Need PV Array Advice/Thoughts/Input…

[AH Trimble]

I am looking for real world experience, advice, input, and thoughts of the mixing of string azimuths on a single PV array.

I have 3 arrays:

1. 1050w @ 155° (6x175w, 2S3P) : PVs on a utility room roof, catches early morning sun through till well after noon. 100/20 MPPT
2. 2200w @ 205° (9x245w, 3S3P) : PVs on ground mount, catches sun mid-morning through late in the day. 150/70 MPPT
3. 4000w @ 180° (16x250w, 4S4P) : PVs on ground mount, catches sun early morning through late-afternoon. 250/70 MPPT

I have 6 extra PVs matching those on Array #2. It is not feasible for me to mount them on the same azimuth as Array #2. It is actually far easier and more aesthetic for me to mount them on a separate ground mount on the same azimuth as Array #1, but the panels’ specs don’t match and the strings wouldn’t match so I can’t add them to Array #1.

I am thinking about adding them to Array #2 because they are the same panels and there is open room on the combiner and within spec on the MPPT after adding 2 more strings. And I am thinking about going with Array #1’s azimuth which is different than the current PVs on Array #2.

Question: On the same array I would have 3S3P at 205° (current PVs) and the new PVs 3S2P at 155°...how would the overall production work? My thought is the string groups would produce individually; the 205’s as they do now, the 155’s earlier in the day. Would they fight each other degrading overall performance, or basically act independently?

FWIW...I can’t add a new MPPT for the 6 additional PVs.

 

[MisterSandals]

You need to specify the panels specs in order for someone to calculate how combining different panels will work.

Your first array:
1050W / 58V charging = 18.1A pretty much maxed out already.

With existing arrays already 3P, it is very possible that the wiring and MC4 connectors are already near the 30A max rating. More info needed.

Would they fight each other degrading overall performance, or basically act independently?

Assuming you would connect the new with the existing in parallel, the amps would be summed. Voltage would be clipped to the lower.

 

[AH Trimble]

Sorry, I thought that I laid that the PVs were the same spec (actually same panels)....additional 6 PVs = PVs already on Array #2.
The 3S is safely below the MC4 amp ratting.
My question was in regards to Array #2, not mixing different PVs with different specs. I was wondering about installing the additional 6 PVs at a different azimuth than the existing PVs...155° vs 205°...and how that would affect overall production. I am thinking it might act like shading...but only shading 3 strings or 2 strings independently and at different times.

The info I shared about Array #1 was for info only, it doesn't factor into the question. FWIW...it runs at about 60 - 80v normally depending on the time of day, etc.
The info I shared about Array #3 (4S4P) was for info only, it doesn't factor into the question. FWIW...it runs at about 90 - 110v normally depending on the time of day, etc.
I included both of those array specs to show production and azimuth spread of all 3 arrays.

 

[MisterSandals]

My question was in regards to Array #2, not mixing different PVs with different specs.

Again, assuming you are adding your additional 6 panels, in parallel to array #2 as 3S2P (need 3S to keep the voltage the same as existing array), you would technically have a 5P #2 array. Only you will know the potential array amps since its such a well guarded secret.

 

[AH Trimble]

Again, assuming you are adding your additional 6 panels, in parallel to array #2 as 3S2P (need 3S to keep the voltage the same as existing array), you would technically have a 5P #2 array. Only you will know the potential array amps since its such a well guarded secret.

Seems we are on way different pages. There is no guarded secret of anything, especially the amps; the amps don't figure into the question at all. Of course I would keep it 3S as I stated. Of course that keeps the voltage the same, that's what series does. Of course it adds amps due to the additional 2 strings, that what parallel does.
What I am looking for is input of two different azimuths; 205° for the original 3 strings, 155° for the new 2 strings and how those two different azimuths, 155° vs 205°, could potentially affect production. It is a hypothetical or theoretical question...not a mathematical equation search.

 

[MisterSandals]

how those two different azimuths, 155° vs 205°, could potentially affect production.

I answered this earlier:

the amps would be summed. Voltage would be clipped to the lower.

Lets assume the 3S strings are 100V for simplicity (33.3V 7.36A panels)

Regardless of how many you combine in parallel, the array will be 100V.

With the strings' amps summed, lets say your 155° strings produce 5A and your 205° strings produce 3A

5A + 5A + 3A + 3A + 3A = 19A
19A x 100V = 1900W

If 155° is completely shaded and 205° fully illuminated:

0A + 0A + 7A + 7A + 7A = 21A
21A x 100V = 2100W


My concern is about what happens when both arrays are producing a bit, like when sun is at 180°. If all 5 strings produce 6A, you are right at the current limits of what i would expect to be 10AWG wire and 30A MC4 connectors.
If your panels Isc is higher than my guess or your panels produce more that my 6A guess, you will have a fire risk.

sorry i edited my math a bajillion times...

