Series VS Parallel Power Difference

[reiters]

I have 12x 445w panels. I have been running them in series/parallel (6x2) connected to grid tie inverter (Growatt 6kW Grid-Tie Inverter | MIN 6000TL-X). They are facing south at 28 degrees tilt which is the "ideal" fixed angle for my location.

I have been unhappy with my production since I hooked them up. I have peaked at 3500w at any one time. With a theoretical limit of 5340 that's only 65% effeciency. I have been running that for a few months. I bought a MPPT meter to test each panel to see if I had a bad one but it was cheap Chinese and released all it's stored smoke instantly when hooked to a single panel. Since the meter didn't work out I decided to try and test using my inverter. I only hooked up one of the series (6 panels) and got 2300w. I then switched to the other series and got 2250w. So why would I lose 1000w when putting in parallel. No shading.

So I decided to do another test and hook them all in series but I can only use 11 panels because of VOC. When I hooked all 12 and measured with the fluke I got 498v so I took a panel out (11 now) and voltage went down to 450ish. The inverter maxes at 500v and I didn't want to risk it if there is a soar flare on a cold day and blow the inverter. Now that I'm using 11 panels in series I am getting 4000w. That's 81% efficiency. Producing more power with one less panel.

All that to ask, why would parallel lose so much power. Can I somehow get that last panel back in so I'm not wasting it. The inverter has a second MPPT input but it would cost too much for more wire for the 445w I would gain.

Second, I have 3 other 6x2 strings on the roof hooked to off-grid. Am I potentially losing that much wattage from them as well? They don't have as high of an input voltage so I'd have to lose 2 panels per inverter to put them in series so I don't think I would gain anything. I typically get up to 3800w from those strings.

If I did put them in series by removing 2 panels per inverter then I would have 6 panels on the roof and one removed from grid tie that are now unused and maybe put them on a second grid tie? They are located all over so lots of wire to use that way. Not sure I'd ever get a return on my investment.

Last question. If the grid tie inverter has 2 inputs, are they 100% independent and won't effect each other. I ask because when I teste each series of 6 panels alone I got 2300w which is 86% efficiency. I would have to run another set of wires but maybe that is the best solution as long as it makes financial sense. I don't want "2 mppt inputs" to just act as a combiner box and gain me nothing.

Thanks!

 

[Hedges]

https://signaturesolar.com/content/documents/GROWATT/1519015-specs.pdf

2 MPPT trackers.
12.5A per tracker.

I estimate your panels put out 10A each.

Connect 6s to one MPPT input,
do not parallel second 6s, instead connect to other MPPT input.

 

[Hedges]

Last question. If the grid tie inverter has 2 inputs, are they 100% independent and won't effect each other. I ask because when I teste each series of 6 panels alone I got 2300w which is 86% efficiency. I would have to run another set of wires but maybe that is the best solution as long as it makes financial sense. I don't want "2 mppt inputs" to just act as a combiner box and gain me nothing.

2 MPPT trackers
1 PV string per tracker.

They are independent. Running the extra set of wires is the best solution and makes financial sense.
You can then expand to 10, possibly 11 PV panels in series. (Check thermal coefficient of Voc and record cold temperature for your location.)

What is max input voltage of our inverter? You said 500V, but you said 6000TL-X for which I read 550V.
I like to leave about 16% headroom above nominal Voc; 10% like you mentioned (450V vs. 500V) would likely be exceeded in freezing weather.

 

[Hedges]

Some inverters allow two MPPT to be paralleled. Read the manual and maybe you can avoid adding wire.

Per code, wire for those two PV string in parallel would need to be sized for 1.56x Isc. Possibly 10 awg meets that, more likely 8 awg. Per code.
For individual strings, 12 awg would comply.

 

[reiters]

I know I threw a lot of info there. I have 3 off-grid inverters, each on 6x2 strings running the house. I have one grid-tie with 12 panels discussed above. I did that because my batteries were filling to quickly and I was throwing away potential so I shut down one of the 4 off-grids and replaced it with a grid-tie. The "retired" off-grid now functions as a standalone 80A battery charger using a generator to feed it for times when the batteries are dead and grid is out. I also run the genset once a month to warm it up and check everything so I do it on a rainy day and push that power into the batteries. I also like having the "extra" off-grid inverter in case one fails during a SHTF time.

The grid-tie has a 550v max input and I'm pushing 545VOC (495v in reality) to it using 11 panels. Adding a new set of wires so I can use both MPPT inputs will run me around $130. I will gain 500w-1000w of solar panel capacity by adding it. I'm paid roughly $0.10/kw I push back. I estimate about a year and a half to break even.

I will probably bite the bullet and run the wires.

My main question is why was I losing so much power when in parallel? Both strings of 6 produce well and roughly the same but when combined I lose 1000w.

 

[Hedges]

The grid-tie has a 550v max input and I'm pushing 545VOC (495v in reality) to it using 11 panels.

That is living dangerously. Cold day and it will exceed max allowed. Do the math, using historical record coldest ever temperature, and shorten the string to be safe.

