Need help with panel array configuration

I have 6x 315W panels currently mounted, they are wired to one 40A charge controller. they are configured in 3 parallel strings each with 2 panels in series. I'm considering adding another 40A charge controller, but the question is how should I reconfigure the panels? Should I make 1 string with 3 in series on each charge controller? or keep one charge controller with 2 strings of 2 in series and only bring over 1 string of 2 in series to the other controller? If i go with the latter, 1 charge controller will be over paneled and limited to 1040W and have 220W less overall than the other config.
The manual for the tracer charge controller says the best efficiency is using 2 in series, and the max is 3 in series.

I guess the real question I'm asking is the charge controller efficiency of going from 2 in series to 3 (having a higher voltage) worse than the 220W I'd be missing out on during peak sun?

Hey BCrippsy38... You should usually follow what they say in your manual, but take a look at this and see if it helps: Figuring out how many panels in series and parallel based on your MPPT.

Usually you put as much as you can in series to reduce wire costs and with an MPPT there are no losses from having them in series. But your situation is unclear to me; that is your post doesn't have enough information to answer it. For example is the controller a PWM or MPPT? What's the battery voltage, what are the amp/volt limits on the controller and panels (or if you're not sure what those are if you provide the model numbers and we can look them up).

Hi svetz, sorry for the lack of info. its a MPPT Tracer 4215N (40A max). Battery / inverter setup is 24V. I think I'm within spec in both scenarios for voltage, solar panel Vmp is 33V and Voc is 41V per panel. the charge controller limit is 138V i think. its just figuring out the efficency differences on the MPPT using 66V (2 in series) vs. 99V (3 in series). and is that efficiency difference larger than 220W i'd be losing in the first scenario.

Also living in NH. The temp co on the panel says -0.40%/degrees C not sure if that is enough to rule out the 3 panels in series or not.

I couldn't find your manual, but the manual for model 4210RN on page 22 had the maximum voltage as 92V and maximum input power at 24V is 800W. As your panel voltage is 41V, putting 3 in series would be to high (123V) for that model. That might be okay if your MPPT is 138V (Not a Tracer, but EPEver has one with that model number and it was 138V); but you need to check the temperature voltage correction factor of your solar panels to be sure (you can find sample calculations in Figuring out how many panels in series and parallel based on your MPPT). Also do the same thing with the current; your post doesn't provide the current from the panels, but I' guessing it's around 8 amps, so two in parallel would be 16. So probably nothing to worry about there, but you should also validate the current correction for temperature too just to be safe.

My mistake, I forgot the b, Tracer4215BN.
Manual says max Voc is 138V or 150V at min operating temperature.
If I assume coldest temp as -30C. Voct = Voc -0.32*(-30) = 40.93 + 9.6 = 50.53V. 3 panels puts me barely over the 150V min operation temp at 151.59V. I would be comfortable with that, southern NH I think those -22F temps are wind chill temps not actual temps.
- Solar Panel Manual
Current from the panel is about 9.5A.

Assuming both configurations are safe to use, how much less efficient is using a higher Pmax of 99V vs. 66V on this particular charge controller? It says 2 in series is best but I can't see how they quantify it.

So you have 6x 315W = 1890W, from page 20 we know the maximum the MPPT can handle at 24V is 1040W; so there's the potential to clip 850W. That's the same regardless of whether there is two or three in series. Page 24 does have a cool efficiency chart, but the thing to pick up on it is that inverter efficiency drops as voltage increases. So 3 in series at 123V is probably another 1% efficiency loss. But that table answers the question "how is my MPPT efficiency affected by the increase in voltage of putting another panel in series".

The real potential loss though is the clipping. How much is that? You could use NREL's SAM to get an actual number. But let's eyeball it....From the graph in How to figure out how much Solar and inverter wattage is needed to charge a battery and knowing the earth rotates 15 degrees/hour then if your panels were at the optimum tilt each day you'd start loosing power at 1040/1890 =.55, watt 55% power the sun's about 35 degrees. So clipping could start at (90-35)/15 as early as 3 hours before solar noon and 3 hours after. Since it ramps up then down it would be roughly half the peak, or 3*850Wh= 2550Wh per day that lost. If the insolation is 5.5, then the maximum would be 1890x5.5 gives a total power of 10395 Wh, so (10395-2550)/10395 = 75%, or a 25% power loss. This is mixing quite a bit of apples and oranges and with other losses and cloudy days probably not this bad.

Is that what you were looking for?

Update: As you already have it hooked up you can tell if the clipping is a problem or not by looking at the actual power generated per day (which includes clipping) and compare it to the what the panels could do theoretically per day.

So.... set it up with 2 groups of 3 panels each, where each group is tied to a separate controller feeding the same battery. Angle each panel array slightly different, one more east, one more west, to gain the most from everything.

thanks for that calculation svetz. that's a good way figure out the clipped loss, in your experience is there any added benefit to clipping to keep max production longer around peak times? if not then i'll get the second controller.

original plan was to get another charge controller, but how many panels in series is the question. if i do 2 panels in series, i'll have 4 panels on one controller and 2 on the other. one controller will be 1260W, 220W over the clip. and the other controller will 630W, well under the max.
If the charge controller efficiency is negligible when 3 panels are in series at around 99V vs. 66V then what Rider is suggesting seems to be most efficient. and i like the idea of tilting them too, just not sure if the ironridge rails have mount to allow that.