 

[AH Trimble]

Isc = 8.61a when new, these are 8 - 10 yo PVs so they're somewhere in the neighborhood of 90% (+/-). But if using 100% Isc that is still only 8.61a each string...less than 30% of the 30a rating of most MC4 connectors. Over the last 30 days the 205° PVs have never topped 20a (6.7a each string, 77% of Isc). More than likely due to age and a bit of tilt not optimized for winter. The PVs have only hit 101v one day...that is 90% of VOC. Again, more than likely due to age and a bit of tilt not optimized for winter.

Strings are wired with 10awg to the combiner...way overkill. Then 10awg underground from the combiner 40' to the utility room that contains the rest of the equipment. The 10awg is still overkill at 2% voltage loss.

Adding the 6 additional PVs would probably equate to 2/3 of the production of the 9 existing PVs at peak production.
Peak production on the 6 additional PVs would come in at around 13a (+/-)...based on existing production records.
So if all the PVs were at peak production at the same time that would be roughly 32a @ 100v. With 2% voltage loss that calls for 10awg wire, which is what I have. And no single MC4 connection will ever have more than 8.61a flowing through it. So the wiring and connectors are within good margins.

Here is what I am looking for...


In the early morning when the sun is directly out of the SSE, will the production of the existing PVs suffer production loss due to the the proposed PVs facing SSW? Or, will the 2 strings facing SSW just not be producing and the SSE PVs will be just chugging along like they have been doing?

And the opposite situation...will the proposed SSW PVs be able to produce at their full potential, or will they be adversely affected by the SSE facing PVs?

I could see a situation where the SSW PVs might be only producing 50v while the SSE PVs are trying to produce 100v. This is where I am lacking firm knowledge due to lack of experience in this situation. My "guess" would be the whole array could be reduced to 50v, but that doesn't really sound right. What "guess" does sound more likely would be the total array amps would be reduced to the lower amp production of the SSW strings. But I don't know that for sure either. This is where my lack of experience with this particular situation comes into play.

The easy, and probably smartest, answer would be just to add the additional 6 at the same azimuth as the existing PVs but I am trying to make this look as good as possible in the yard for my wife

 

[MisterSandals]

And no single MC4 connection will ever have more than 8.61a flowing through it.

So some kind of combiner box. Any fuses in the combiner box? In and/or out?

I could see a situation where the SSW PVs might be only producing 50v while the SSE PVs are trying to produce 100v. This is where I am lacking firm knowledge due to lack of experience in this situation. My "guess" would be the whole array could be reduced to 50v, but that doesn't really sound right.

This is an insightful question. As it turns out, when a string is producing so little, it mostly behaves (when in parallel) as if it was not there and will not noticeably affect production of the other parallel strings.

The easy, and probably smartest, answer would be just to add the additional 6 at the same azimuth as the existing PVs but I am trying to make this look as good as possible in the yard for my wife

Can you rig up 3 panels (3S) temporarily and add them to your 105° array and see how it affects the output? Seems you have quite a bit of experience/data with the existing array so adding one more parallel string in various shading would be a good experiment. If you get something other that what i thought, please come back and share the knowledge with the collective (i've eaten crow before too).

 

[AH Trimble]

So some kind of combiner box. Any fuses in the combiner box? In and/or out?


This is an insightful question. As it turns out, when a string is producing so little, it mostly behaves (when in parallel) as if it was not there and will not noticeably affect production of the other parallel strings.


Can you rig up 3 panels (3S) temporarily and add them to your 105° array and see how it affects the output? Seems you have quite a bit of experience/data with the existing array so adding one more parallel string in various shading would be a good experiment. If you get something other that what i thought, please come back and share the knowledge with the collective (i've eaten crow before too).

fused per string 15a, 63a circuit breaker for the combined strings, SPD for the box as well.

yup, great idea!!! I can cobble together a temporary mount and lay wire on the ground for a string of 3 PVs and orient them SSE, record for a week. Then shift it to SSW, record for a week. Compare results...allowing for differences in weather.
And yes...return and report.

That is what I was looking for...thinking outside of "my" box. I hadn't even considered doing a temp test like that. Well done!!!!

"Crow" tastes great when it makes you grow That means I should be a genius by now or 900lbs