Too many solar panels?

[Prefersdirt]

I am wondering what happens if you have too many solar panels hooked to your system.

Here is my specific scenario:

Using the Growatt SPF 3000TL . What happens if I put 10 250W 24V panels out for the charging the system. Since everything will be 24V, things will be in parallel. While I know these are used and will produce less than 100%, I would rather plan for maximum output for safety (and 200W for actual production value).

First, Amps. 10 panels at 100% production would be over 80A input. Will the Growatt just not accept the extra, should it be made?

Next is Volts. Since the Growatt is limited to 2kW input, what happens if more than 2kW is being generated?

I can order less than 10 of these panels. Since shipping is the same for 1 or 10, I figure I might as well get 10 of them. Or should I just get fewer?

I really do appreciate the help of this forum.

 

[tictag]

Generally speaking, so long as you don't exceed the PV input limits, you can over-panel any modern MPPT-based SCC and it will dynamically control the input power to whatever the output needs. For example, if you were charging a 24V battery at 20A, that's a peak power requirement of 584W (29.2V x 20A = 584W). You could over-panel your array to say 1,000W and your SCC would dynamically manage the input to only use 584W of it i.e. leaving 416W 'on the table' (ignoring conversion efficiencies).

Generally speaking ... but in your case, this unit's specs are confusing (maybe it's me?).

According to here, I have to assume that the "Maximum Solar Charge Current" refers to the maximum charging current it can supply to your battery and not the PV array's Isc (Current Short Circuit) i.e. the maximum current the array can generate. Why? Because 145V x 80A = 11,600W (!) and your maximum PV power is only 2,000W! So assuming battery charge current, that is 80A x 29.2V = 2,336W, which is still more that the maximum PV power specification. This makes no sense to me.

I am certainly confused by this unit's specifications, but let's see if we can answer your questions.

What happens if I put 10 250W 24V panels out for the charging the system. Since everything will be 24V...

To clarify, these are not 24V panels, that's just a marketing term, they are 37.6V panels i.e. Voc = 37.6V.

While I know these are used and will produce less than 100%,

I doubt you will find much difference. Solar cells can usually be relied upon to deliver their stated output for 25+ years, unless they've been physically damaged or mistreated.

First, Amps. 10 panels at 100% production would be over 80A input. Will the Growatt just not accept the extra, should it be made?

10 panels connected in parallel could potentially generate 10 x Imp (8.27A) = 82.7A at Vmp = 30.3V = circa 2,500W. But you would never, ever do this because this array configuration would maximise losses and minimise the efficiency of your SCC.

Next is Volts. Since the Growatt is limited to 2kW input, what happens if more than 2kW is being generated?

This is more difficult to answer because, as already mentioned, in order to deliver its rated charge current, the SCC would need 2,336W on the input (ignoring conversion efficiencies), which is above the maximum PV rating! ? But to try to answer the question, so long as you didn't attempt to draw more than 2,000W on the output and so long as you do not exceed 145V on the input, you should be OK i.e. install as many panels as you like.

For example, with a 145V input limit you could connect up to 3 panels in series (3 x 37.6V = 112.8V, which is OK, but 4 x 37.6V = 150.4V BUST!). You would then connect 3S strings in parallel, so:

• 3S1P (3 panels) = 750W
• 3S2P (6 panels) = 1,500W
• 3S3P (9 panels) = 2,250W (but don't draw more than 2,000W on the output)
• 3S4P (12 panels) = 3,000W (but don't draw more than 2,000W on the output)

You might also want to clarify with Growatt how they intend to deliver the rated charge current into a 24V nominal battery with a 2,000W input restriction.

p.s. To clarify, it is quite common for people to over-panel their SCC i.e. leave power 'on the table', because over-panelling in this way maximises the power you can achieve on cloudy/overcast days. But it's a balance. The more you leave 'on the table' the more power you are wasting on sunny days. A common rule of thumb is to over-panel by around 30%.

Edit: added postscript.

 

[DThames]

I am wondering what happens if you have too many solar panels hooked to your system.

