Combiner Box for House Off Grid? (And a couple more questions...=)

[RonS]

3 Noob questions. I’m in North Florida and considering buying the following:

12 – 250W Used Canadian Solar solar panels
1 – MPP Solar LV2424
8 – CALB 230ah, 3.2v LiFePO4 Cells to create 24V battery

Question 1: I’m thinking about splitting the panels 6 on the East side of the house and 6 on the West, ground install. But: do I need a combiner box since the power will be rather different from both sides as they will be getting diff sun at diff times of day? Another thread talked about Combiners and it does sound like I would need one if I have panels on 2 sides of the house and that seems to be what Will indicates (https://swisscows.com/video/watch?query=will powers youtube combiner box&id=E7D4F3EFBD1BA5ABCF69E7D4F3EFBD1BA5ABCF69).

Question 2: Will the LV2424 controller prevent over-input just in case there's a really good sun and the panels are feeling good or do I need to just do 8 panels (4/4; I'll have to pair them in series since the VOC is only 37.4—again, Will said the VOC needs to be at least 40V)?

Question 3: Is there a way to figure the exact angle for best sun in particular area (like North Florida)? I know the general angle is 30º, but that can't be universal.

I just found the OCPD article posted elsewhere and I think my head is exploding. Does it suggest I need to have heavier wire for a long run to my panels? Unfortunately, since they're not on the roof I have about a 45' or so run.

Any comments on potential system encouraged. I'm new to this.

 

[RonS]

Also found an OCPD article PDF referenced in another thread and I think my head is exploding. It seems to indicate I may need a bigger wire if my panels have a long run to the inverter? Because of the shadow from the house I have to put the panels about 20' from the house and then it's another 25' to the electrical panel. Being in North Florida the coldest it gets is probably 14 degrees (rarely), so that would be a Factor of 1.14, which with a VOC of 37.4 would = 42.6. Since I would have more than 3 pairs of panels in series, the article suggests I would need an OCPD between the panels and the inverter.

The ISC on these is 9, so 9x1.56 = 14, which should be rounded up to 15A? Does that mean I have to use 15A wire for the whole thing?

If I put them all on the same side of the house with 4 pairs in series paralleled, does that mean 9 x 4 x 1.56 = 56A, rounded up to 60A?

Just checking to see if I'm following the article and wondering about the extra AWG. If I'm reading it correctly, 60A requires 6 Gauge 6AWG cable.

Is there a better way of doing this?

 

[Hedges]

Thicker wire for longer run only if voltage drop V = I x R is too high.

Fuse on PV string, the amperage which panel label says, if and only if 3 or more PV strings connect in parallel.

If Voc x number of strings x 1.14 factor > max voltage rating of device, DON'T CONNECT IT.

14 degrees does not mean factor of 1.14
The math uses "temperature coefficient of Voc" and difference between your location's historical record coldest temperature and 25 degrees C (nominal temperature where panel specs are measured.)
That said, 1.14 is pretty reasonable. I use 1.16 to be very conservative. But if it gives too high a voltage you can sharpen your pencil and do the temperature coefficient calculation.

your math for 4 "pairs in series" (series strings) in parallel looks correct. Sure, round up to 60A ampacity. Although it's so close to 55A ampacity of 8 awg I wouldn't worry about it.

I would consider orienting some PV panel pairs toward the morning sun, some toward afternoon. That reduces peak current and increases hours of output. Combiner box not required on account of orientation; strings can just be wired in parallel. But a box is one way to connect fuses. MC4 fuse holders is another. Combining in parallel could use "Y" cables, except connectors only rated for 30A. To support 60A output, last connection could be made in a box. Or, does inverter have two input terminals? Inverter can be the box; I join wires with a split bolt and tape.


I haven't looked up specs of LV2424. Link to data sheet for all components would help.

 

[RonS]

Sorry, I was going by it got down to 16º F here last winter and usually does at least once a year here—I'm just over the boarder from GA, and was using the Factor from the chart which was for lows of 20º-14º F.

I had been considering angling for morning and afternoon particularly from how the shadows play. Neighbor has a big oak tree that covers late afternoon, but my house covers early morning till about 9am.

Attaching the MPP LV2424. The Growatt most similar to the LV2424 also attached. Should I consider biting the bullet on the extra and going 48V?

Interesting, the Growatt Multiplus II suggests using a battery controller before the inverter box...