BCrippsy38 said:

and i like the idea of tilting them too, just not sure if the ironridge rails have mount to allow that.

Click to expand...

It's a bit of "6 of one, half dozen of another". Tilting allows earlier and later sun, but maybe slightly less watts at mid-day. Aiming them all to the solar noon gives you maximum mid-day watts, but maybe a few less earlier and later. And how far to tilt before it's actualy diminishing total KWH? Experiment time.....

with one controller on a sunny day i was generating around 7.3kWh i think...

Thinking about the tilt.... I have a single 270w panel pointed at solar noon. I was getting my max watts (190) as early as 10am. That kind of implies one panel set to the 10am sun and the other to the 2pm sun. Both would still be getting near max at noon x 2.

Rider said:

Thinking about the tilt.... I have a single 270w panel pointed at solar noon. I was getting my max watts (190) as early as 10am. That kind of implies one panel set to the 10am sun and the other to the 2pm sun. Both would still be getting near max at noon x 2.

Click to expand...

what's the power generation of the panel currently at 2pm? close to max? its an interesting experiment to see how much of a net gain you would see if any.

BCrippsy38 said:

what's the power generation of the panel currently at 2pm? close to max? its an interesting experiment to see how much of a net gain you would see if any.

Click to expand...

Pretty much the same as 10am, 4 hour window. Drops off rapidly after that due to the sun angle. Hard to determine because by noon my battery is fully charged and watts drops off anyway.

BTW, this is in Maryland, so my sun strength is a bit more than the northeast.

Glad to help! Coming up with a back-of-the-envelope-questionable-way-to-calculate-clipping was fun.

SAM's the best way to play with tilt angles and get real clipping numbers. I know for my system I lose on average 10% for having them flat rather than at the optimum tilt (obviously only losing a little in the summer, but more in winter). But flat is good for my roof due to hurricane wind forces.

BCrippsy38 said:

with one controller on a sunny day i was generating around 7.3kWh i think...

Click to expand...

Wow! That's darn close to the clipping number calculated above.

BCrippsy38 said:

...in your experience is there any added benefit to clipping to keep max production longer around peak times?

Click to expand...

If you have fixed panels you'd only ever see max possible production for a couple of hours for a couple of weeks assuming the panels were sparkling clean, no clouds in the sky etc. This article says 1.2 (e.g. a 12 kW panel is okay on a 10kW MPPT/Inverter); by their standard you're overclipped. But really it depends on two things, the cost of the second controller and the importance of that power to you. If you've got all the power you need no worries, if you could use an extra couple of kwh/d then maybe it's worth the price.

@Rider's suggest of east/west requires two controllers, and once you have the second controller you're good to go.
The problem with

... one panel set [perpendicular] to the 10am sun and the other [perpendicular] to the 2pm sun. Both would still be getting near max at noon x 2.

Click to expand...

True at noon, but from the chart you can see the max solar radiance at 8 AM and 4 PM is only a fraction of what it is at noon; they are not the same.

The downside of any tilting east/west in a fixed setup with MPPT is you get the maximum irradiance when there's a minimal amount of atmosphere between you and the sun; so perpendicular at solar noon usually wins. If you can tilt your panels thoughout the day, you can get an extra ~20% power by readjusting them east/west every couple of hours. Other reasons to focus more east or west is if it's common to have morning/afternoon clouds. I've even heard of some die-hards that prefer a little east as it's generally cooler in the morning and the panel's produce more when they're cooler (see Pmax discussions).

svetz said:

True at noon, but from the chart you can see the max solar radiance at 8 AM and 4 PM is only a fraction of what it is at noon; they are not the same.

Click to expand...

Not disputing that at all, just seems by tilting you expand the window of ANY sun gain, direct or angled, and even more so between 10 and 2.

When I get a chance I think I'll put it to the test, and see just how much you gain/loose overall. It may work, it may be a bust.

Rider said:

Not disputing that at all, just seems by tilting you expand the window of ANY sun gain, direct or angled, and even more so between 10 and 2.

When I get a chance I think I'll put it to the test, and see just how much you gain/loose overall. It may work, it may be a bust.

Click to expand...

If you're mean tracking the sun you'll definitely see a gain. But otherwise the best you can do is with the minimum amount of atmosphere between you and the sun. Although based on location, SAM shows changes in azimuth can increase yield, which surprised me and I don't fully understand (suspect it has to do with local weather).

But let me know what you find! I was wondering after that post if the east/west trick might work with PWMs since they're generally less power anyway so a longer exposure might be better.

svetz said:

If you're mean tracking

Click to expand...

No, fixed point. Set to the 10am point, monitor thru the day, then again at noon, then again at 2pm.
I was thinking of not using the SCC at all, just the raw VOC and amps from the panel.

Sounds like a fun experiment! Theoretically it should match up to the chart; but it's a yearly average. It would be cool to see your experimental data over time; it could answer things a lot of weird questions (e.g., When is the atmosphere thickest? How big of an affect does air mass have?)