Here is my specific scenario:

Using the Growatt SPF 3000TL . What happens if I put 10 250W 24V panels out for the charging the system. Since everything will be 24V, things will be in parallel. While I know these are used and will produce less than 100%, I would rather plan for maximum output for safety (and 200W for actual production value).

First, Amps. 10 panels at 100% production would be over 80A input. Will the Growatt just not accept the extra, should it be made?

Next is Volts. Since the Growatt is limited to 2kW input, what happens if more than 2kW is being generated?

I can order less than 10 of these panels. Since shipping is the same for 1 or 10, I figure I might as well get 10 of them. Or should I just get fewer?

I really do appreciate the help of this forum.

This is somewhat speculative, but it appears some low end chargers might be thermally protected but are not current limited. So if you over panel one, they will get hot and shut down, get hot and shut down, over and over. I have some of those (cheap PWM type) that say to limit the input watts, normally allowing double for 24v over 12v (those that auto switch to the battery type).

I have 10amp MPPT that no matter what the input (within voltage limits of course), the output to the batter will max out at 10A.

If the panels are not expensive for you and you can store them, get extras. You might want to grow your system or make some other system. I recently took a large spare panel that I had and set up a lighting/battery/solar system at a hunting shack that we visit a few times in the fall.

 

[apctjb]

Generally speaking, so long as you don't exceed the PV input limits, you can over-panel any modern MPPT-based SCC and it will dynamically control the input power to whatever the output needs.

The Voc of the array does should not exceed the maximum voltage limit of the MPPT controller (145V for the Growatt 3000). Agree with @tictag, over paneling (is that a word?) is common and the MPPT will limit the input to not exceed its wattage limit.

 

[boondox]

If your CC can't handle a larger array, get rid of it. We are using the term "over paneled" but is it really? I have more than my CC can handle and what it does for me is get me by on marginal days without running the genny. I don't feel "over paneled" at all.

 

[tictag]

I don't feel "over paneled" at all.

Over-panneling just means that you have more PV watts than needed to generate the SCCs nominal output. For example, if your SCC is rated to output 10A into a 12V nominal battery, its nominal output is 146W (14.6V x 10A = 146W) so adding more than 146W on the input will leave power 'on the table' in ideal circumstances (i.e. STC). Adding a 200W PV array means that you never going to use 54W of it (on the table) but having a 200W array helps you during cloudy/overcast days.

If you have more PV than the nominal output of your SCC, you are over-panelling.

 

[apctjb]

With ongrid systems it is common to "over panel" by 10-15% to account for the fact that actual output from the array is rarely equal to the array nameplate rating due to soiling, temperature derate and other system losses. No sense is having wasted inverter capacity. PV panels are rated at standard test conditions (STC) , what they produce in the real world is typically less...

 

[Hedges]

You can overpanel by 50% without any issues or waste of production capability if you put the strings in two groups, oriented 90 degrees from each other. Trig tells you that presents 1/sqrt(2) = 0.7 times as much area toward the sun, so peak power output is reduced to 0.7 of what you'd get all oriented the same. (more or less according to how off-angle light gets into the panel vs. reflecting off the glass.) You get more power earlier and later in the day, broadening the curve. The panels could be oriented toward 10:00 AM and 4:00 PM sun, for instance, 6 hours and 90 degrees different. The idea could be extended to seasons as well as time of day, and optimized for when you expect more consumption.

A one-axis or two-axis tracker would accomplish the same thing with smaller array, but in most cases buying more panels is more cost effective and practical than trackers.

I used to orient 100% toward 2:00 PM to maximize production during Noon to 6:00 PM peak utility rates. Now that peak rates are 4:00 to 9:00, I've added some oriented later in the day. I plan to add some oriented for morning sun. With PV panels priced around 1/20th of what I originally paid, over paneling to broaden the production curve while staying within my backfeed ceiling is cost effective. (utility rates are 3x to 9x my PV panel cost, depending on time of day.)

Panels are a commodity that come and go. In 3 months you may not find the same model available. When you pay $90 for a 300W panel plus $200 shipping per pallet, you might consider getting several extra. If you break one, or if you want to use a 600V string inverter later or just double your system, now is the time to get more. I got multiples of 8, nominal 36V panels plus a few spares. An inverter I was considering had a fairly narrow MPPT range, and 8 panels fit that well.

