Different Panels with Different Ratings

Re: Different Panels with Different Ratings

icarus wrote: »

So Bill,,

If I am reading you correctly,,, which may not be the case,,,

I don't think you are following me yet--but your case would be confusing because it is not a symmetrical system. But--never fear--I can make up an answer that will fit.

If I have 5 panels (I do!) on two different parallel strings,,, each string on a home run, and let's say that string #1 puts out 10 amps (2 panels) String #2 puts out 15 amps (3 panels). What you are saying is that I should have a 6 amp fuse BETWEEN every
panel, AND a 20 amp between the panels and the junction of their home runs at the controller?

We can go through, what I think, is the proper thought processes to analyze the system topology and possible current sources and failure modes. Note, I am using Imp (measured) but should probably using Isc--but you did not give me those numbers--but Isc~1.1*Imp (at least for the BP panel)--so it is not a big difference to worry about.

String #1. 10 amp output (no series fuses listed in spec.)... estimate that:

2x 10 amp fuse = 20 amp would be both safe for panels and would not trip in normal usage
3x 10 amp fuse = 30 amps -- and I would round down from that fuse

20A=<XX fuse<30A -- so 20 to 25 amp fusing/breaker in series with string #1 would be appropriate.

Double check wire size--12-14 AWG wire would be the minimum gauge wire to run with a 20 amp fuse.

String #2. 15 amps (no series fuses listed in spec.)... estimate:

2x 15 amp fuse = 30 amp
3x 15 amp fuse = 45 amp

30a=<xx fuse<45 amp

A 30 amp fuse should be fine in series with string #2... Minimum wire gauge should be ~10 AWG.

Now, look at a short in string #1. We have the 10 amps (Isc~11 amps) going into the short, and we have the rest of the energy from the parallel strings coming into through a 20 amp fuse. If there were 2 or more additional strings whose power added up to more than 20 amps--then the fuse would pop and protect the wiring/fire safety of the 10 amp panels in string #1.

In your case, the amount of current from string #2 is only 15 amps (or 1.25*15A=18.75amps with NEC safety margin)--so a 20 amp fuse is redundant in this case. But we are not done with the analysis yet.

And for string #2, there is no source(s) available that is > 30 amps--so its fuse is also redundant in this case. But--we are still not done yet.

Now--there is the question of back-feed through the controller... A typical GT inverter (isolated supply) cannot back-feed in a single fault mode ("guaranteed" by "UL"/Spec./Requirements -- at least for Xantrex's designs).

But you have an off-grid controller--so lets assume that it is fused between the controller and the battery as (assuming is MPPT controller with 25 amp maximum output) and the controller is not "double isolated--such as would be in a GT type inverter--a single fault can short between Batt + and PV + (I am guessing here--but I think I am correct that a typical off-grid MPPT/PWM controller may assume to be able to fault shorted--so I will continue with this assumption through this example):

25 amp * 1.25 = 31.25 amps or 30 amp fuse controller/battery fuse

Lets assume that this controller, if an internal FET/inductor/etc. failed shorted, there would now be 30 amps available from the battery out to the array.

So, now the wiring from the controller to the solar panel combiner box/junction would have 30 amps available. And therefore the wiring must meet a minimum of 30 amps (10 AWG per NEC?--other specs. may allow smaller awg wire).

At the combiner box, there is 30 amps available--that is certainly able to pop the 20 amp fuse on String #1, and may or may not pop the fuse on String #2 (assuming both the controller had an BAT output to PV input short, and one of the panels had a short to ground too).

But, notice that we may have power from both the faulted controller and string #1... So the total current available to string #2:

30 amps (controller fault) + Isc of 11 amps of String #2 = 41 amps

Or more than enough current to pop String #2's 30 amp protection fuse.

So, string #1 fuse would still be appropriate and string #2 fuse would still be needed in this case too because of the controller+string #1 available current.

Anyway, that is how I would look at your setup. My analysis may or or may not meet the NEC version of codes (1.25x 1 or 1.25x1.25 double derating, etc.)--but the idea is to be safe--and the reasonable failure modes (one fault and the consequences--Designs do not need to take into "double faults"--i.e., both the device and the fuse/breaker fail shorted).

I recall (perhaps not clearly) a conversation we had on this site about the requirement of a fusing small arrays as a safety issue. My recollection of that thread was that there was no safety reason that the panels needed to be fused between the panels and the controller.

Of course my major concern is having a safe installation (from fire) especially while the system is unattended.

Now you have me confused,,, an easy thing to do I confess.

To be clear--end recommendations (based on my assumptions above)

• 30 amp fuse between controller Batt + and battery bank (or whatever your manual says)
• 20 amp fuse in string #1
• 30 amp fuse in string #2
• No fuse required from Controller PV input to PV Combiner Box

You can use slight larger fuses (round up to stand fuse size and wire gauge). And you can use smaller fuses--with the risk of false tripping (running a fuse at 1/1.25 or 80% rated current is pretty close to trip point and I would not go any closer--an a bit farther is better--fuses/breakers are designed to be "unreliable" -- i.e., designed to open if there is a problem).

