Overpanel a lot - What's the worst that can happen?

[AgroVenturesPeru]

The vendor may have also been working on the principle that your array probably will never produce full power, and even if it did it would only be at the peak of the day so it may be acceptable to use a lower rated controller that misses out on some peak hour production but will still benefit from the larger array at non-peak sun times with the gain being saving in system costs.

How would it save system costs over the 150/85 mc4 with 3 separate groups of 2s2p?

 

[gnubie]

I'm sure you can get prices and perform the calculations yourself I don't know what prices your vendor offered you on what ever configurations you asked for or the vendor came up with.

 

[AgroVenturesPeru]

The problem I saw with the 150/70 was the 63.66 amps ISC is greater than the absolute max 50 amp input rating by Victron. If you went to 3 panels in series and risked the voltage limit at 147.3 out of 150, then the ISC of 4 in parallel would be 42.44 and squeak under the 50 amp limit. This might work, but a single cold morning might be the end of it. The 150 volt limit has a note (absolute maximum under coldest conditions) so to me, that is just way too close.

The specs of the 250/70 say max pv input current is 35A; so I'm assuming the vendor wants me to go 4s3p.

Still I don't like the idea of being limited from achieving the max potential

 

[AgroVenturesPeru]

Well the 150/85 is a bit cheaper than the 250/70, but I'm sure you knew that, so I'm going to assume the costs savings have something to do with the price of mc4 connectors vs. whatever gauge wire is required of the screw terminals on a 4s3p array. I don't know though. Is this a trick question on your part?

I'm sure you can get prices and perform the calculations yourself I don't know what prices your vendor offered you on what ever configurations you asked for or the vendor came up with.

 

[Hedges]

The specs of the 250/70 say max pv input current is 35A; so I'm assuming the vendor wants me to go 4s3p.

Still I don't like the idea of being limited from achieving the max potential

Do you mean clipping production of PV panels can produce more than the charge controller will deliver?
Aim half your panels at 10:00 AM sun and half at 4:00 PM sun. That will reduce peak current & power to 0.7 as much, but flatten the curve to deliver (almost) as much power over the day. "almost" because sun isn't as intense when lower in the sky; longer path through atmosphere.

But you can still consider a larger charge controller, to support more panels in the future.

 

[Hedges]

Well the 150/85 is a bit cheaper than the 250/70, but I'm sure you knew that, so I'm going to assume the costs savings have something to do with the price of mc4 connectors vs. whatever gauge wire is required of the screw terminals on a 4s3p array. I don't know though. Is this a trick question on your part?

Give a man a fish ...

?

 

[AgroVenturesPeru]

For a few more C notes ... how about 250/100? Then you can do 3 (4?) in series AND stay under the current limit. Overpanel to your heart's delight, which was what this thread was about, anyway.

I can achieve the goal of harnessing the full array potential on a 150/100 mc4, right?

Or...is there some sort of advantage to higher series (voltage)?

 

[Hedges]

If the second number "100" is amps into the battery, then 150/100 and 250/100 could both deliver the same watts.
Haven't checked the rest of the parameters, like max allow panel short circuit current.
It's a matter of determining which panels and SCC fit best. Right now I've got a 140V max charge controller and some panels that would exceed that on a cold day for strings of 2, but would be insufficient voltage if only one panel.

Maybe, maybe not. It depends.

One advantage of more in series is less copper wire. Another, if just 2 strings, is no fuses.
Then there is shading. Will you get partial shading on some panels? Ideally in that case you have separate MPPT for each string.
Strings of multiple angles is fine on one MPPT, but some strings partially shaded can lose the rest of the panels (and diode-bypassed portions of panels) in that string depending on voltage generated by the rest, and algorithm of MPPT.

It is possible that with 2s3p, one partially shaded panel takes 2 entire panels out of production, but 3s2p only loses the shaded portion of a panel.
After someone came to tout the benefits of panel-level optimizers I tested my 9S2P array by covering one panel. Only lost 1/18th of the power.

 

[AgroVenturesPeru]

Give a man a fish ...

?

I need

Give a man a fish ...

?

some tools to go fishing.

Assuming I'm only running the cables from the panels 15ft. With the mc4 controller there will be 3 separate cables coming from the roof. With the TR (screw terminal) controller. There will be one cable but it will need to support ~32-33amps vs. The mc4 lines would only need to support ~22amps. I haven't priced cables, but the difference doesn't sound too big. Is there some other factor I should consider?

Fuses?