 

[Hedges]

You can do the math of temperature delta and temperature coefficient of panels so you understand it, rather than just relying on a chart.
Same for voltage drop in wire runs.

If house makes a shadow encroaching from a particular direction you can consider arrangement of panels and orientation of panels to minimize drop in production, and so multiple parallel strings get similar shading. might be able to keep production a bit higher that way.

What instructions for some panels explicitly prohibit is hard shadows for an extended time while rest of string gets direct sun. That puts full current through bypass diodes, and some panels aren't engineered to tolerate it. They mean the diodes to only get used when sun is lower in the sky and off-angle. But 9:00 AM is early, shouldn't be a problem for diodes.


48V is the way to go for larger systems. It is becoming a commodity e.g. server rack batteries. Depending on SCC, it could support a wider range of PV voltage (but for some SCC would mean narrower range.)

 

[RonS]

Hey Hedges,

I really appreciated your patient response to my above questions. I was wondering if I could bother you more.

The panels I have are Canadian Solar (used) 255W; VOC: 37.4 V; VMP: 30.2 V; ISC: 9.0 A; I: 8.43 A

I'm now looking at an Eg4 3000 EHV-48V Inverter / charger. It has a max VOC of 500VDC, and max input Charging Current of 80A (see attached).

If I'm reading it correctly, that would mean I could put all 8 panels in series, the amps from all the panels would stay 8.43A—as they are individually—while voltage is multiplied to 299.2V (37.4V x 8), which is under the Eg4 3000's 500V max? That would allow for a smaller gauge on wire, right?

Or am I miscalculating something... (Dangit, Jim, I'm an artist, not an electrician! )

 

[Hedges]

All correct.

500V / 299.2V = 1.67, plenty of headroom (I want at least 1.16, otherwise do the math more carefully)

"PV array 120VDC - 450VDC" well above the 120V minimum.

"Max Input Current Draw 18A" so one or two 9A Isc strings in parallel could be used.

255W x 8 = 2040W, 2040W / 48V = 42.5A, well under 80A.

9A Isc x 1.56 = 14A ampacity of wire per NEC. That could be 14 awg or larger.
If you use 10 awg it would be OK for 2 parallel 9A strings, for future expansion.

 

[RonS]

Ah, music to my ears. The Eg4 3000 is only $750, and if I needed more than that the next stop is $1300.

I do think I'd want to have the 10AWG for future expansion. If I had a 8AWG I could have more strings since the the amp limit for the inverter/charger is 80A, or should I just figure on a combiner box to string in more?

Thanks for your kindness on this!

 

[Hedges]

16 PV panels, 255W wired 8s2p would put just about 80A into 48V.
If more than 2 strings in parallel, need a fuse per string. Does the next model have two MPPT inputs? If so, a pair of 10 awg per input and still no fuses needed.

I have a pair of 12 awg per string. That gives me flexibility to rearrange in the future (I have multiple GT PV inverters.)

Is the next model 120/240V split phase? The EG4 3000 is single phase, but says you can connect as man as 12 of them 6p2s for split phase. Or 4p3y for 3 phase. So multiple of this model could be good. Also redundancy if one goes down.

 

[RonS]

Ah, it's good to have someone who can speak amps. I thought it was getting close to the 80A limit of the EG4 3000.

The EG4 6500 EX-48 does have 2 MPPTs and can stack up to 6 for multiple phases. The speck sheet says "Dual MPPT’s can be used individually or in parallel with each other," so I think that means they can be dedicated to two different voltage outputs if desired? Getting a bit steep for me to buy more than 1 of the 6500 for now. I could consider 2 of the EG4 3000, but the pennies are starting to squeak... I suppose it might be logical to start with one EG4 3000 and add later if my checkbook gets fatter. =)

 

[RonS]

Oops, did a bit more research on the EG4 6500 and it's a single-phase as well. You have to put both models in parallel to do multiphase. You can put 12 of the EG4 3000 in parallel, you can put 6 of the EG4 6500 in parallel.

Given my budget it would be easier to buy a series of 3000s if I needed to. Unless I'm missing something the only real advantage to the 6500 vs two 3000s is that the 6500 (obviously) can run 500W more continuously and can charge batteries with 120A and the 3000 only 80A. Alternately, it is $200 cheaper than two 3000s, but I think my budget fits the 3000 a bit better.