 

[Fred S]

This is somewhat speculative, but it appears some low end chargers might be thermally protected but are not current limited. So if you over panel one, they will get hot and shut down, get hot and shut down, over and over. I have some of those (cheap PWM type) that say to limit the input watts, normally allowing double for 24v over 12v (those that auto switch to the battery type).
I have 10amp MPPT that no matter what the input (within voltage limits of course), the output to the batter will max out at 10A.

That's true, PWM type controllers can't limit current, unless they have some kind of protection circuitry built in. So the overpaneling advice is only for true MPPT controllers, as tictag was careful to say. Like your experience, my 40a mppt controller hits a hard ceiling at 40.0 amps charge rate (1.1kw at say 28v), even though the array is 2.1kw (75a at 28v).

 

[boondox]

Over-panneling just means that you have more PV watts than needed to generate the SCCs nominal output. For example, if your SCC is rated to output 10A into a 12V nominal battery, its nominal output is 146W (14.6V x 10A = 146W) so adding more than 146W on the input will leave power 'on the table' in ideal circumstances (i.e. STC). Adding a 200W PV array means that you never going to use 54W of it (on the table) but having a 200W array helps you during cloudy/overcast days. excess

If you have more PV than the nominal output of your SCC, you are over-panelling.

I get that Tic. I am just saying that "over paneled" isn't the best expression because all (off grid) solar system should have some excess capacity to be "properly sized". For one thing you never get full rated power out and then you need overhead for those marginal days.

Any CC that can't safely current limit itself is a POS that will never be part of any system I design. Just what makes sense to me.

 

[gnubie]

It's not just about the controller current limiting. Quality brands will also tell you the absolute current that can be made available to the controller. The controller will be designed to contain the effects of an internal short at that current, but if you can supply 3 times that, all bets are off. A fuse on the PV input with the controller's absolute rating will stop things catching on fire for configurations that can supply more amps than that.

 

[Ampster]

I am wondering what happens if you have too many solar panels hooked to your system.

It is a fairly common scenerio in the grid tie world to have DC to AC ratios in excess of one to one. Grid Tie inverters are mostly MPPT anyway. For example my GT inverter has a DC to AC ratio of 1.5 to 1.
The result is clipping of the typical sine wave like curve and instead a flat top. The benefit is more energy recovery in the morning and afternoons for that particular inverter size. In my case I think the power company did not want me generating more that 4kW at any one time so the installer specified a 3.8 kW inverter with 5.7kW of panels. Every inverter has a limit on how much it can tolerate as earlier posters have mentioned. For off grid inverters there may be an equivalent ratio expressed in DC terms.

 

[Fred S]

It's not just about the controller current limiting. Quality brands will also tell you the absolute current that can be made available to the controller. The controller will be designed to contain the effects of an internal short at that current, but if you can supply 3 times that, all bets are off. A fuse on the PV input with the controller's absolute rating will stop things catching on fire for configurations that can supply more amps than that.

Yes, good point, in my comment I should have included the assumption that the PV is fused. E.g. I am fused at 25a for my array, because my 2S4P setup nearly always results in output of at least twice the amps from the controller than from the array.

So Prefersdirt, the fuse size for the array's input to the controller will depend on the array config, but can never be greater than the wire's ampacity limit.

 

[Hedges]

So Prefersdirt, the fuse size for the array's input to the controller will depend on the array config, but can never be greater than the wire's ampacity limit.

Additionally, if more than two panels or strings in parallel, each string needs its own fuse no larger than specified on label of PV panel. The wire may be 10 AWG good for 30A, and the panel may put out 8 A nominally, but the panel itself is probably specified for 15A or 20A fuse.

So if 10 panels as 1s10p or 2s5p, you should have 10 or 20 fuses so both ends of each panel (or string of panels) is fused.

If you have an MPPT charge controller that accepts higher input voltage (like 140V or 250V) then you can have fewer strings. Some of my systems used to have only two strings in parallel and no fuse was required, but now I have a single larger inverter with several strings; it has fuses in a combiner box.