 

[AgroVenturesPeru]

If the second number "100" is amps into the battery, then 150/100 and 250/100 could both deliver the same watts.
Haven't checked the rest of the parameters, like max allow panel short circuit current.
It's a matter of determining which panels and SCC fit best. Right now I've got a 140V max charge controller and some panels that would exceed that on a cold day for strings of 2, but would be insufficient voltage if only one panel.

Maybe, maybe not. It depends.

One advantage of more in series is less copper wire. Another, if just 2 strings, is no fuses.
Then there is shading. Will you get partial shading on some panels? Ideally in that case you have separate MPPT for each string.
Strings of multiple angles is fine on one MPPT, but some strings partially shaded can lose the rest of the panels (and diode-bypassed portions of panels) in that string depending on voltage generated by the rest, and algorithm of MPPT.

It is possible that with 2s3p, one partially shaded panel takes 2 entire panels out of production, but 3s2p only loses the shaded portion of a panel.
After someone came to tout the benefits of panel-level optimizers I tested my 9S2P array by covering one panel. Only lost 1/18th of the power.

When you say string what do you mean?

How many strings are in an 4s3p array?

 

[Hedges]

I need

some tools to go fishing.

Assuming I'm only running the cables from the panels 15ft. With the mc4 controller there will be 3 separate cables coming from the roof. With the TR (screw terminal) controller. There will be one cable but it will need to support ~32-33amps vs. The mc4 lines would only need to support ~22amps. I haven't priced cables, but the difference doesn't sound too big. Is there some other factor I should consider?

Fuses?

Here's a table of wire ampacity, assuming bundle of no more than 3 current carrying conductors (other tables allow higher for single conductor in free air). It gives derating for more than 3 wires or for higher temperature.

https://www.usawire-cable.com/pdfs/nec%20ampacities.pdf

If 2 strings of panels, no fuses (or breakers) needed in that part of the circuit. Size wires carrying current from one string at 156% of Isc for a panel.
But it is still convenient to have a breaker or switch, either per string or for the combined wires.

If more than 2 strings, use a fuse/breaker per string, value from label on panel. Select wire with ampacity at least as high as the fuse where it carries current from just one string.

Size wires carrying current for n strings (n = 2 in the first example) at n x 156% of Isc for a panel.

Also run a ground wire to the panel frames and mounting racks. Scuffed wires or cracked panels can electrify the frame with lethal voltage and current. Other NEC tables will tell you what size wire.

Also put appropriate fuses/breakers/switches between battery and charge controller.
We size fuses/breakers 125% of max continuous current so they don't cause a nuisance by tripping. Wire needs to be at least large enough for the fuse rating. (panels got 156% because they can produce higher than rated current, and safety for unfused wires.)

4s3p is 3 parallel strings, each string having 4 panels in series. (I guess its inconvenient "series" and "string" both have an "s")

 

[carlosspicyweiner]

Eh, it's a tradeoff - you can either get multiple cheap controllers so each one isnt operating near it's max capacity so it doesnt cook itself, or spend more per unit and get quality charge controllers that can operate at their max amperage all the time.

You can have a panel array the size of Arizona behind it but the charge controller is only going to draw what it's going to draw whether its got 3 panels its dealing with or 300 paralleled panels. As soon as you hit the max amperage thats it, its not going to draw more.

You waste power this way when the panel is producing more (technically this is called 'clipping') and the word waste doesnt really fit, it's just unused potential power and the only real effect is that the array voltage goes up, thats why you have to pay such close attention to the panels max VOC because unused power makes the voltage go up.

When I get it finished I'm going to be badly overpaneled in the summer because I'm designing for winter conditions and the days are short and the sun is low in the sky so I need to suck down as much power as I can as fast as I can which means a big array, come summer I'll have power coming out my ears.

This is what i've been looking for - can you elaborate on how/how much the voltage will increase with clipping? I had always assumed that "too much clipping" might just mean excess heat put on your CC.

Tinkering with victron 100/50 12v setup- wondering how 4x parallel strings of 70v x 8amps (70x32 = 2240) CC rated at 100v/60amps input. any calculator to work out high your voltage will go for X amount of clipping?

(yep I'm aware of the absurdity/inefficiency, I have dozens of panels lying around to experiment with)

 

[Hedges]

"Clipping" production doesn't heat the SCC. It just draws less current, lets voltage rise higher so greater percentage of PV production leaks back through PV diode. Voc is 100% leaking, zero current drawn out.

 

[Ampster]

any calculator to work out high your voltage will go for X amount of clipping?

Hedges explained it well. I just wanted to clarify voltage is more a function of temperature. Current potential varies with light and actual current varies with the load.pulled by the inverter.