Solar Panel Config Oddity

[BriteStar]

Is it possible to configure solar panels in such way that the calculated capacity exceeds the nominal rating? I also queried Renogy and they confirmed my math but implied it’s very unusual and made me feel lucky about it! But am still skeptical, so wanna crowdsource the basic physics to you guys. Here’re the specs:
2 x renogy 100W (Voc=21.6V, Isc=6.24A)
1 x renogy 50W (Voc=22.6V, Isc=2.92)
1 x ecoworthy 100W (Voc=23.4V, Isc=9.12A)
The configuration is 2P1S in the order listed, with a resultant output of 44.2V x 12.04A = 532W, which is a whopping 52% higher than the rated 350W panels. Will I ever see anywhere closer to this figure in real life (of course after accounting for conversion efficiency, etc), or is this just euphoria caused by bad math compounded by an even weaker grasp of physics?

 

[rmaddy]

2P1S is just two panels in parallel but you list four panels. Can you please clarify how exactly the four panels are connected?

Also note that panel wattage is Vmp x Imp. Neither Voc nor Isc are used in determining panel wattage.

 

[BriteStar]

2P1S is just two panels in parallel but you list four panels. Can you please clarify how exactly the four panels are connected?

Also note that panel wattage is Vmp x Imp. Neither Voc nor Isc are used in determining panel wattage.

By 2P1S, i meant the first 2 panels are in parallel as are the last 2, then the 2 parallels are configured in series to complete circuit. My understanding from reading Victron’s Wiring Unlimited and consulting Renogy engineers was that you always use Voc and Isc for sizing panels and charge controllers.

 

[BriteStar]

2P1S is just two panels in parallel but you list four panels. Can you please clarify how exactly the four panels are connected?

Also note that panel wattage is Vmp x Imp. Neither Voc nor Isc are used in determining panel wattage.

Just looked up my spec sheets for the panels and the Vmp and Imps are 17.9V/5.72A for the 100W renogy’s, 18.5V/2.71A for the 50w renogy and 19.2V/5.21A for the 100W ecoworthy.
When i recalculate the output this way, I get about 82% of the rated wattage, which seems reasonable for this set up. Perhaps I have the mppt sizing exercise confused with estimating the reconfig wattage(?)

 

[rmaddy]

By 2P1S, i meant the first 2 panels are in parallel as are the last 2, then the 2 parallels are configured in series to complete circuit.

Ok , that’s 2P2S.

My understanding from reading Victron’s Wiring Unlimited and consulting Renogy engineers was that you always use Voc and Isc for sizing panels and charge controllers.

You use Voc against a controller’s max PV input voltage. You use Isc for working out the current for choosing wire. You use panel wattage and battery charge current to work out what size controller (charge current) you need.

But panel wattage is Vmp x Imp.

The big issue is putting the 50W panel and the 100W panel in parallel. The 100W panel will be pulled down to the lower specs of the 50W panel. So effectively you have 18.5V and 5.42A for the 2nd two in parallel.

The first two panels in parallel will be 17.9V and 11.44A.

Now you attempt to put those two mismatched pairs in series. I’m not sure exactly what the results are for mismatched panels in series but I think you end up with 36.4V at 5.42A which is only 200W. Far from the 350W of the separate panels.

 

[BriteStar]

Ok , that’s 2P2S.


You use Voc against a controller’s max PV input voltage. You use Isc for working out the current for choosing wire. You use panel wattage and battery charge current to work out what size controller (charge current) you need.

But panel wattage is Vmp x Imp.

The big issue is putting the 50W panel and the 100W panel in parallel. The 100W panel will be pulled down to the lower specs of the 50W panel. So effectively you have 18.5V and 5.42A for the 2nd two in parallel.

The first two panels in parallel will be 17.9V and 11.44A.

Now you attempt to put those two mismatched pairs in series. I’m not sure exactly what the results are for mismatched panels in series but I think you end up with 36.4V at 5.42A which is only 200W. Far from the 350W of the separate panels.

There’s only one series in that config and 2 parallels, so it seemed intuitive to me to call it “2P1S”, but am oblivious to any conventions, so i stand corrected! Yes, the 50W stands out, but i didn’t wanna drop it all together, since mismatching isn’t an issue when either voltage or current are closely matched to the others. The voltage is comparable to the other 3 panels, so linking it in parallel with the highest Imp/Isc panel would only add to total current without drastically limiting the voltage. You actually end up with 17.9V x 11.44A for first parallel (P1) and 18.5V x 7.92A for P2, yielding a series output of 36.4V x 7.92A = 288W which is ~82% of the rated output of 350W - still much lower than ideal!

