Panel breaker bigger than manufacturer recommends

stephendv

Hi all,

I have some new panels and they're rated with an Imp of 2.7A and an Isc of 3.3A and I need to put 7 in parallel. The manufacturer recommends a 5A fuse for each panel, but I already have some 8A midnite breakers which I'd like to make use of. So, since this will be 7 in parallel, if one panel were to develop a short circuit, there would be 6 x 3.3A =~ 20A flowing through it. Surely an 8A breaker would be good enough?

What say you?

Cariboocoot

Re: Panel breaker bigger than manufacturer recommends

Look at it this way:

The purpose of the series fuses/breakers is to prevent the other panels from back-feeding too much current through one if it should fail. The Isc (plus a bit) is the most current they are designed to handle.

So with an Isc of 3.3 any two panels could push 6.6 Amps into a shorted third panel. With seven panels you would indeed get 19.8 Amps into one shorted panel which should definitely trip an 8 Amp breaker.

Provided only one panel shorts.

That's where it gets tricky. With that one out if another one goes it can have (5 * 3.3) 16.5 Amps and still the breaker will work. When the third panel shorts the other four will hit it with 13.2 Amps. The 8 Amp breaker still works. Number four goes down and the other three feed it 9.9 Amps. 8 Amp breaker still trips. Now the fifth panel shorts and the other two hit it with 6.6 Amps. Breaker will not protect it.

Sure, it's Panel Armageddon and extremely unlikely to happen. But there's a saving grace: that "plus a bit" mentioned before. The panels normally have a "maximum series fuse" rating. In this case they'd said 5 Amps (3.3 Isc "plus a bit"). As you can see with 5 Amp fuses in place they'd blow all the way down to the last two panels standing.

So what are the chances of such a massive cascade failure of the array? Almost nil. If you're happy with "almost" the 8 Amp breakers will work.

NorthGuy

Re: Panel breaker bigger than manufacturer recommends

But of course, the panel will short when it's a little bit cloudy, so that other panels will not produce their full current, but only 8A - not enough to trip the braker, but enough to put the shorted panel on fire.

I'd use 8A anyway.

stephendv

Re: Panel breaker bigger than manufacturer recommends

Thanks guys. I didn't mention a key part of this install.. that these panels will all go in parallel with an existing 2.8kW array, so there'll be even less chance of an 8A breaker not working.

8A it is

Cariboocoot

Re: Panel breaker bigger than manufacturer recommends

*creek of brain*

Uh, are you sure the new array is compatible with the existing one? Differences in current between strings can usually be handled by circuit protection. Differences in Voltage (especially reverse Voltage) is another matter. We've actually had a case here where several "small" panels in series to match the Voltage of "larger" panels did a bit of reverse-conducting when shaded and created a problem.

Always best to check these things twice. At least.

stephendv

Re: Panel breaker bigger than manufacturer recommends

Cariboocoot wrote: »

*creek of brain*

Uh, are you sure the new array is compatible with the existing one? Differences in current between strings can usually be handled by circuit protection. Differences in Voltage (especially reverse Voltage) is another matter. We've actually had a case here where several "small" panels in series to match the Voltage of "larger" panels did a bit of reverse-conducting when shaded and created a problem.

Well they are compatible voltage wise... not sure about reverse voltages? Existing array is composed of 4 strings of 3 x Yingli poli with Vmp 30.1 so the string Vmp is 90.3 and the new array will be composed of 7 strings of 2 x Sharp NAF micromorph with Vmp 45 each, so 90V for two. Summer evenings will see the yinglis entirely shaded, while the sharp's will be in sun. There will never be a case where the sharp's are shaded.

How and when would reverse voltages be generated?

Cariboocoot

Re: Panel breaker bigger than manufacturer recommends

stephendv wrote: »

Well they are compatible voltage wise... not sure about reverse voltages? Existing array is composed of 4 strings of 3 x Yingli poli with Vmp 30.1 so the string Vmp is 90.3 and the new array will be composed of 7 strings of 2 x Sharp NAF micromorph with Vmp 45 each, so 90V for two. Summer evenings will see the yinglis entirely shaded, while the sharp's will be in sun. There will never be a case where the sharp's are shaded.

How and when would reverse voltages be generated?

That is probably not enough difference to matter.
Trouble sometimes arises when the lower Voltage panels are shaded and the higher Voltage panels are not. The higher Voltage will flow back through the lower and bleed off power. It's a weird thing that only happens under just the wrong conditions.

But in this case I don't think the reverse break-down Voltage of your Yingli panels is low enough to be affected by the Sharps. In the one case where this did arise a blocking diode (which we usually have to advise people not to put in) solved the problem. As I recall, that had one string made up of panels around 17 Vmp against some made up up 60+ Vmp panels (oddballs).

stephendv

Re: Panel breaker bigger than manufacturer recommends

Pheww, that's a relief. Although... the sharps are thin films so they'll have an initial Vmp of 48V which will stabilise to 45V, so for the first few months there'll be a 6V gap between the two arrays.