Above fuse recommendations require that wire at each fuse be capable of handling rated (fused) current flow (and obviously, the controller to battery rated current * 1.25 capacity too).

Almost confused myself--and if you have/use real Isc or Series Fuse Rating instead of my guesses--that would only be better.

Also, I had myself convinced that you would not need series fuses for string #1 or #2 -- until I went through the process and realized that a faulted Off-Grid controller can (could?) also back-feed into the solar panel string (something that a double isolated, which is required by code, GT inverter cannot do by design).

I am sure that Solar Guppy and/or Jim/Crewzer (or others) may find some holes and/or suggested improvements in my analysis -- but that is how I would start.

-Bill

PS: There is only one fuse required per panel string--we are not going to assume that this is a fault that can bypass a fuse and require a fuse per panel (three panels in series, three fuses). One fuse is enough to protect one series string.

Re: Different Panels with Different Ratings

icarus wrote: »

On the roof, the two panels come from the one panel J-box into the other j-box, and then joining the #4 stranded. There are no fuses or blocking diodes other than the ones built into the panels themselves.


So the 20 amp and 30 amp string fuses,,, would they go at the panel,,, or at the beginning of the wire run? I think in the previous (mostly forgotten conversation) we were worried about battery current burning up the wire,, not Pv current,,, than would assume that the fuse should be at the j-box where the strings connect. (Fairly easy to do,,, and out of the weather)

Tony,

I am not 100% sure I understand... Is String #1 home run back to the Controller PV connections, and String #2 is home run back to the controller connections?

Or is String #1 run (on 10 awg or what ever wire) run x feet to a combiner j box. And String 2 (on 10 awg or whatever) run y feet to the same combiner j box? Then every thing comes back on one (or is it two) pair of #4 wire back to the controller?

Stranded or solid does not matter with DC or low frequency short AC runs. What ever terminates well and is easy to handle for you (assuming no flexing like a tracker or moveable panel array on ice--then a 10 awg sun rated shop cord would probably be better in the long term).

Lets see if I can do some ASCII pictures:

[FONT=Fixedsys]
  4AWG  30A                 4  AWG   20A 10 AWG
Batt +==*~*==[controller]============*~*------ String #1+
                                   | 30A
                                   ==*~*------ String #2+

Batt -=======[controller]===========-----------String #1-
                                   |
                                    -----------String #2-[/FONT] 

Hopefully, the drawing comes out properly aligned on your computer screen.

But, look at each point above---Every wire is protected from a maximum current flow (not to exceed its rating) by a fuse.

Note, that these wires are assumed to have current sources on both ends.

One short in the middle (say 30 amps from the "left" and 30 amps from the "right") will result in 30 amps at the short--but each path in the wiring only has 30 amps in it--so it is "safe".

Now, the alternate wiring (A):

[FONT=Fixedsys]
  4AWG     30A                  4 AWG   20A 10 AWG
Batt +=====*~*===[controller]===========*~*-- String #1+
                            |           30A
                            ============*~*-- String #2+

Batt -===========[controller]===============--String #1-
                            |
                            ================--String #2-[/FONT] 

And alternate fuse location for PV runs (B):

[FONT=Fixedsys]
           30A                 20A[/FONT][FONT=Fixedsys] 4 AWG   10 AWG
Batt +=====*~*===[controller]==*~*============-- String #1+
                             | 30A
                             ==*~*============-- String #2+

Batt -===========[controller]==================--String #1-
                             |
                             ==================--String #2-[/FONT] 

Either (A) or (B) above is an OK location for the 20/30 amp PV fuses... It is perfectly OK to "Over fuse" a 4 awg wire or put the fuse out at the end at the 4 AWG to 10 AWG transition. The strings max current (by design Isc) is handled by either wire, and the controller+other string excess current is protected by either fuse location.

With solar systems, "extra heavy wire" is not only a good thing for lower line loss/voltage drops... It is also good for "fusing"... A XXX feet of 14 AWG wire may have so much voltage drop that it could not pop a 15 amp fuse.

However, a X AWG that is XXX feet long can still pop a 15 amp fuse handily. (one of the design requirements is to make sure that there is low enough resistance that a short will still cause a fuse to pop--and not cause the entire run of wire to "cook"--and don't make me look up real numbers and do the calculations tonight -- ).

-Bill

Re: Different Panels with Different Ratings

Tony,

Are those little 2/3 panel "wingy things" in series or parallel?

-Bill

Re: Different Panels with Different Ratings

Just being silly--the drawing looked a bit like a lopsided dragon fly to me...

In the end, each panel should have an appropriately rated fuse (based on the panel's spec. for Series Rated Fuse or on ~2xIsc rating, or ~2.2xImp).

In your case, panel wise, you have 5 strings (one panel each string) and you will be protecting each panel/panel wire with a fuse.

The main wire (4 AWG) is heavy enough to be protected by the 30 amp fuse on the battery side and other smaller fuses/limited current from the panels.

Would 5 fuses on 5 panels be overkill--seems like it--but it is the price you pay for reducing risk.

-Bill