 

[rmaddy]

It’s 2S because you put 2 things in series. You count the items, not the connections.

Even with a result of 288W you are clearly not getting anywhere near the 532W you originally asked about. So no oddity or unexpected excessive results unfortunately.

 

[BriteStar]

Ok , that’s 2P2S.


You use Voc against a controller’s max PV input voltage. You use Isc for working out the current for choosing wire. You use panel wattage and battery charge current to work out what size controller (charge current) you need.

But panel wattage is Vmp x Imp.

The big issue is putting the 50W panel and the 100W panel in parallel. The 100W panel will be pulled down to the lower specs of the 50W panel. So effectively you have 18.5V and 5.42A for the 2nd two in parallel.

The first two panels in parallel will be 17.9V and 11.44A.

Now you attempt to put those two mismatched pairs in series. I’m not sure exactly what the results are for mismatched panels in series but I think you end up with 36.4V at 5.42A which is only 200W. Far from the 350W of the separate panels.

Hey, I wanna thank you for sorting out

It’s 2S because you put 2 things in series. You count the items, not the connections.

Even with a result of 288W you are clearly not getting anywhere near the 532W you originally asked about. So no oddity or unexpected excessive results unfortunately.

Yes, and I wanna thank you for helping me sort out that confusion. I had been sizing my MPPT using Voc/Isc and erroneously adopted those same parameters to predict my output. Big thanks!

 

[BriteStar]

It’s 2S because you put 2 things in series. You count the items, not the connections.

Even with a result of 288W you are clearly not getting anywhere near the 532W you originally asked about. So no oddity or unexpected excessive results unfortunately.

Hey, do you know of any app or online tool that can take a set of different panels and can predict the best/optimal configuration for them? Tried Victron, but didn’t find any…

 

[rmaddy]

Hey, do you know of any app or online tool that can take a set of different panels and can predict the best/optimal configuration for them?

I don't know one but there are a few threads on this site about connecting mismatched panels and I've seen some references to links that might be helpful. A little searching here might dig up one of those links.

 

[BriteStar]

Ok. I thought that would be useful tool to build an array, rather than manually iterating through random configs. When you say “mismatched” do you mean of the rated wattage or Vmp/Imp specs? Granted they’re all related by Watt’s law, but how you configure them (series vs parallel) seems to matter more based on how matched any one of these parameters are to the other panels. I’ve read a few papers debunking the myth of excluding panels strictly based on the rated wattage. But I’ll definitely explore this forum for added insight.

 

[rmaddy]

When you say “mismatched” do you mean of the rated wattage or Vmp/Imp specs?

I do mean mismatched Vmp/Imp.

 

[BriteStar]

Both or either?

 

[BriteStar]

Only one need to match to optimize the config as series or parallel without limiting the output.

 

[rmaddy]

Either depending on which is different, how much the difference is, and how the panels are arranged. The right combination has little impact. The wrong combination has a big impact.

 

[BriteStar]

Yes, how they’re configured becomes more critical. That’s why I’m shocked none of the big PV manufacturers have an online tool that can optimize large scale configurations of panels with varying specs. It would appear a simple algorithm but powerful tool when building new or expanding upon existing arrays…

 

[Minimoose]

Hey, do you know of any app or online tool that can take a set of different panels and can predict the best/optimal configuration for them? Tried Victron, but didn’t find any…

Van Life Solar Setup | Wire Solar Panels in Series vs Parallel

Solar panels perform best when configured in the most optimum way for your combination of panels. Not all wiring setups are the same. So how do you decide if you should wire your panels in series or in parallel?

mowgli-adventures.com

At the bottom of this page is a handy little calculator that might help?

 

[BriteStar]

Hi. Thank you for this link. That was along the lines of what I had in mind for a tool, but unfortunately, this calculator doesn’t exhaust all possible permutations of configs to optimize for maximal output/minimal loss. For instance, I could combine my panels in a 2P3S config to yield 311W (54.3Vx5.72A) for an 11% loss, which is not listed by that calculator. Especially, when you’re working with large arrays, you’d need fancier ML-type computation to optimize configs/predict system outputs and losses. I’m almost sure such tools likely exist within the trade, but none accessible for diy folks like us
PS: The reason i can’t configure them with all parallel (for 4% loss listed as the best) is bc i have a 12V system where the victron MPPT is triggered by a voltage differential of >5V bn PV and batt before charging. A parallel configs with Vmp of 17.9V would necessitate my batteries dipping below 12.9V before charging, which seems a wasted solar opportunity!