 

[heirloom hamlet]

Yes, good point, in my comment I should have included the assumption that the PV is fused. E.g. I am fused at 25a for my array, because my 2S4P setup nearly always results in output of at least twice the amps from the controller than from the array.

So Prefersdirt, the fuse size for the array's input to the controller will depend on the array config, but can never be greater than the wire's ampacity limit.

What are the total specs of your 2S4P array that led you to select a 25a fuse?

 

[Fred S]

What are the total specs of your 2S4P array that led you to select a 25a fuse?

I didn't use a scientific approach. At 40 amps delivery to my 24v batteries (the max that my SCC can deliver, and a rare occurrence), I saw at most about 21a draw from the array. So I "tried" a 25a array main fuse. This has never blown, so I never changed it. A 30a fuse would be fine as well.

A more scientific approach would be

• What can the wire accommodate? 40a
• What is the panel Isc (max short circuit amps)? 8.5a
• How many panels in parallel? 4
• (Theoretical) highest amps from the array if there is a short circuit? 34

Fusing for 40a would be "ok" (wire protection), but this would be an example of where I don't want a power source (the panels) to feed into a short circuit and not blow a fuse. The wire would be fine, but I don't want to overheat a component, such as the SCC, or stress the panels with a sustained short.

 

[Hedges]

A fuse of 25A in that case might cause nuisance blowing, but perhaps the charge controller would never draw enough to blow it.
If 21A was the maximum continuous current, then 25% headroom would suggest 26A, round up to 30A. Or your 25A is probably close enough.
But if panel label says 8.5A Isc, that isn't the maximum current. That is just the current under STC. Cold day and direct sun plus extra light reflected or diffused by clouds can make the current higher. Maybe charge controller would make an excursion during it's peak-seeking algorithms.

The goal is not to blow the fuse if the wire has a short, but rather to limit the current (or time at excessive current) to what the wire can handle.
A PV array fuse ought to never blow, even in the event of a short in the wiring. (but fuses for individual strings will blow if that string is shorted.)
But, if the charge controller shorts battery output to the PV wiring and the PV wiring has a short, then the fuse would be expected to blow.

So I would say about 40A would be appropriate, assuming wire ampacity is high enough.

Separately, each PV string ought to have its own fuse.
(both ends of each string, at least in the case of my system where neither is at ground potential.)

Panels aren't stressed by a sustained short. They produce current in the sun, and if open circuit voltage rises until all current produced leaks through the diode that is the PV cell. If shorted, that current flows in the wire. At mpp, most current flows in the wire and a bit leaks through the diode.

 

[DASH]

If you don't wish to damage the battery, don't forget to stay within the range of the Battery Specifications . Batteries have a maximum safe rate of charge. Depending on the battery chemistry, the maximum C rate of charge is generally between .1 to 1 C. I heard that one can set the maximum rate of charge to the battery with some Midnight Solar Controllers. This is not possible for many other solar controllers.

 

[HRTKD]

Instead of a combiner box with fuses/breakers, what are your thoughts on using the MC4 inline fuses? I see Renogy has them in 10, 15, 20 and 30 amp ratings.

I'm looking to control costs on my design and MC4 inline fuses are much less expensive than a combiner box and way easier to install too.

 

[Hedges]

Initially I avoided fuses by going with 1 or 2 strings. Multiple strings to a larger inverter now and I put a bunch of surplus touch-safe DIN holders on rails in a surplus box. Wasn't too much money. I was pulling wire in conduit anyway so this was convenient.

I can see that MC cable all the way from panel to inverter would be convenient place for the in-line fuse holders.

I plan to use MC4 holders for a couple strings that will be paralleled at the end of one wire run.

Note that MC4 fuse holders are not touch-safe, so you should disconnect the MC4 connectors at both ends before opening.

Neither the MC4 holders nor the touch-safe holders I used are for make/break while hot, so a disconnect switch is needed too. I have heavy duty (600 VDC) knife switches, but the newer inverters have rotary switch attached.

Depending on voltage you can also use Square-D QO with 2-pole breakers (one opens positive, the other opens negative PV wire). Square D says OK up to 125 VDC.