Adding a new set of wires so I can use both MPPT inputs will run me around $130. I will gain 500w-1000w of solar panel capacity by adding it. I'm paid roughly $0.10/kw I push back. I estimate about a year and a half to break even.

I will probably bite the bullet and run the wires.

I looked in manual to see if two MPPT could be paralleled, but see no mention.
Because that would draw full current, could exceed what wire is allowed per NEC.

Adding another run is the thing to do. Maybe you can get a bargain spool on eBay. I've seen prices all over the map.

Do you have a ground wire from panel frames to inverter? You should. People have been seeing 60Vrms, 0.5A, feeling a shock.

My main question is why was I losing so much power when in parallel? Both strings of 6 produce well and roughly the same but when combined I lose 1000w.

I answered that. Your two strings in parallel are capable of delivering 20A, but one MPPT limits itself to 12.5A

If your strings had significantly different orientation, peak current would be lower for little clipping (if 90 degrees apart) or none (approaching 60 degree acute angle). More hours of production. A bit less total watt hours per day available.

 

[740GLE]

If $130 is breaking the bank and cost recovery is painful, I guess you can’t really complain about lost KWhr production.


When you were initially designing the system what was your thought process to only use on MPPT input vs two? Save the $130 in only running one conductor string vs two?

 

[reiters]

Hedges,
I think I get it now. Because I changed them from parallel to series I raised the voltage and dropped the amps so the single input can now handle the full output of the panels. So 12.5a @ 275v (typical) I get capped at 3437.5w which is right where I was at. Makes total sense. Because I don't actually get to measure the amps and inverter doesn't tell me, I didn't think to use a calculator to troubleshoot. Thanks.

740gle,
The grid-tie wasn't in my original plan. I had a very hard time finding a "mentor" to coach me but never found anyone willing to discuss a setup that wasn't in sales. Hard to ask the forums when you don't know what to ask. I read everything I could get my hands on for months but nothing like hands on for the full picture to become clear. After my first 6 months I observed that my batteries were full early in the solar day and power from there was wasted so I peeled one of the 4 stings off to grid-tie them. I didn't go grid-tie all the way because my whole idea was power security. I had one set of panels that were not receiving great sun because of location so I took them off the roof and build a rack in the back yard for them and made that my grid-tie set. The ones on the roof face east/west because that is what I have to work with and longer days is better than high output in the middle of the day when I have limited battery capacity. I can't buy enough batteries for multiple days. I have a large generator for that. I spent a large amount (to me) of money on the system full well knowing what I was getting. Now I can't keep dumping money into forever tweaking the system without a much shorter payback. $130 isn't a ton of money but I could easily spend $10,000 making it better if I wasn't worried about the cost. I try to learn more and tweak where I can to get the most out of the system I have.

 

[reiters]

I want to throw out a big thanks to you Hedges. I reduced the sting to 10 panels temporarily while I order more wire. I am not looking forward to running the wire. I have a low pitched roof and I'm 6'7" and overweight so it's a long painful trip to the other side of the house. It's ok though. I think I'm gonna run a 2" PVC across and put a box on the other end of the house so I only have to make the trip one last time. I am getting 4279w currently and still climbing but probably close to peaked out. Way better than the 3500w peak I was getting. After adding the new wire I can put the 2 panels back in (445x2) and see where that takes me. It's running at 400v under load and would be 500 VOC so I think im safe until the wire arrives.

 

[Hedges]

Decide whether PVC vs. EMT or other is appropriate considering NEC and fire/fume hazards.
I don't know the location where you are routing it, but toxic fumes from PVC would be a problem indoors.

I ran PVC exposed outside, and inspector rejected horizontal runs. Said it could only be used vertical (e.g. down from a box to under ground.) I replaced with rigid, and repurposed PVC for data and sprinkler control. But exposed to the sun, it expands/contracts a lot (and gets sunburn). That pulled apart couplings, broke a junction box. Maybe if it makes an "S" bend, it can handle that. I've also seen boxes with slip joints to accommodate motion.

Roof work is one of the more dangerous. You might want to enlist someone more spry, if that would keep you safe. I'm skinny so I work on my roof, but strength vs. size and weight of panels is a problem.

 

[reiters]

I dont mind getting ON the roof. It's the crawling in the attic that I don't like. It's maybe 3 ft tall at the peak and filled with ductwork. I hadn't thought about pvc fumes. I'll see how much rigid costs too. No inspector here to reject it but I do want it safe.

 

[Hedges]

Rigid has become incredibly expensive, like $100 for 20' 2" conduit.
Indoors, EMT should be good. Thinner metal, compression or set-screw fittings.
I agree regarding attic (and mine is 4')! Some of my eaves are exposed to the point I can stuff from one end and reach from middle of attic. Porous to rats, too. Need to box up tight when I finally take care of that "deferred maintenance".

My attic has trusswork, less space to fit through. I had a bunch of tongue and groove bamboo left over, so I put that in to walk on.