Imp. The 2 panels should have the same orientation so they follow the same trend with irradiance.

1 Like
3

Try Solar Panel Series & Parallel Calculator - Footprint Hero

Gives you an idea of any losses with different configurations.

Obviously always best to have identical panels.

4

Yes for different panels but the number of panels doesn’t matter then? So one single panel can be paralleled to a string of 3 with no issues as long as Vmp and Imp are obeyed?

Also what does the orientation matter? Actually they won’t be the same because some are flat and fixed while the other new ones must be ground mounted as there is no more space to put them where the others are. They will still be in the same general vicinity, facing the same direction, it is just the others will be ground mounted and I thought I would angle them for more exposure in the duller months, but the won’t be fixed so I could angle them as I wanted.

5

As long as Vmp is similar number of panels does not matter.

Orientation, if one is flat and another angled probably OK, Vmp does not change too much with irradiance. The trouble comes if one is say east facing and one south facing, the two maximum power points may be further apart so you only get one of the strings working at a time.

6

1 thing you seriously need to consider when paralleling strings of different panels.

Series fuse rating of panels.

You dont want the string with the higher Isc to set fire to the string with the lower Isc. Should any bad things happen to the lower Isc panels.

It is best to keep the Vmp, Isc and Imp the same for both strings.

7

I just tried it. Doesn’t give the configuration I want. It only gives all parallel or all series whereas I want a mix of both.

Oh that is good and what I had thought myself just people online just discouraging without reason, probably they don’t even understand themselves just afraid of trying anything out of the box and say it is wrong because it isn’t the orthodox way.

Check this and see if it looks ok if you don’t mind:

Existing 100W panels, currently in series:

Vmp: 19.4

Imp: 5.2A

= 58.1V x3 in series.

Planned new panels:

430W rated each:

Vmp: 32.9

Imp: 13.07

= 65.8 x2 in series.

So the 430W panels, being the higher voltage figure would take the efficiency hit. The optimal output would be 860W. With the voltage drop would get 759. So 12% performance hit. 300 + 759 = 1059W (compared to 1160W full power).

Do those calculations look correct and ok?

8

Can you explain this please so I can check?

9

All quality panels will have the “series fuse rating” on the panel label and datasheet.

10

I just looked and the fuse rating for the single 100W panels in 15A. Is it ok since the Isc for the big panels is 13.86?

If they are in parallel and not series then why would the current of the bigger panels affect the smaller ones? or is it only in case of malfunction?

11

This.

The series fuse rating relates to how much reverse current the panel can handle. Current through the cells, cell interconnects, junction box, cabling and connectors.

12

Ok, but if it was above the fuse would just blow wouldn’t it? but in this case, 13A for the big panels, and 15A max for the small fuses, it is under, so it is ok?

13

I’m aware that the amps of both parallel Imp strings will be additive, so 13.07+5.2 = 18.27A but does that full current also travel through the smaller 300W string in normal operation? I thought it will only be the sum at the Y junction where they connect.

14

Sure in an ideal world.

But fuses suck. They get warm and blow. They go through a couple of hot cycles and blow. They drop voltage and cost output. They cost money (4 fuses and 4 holders for a parallel string)

Best to stick with strings with the same specs.

15

I don’t understand what you are putting here. You are basically saying you can’t depend on fuses? Please explain what the problem is with the setup as I have explained it? It seems the fuses would protect any malfunction so please explain why not if you think that.

16

This is also subtractive.

Your 5,2a string shorts out. It is now subject to 8a of current. Might not blow that strings fuse, but now that string is absorbing 8a of current.

That is going to get that string toasty warm. Enough to blow the fuses? Enough to damage the panel?

17

Footprint Hero doesn’t take system voltage and controller voltage into consideration, so I think it is pretty limited.

I think I received better info from ChatGPT - asking whether 6 in series - 6S -was better than 3S/2S 2P/3P with a 150v Controller and 21.8v Voc panels/Vmp18.18.

ChatGPT reckoned that 6S would exceed 150v on a cold clear morning, blowing controller, and exceed system voltage by too much to get any better efficiency.

The MPPT manual “Best” is 3S.

18

Where did you get 8A from? Wouldn’t it be 13.07A?

Anyway can’t all these issues be circumvented with additional fuses at certain places?

EDIT: Also I am confused by your cynicism towards fuses. I thought they were generally understood as something you can rely on, over something like breakers which may or may not have the same level of efficiency, due to their simplicity. If you are not confident in the ability of fuses then how do you build a system?

19

Limited indeed, and not doing any combination of series-parallel makes it pretty pointless except for the most novice person but they should know how to do the simple calculation for basic series and parallel anyway if they know enough to put in the data to the fields so it makes it pretty redundant. :slight_smile:

Not saying I am an expert but seems rather redundant as per the above.

20

Agreed footprint hero only can calculate series /parallel at one time generally.

If a mix is planned, you have to work out each section at a time then put that section into calculator.

Eg: you plan on 3S string being parallel with a 2P string, simply calculate the two strings first, then those two strings figures go into calculator giving an overall figure for the planned configuration.

Gives a guide of different configurations with losses.

Fairly easy to set up an excel type spreadsheet if you need further configurations, personally I found it useful to give suitable config variants for my own system with different panels.