Additional Solar Capacity

elbolillo

CatDaddy said:

I won’t debate the objective facts either. FACT: Shading definitely affects panels wired in series more than if wired in parallel. https://www.youtube.com/watch?v=ofo1HQyGG8s&t=43s 

I can provide SPICE simulations that refute your "facts" if you're interested. Otherwise, you do you.

The real world is great for me, thanks.

AnOldUR

elbolillo said:
The real world is great for me, thanks.

Me too. I've always believed/read that parallel was best for shade, but this discussion may motivate me to test for myself.

boathealer

What is missing from this discussion are some relative numbers of what "better" is.  Many times (not necessarily in this case) "better" or "best" ends up being, say, and improvement of 0.03 watts.  Yes, it's the best way, but at what cost (24 or 48V systems, converters, etc.) to increase your output by, for example, 0.03W?  I could accomplish that by upgrading my wire size for example.

Not saying this is the case here, but I have seen this happen in my engineering career over and over.  What is "good enough" for your system cost, or implementation complexity?  Don't get lost in the theory.

elbolillo

boathealer said:

What is missing from this discussion are some relative numbers of what "better" is.  Many times (not necessarily in this case) "better" or "best" ends up being, say, and improvement of 0.03 watts.  Yes, it's the best way, but at what cost (24 or 48V systems, converters, etc.) to increase your output by, for example, 0.03W?  I could accomplish that by upgrading my wire size for example.

Not saying this is the case here, but I have seen this happen in my engineering career over and over.  What is "good enough" for your system cost, or implementation complexity?  Don't get lost in the theory.

This was what I was trying to communicate but maybe I missed the boat.

There are so many factors to take into consideration and there are many trade offs and advantages to different ways of doing things

CatDaddy

boathealer said:

What is missing from this discussion are some relative numbers of what "better" is.  Many times (not necessarily in this case) "better" or "best" ends up being, say, and improvement of 0.03 watts.  Yes, it's the best way, but at what cost (24 or 48V systems, converters, etc.) to increase your output by, for example, 0.03W?  I could accomplish that by upgrading my wire size for example.

Not saying this is the case here, but I have seen this happen in my engineering career over and over.  What is "good enough" for your system cost, or implementation complexity?  Don't get lost in the theory.

I'm not lost in the theory.  I will give an example:

Consider a system consisting of three identical panels A, B and C.  Each are 12V, 120W panels (10A) under full illumination. For this example, assume A is getting 100% illumination, B getting 75% illumination, and C getting 50%.

If they are wired in parallel, the resulting output is 120W because only A is contributing power to the system. Panels B & C aren't doing anything because their output voltages are significantly lower than that of A. Panels in parallel do not "share" the load very well unless they are receiving nearly identical illumination, making their output voltages very similar.

If they are wired in series, a rough estimate of the output power is 50% of A+B+C or about 180W. This is because Panel C limits the current to half it's peak value because it is the most shaded. 

If they are wired with individual MPPT controllers you get 120W from A, 90W from B and 60W from C or 270W total.

These are very rough numbers that do not consider secondary effects but they illustrate what I'm saying. One would have to look at the VI curves for the specific panels to get a more refined estimate and take into account I^2R and power conversion losses.

One can easily cherry pick shading scenarios that show the parallel configuration doing better but in general, the series strings will product more power, especially if A, B and C have bypass circuits to short out panels that are mostly shaded.

The one MPPT per panel configuration will be tops in all shading scenarios except for the series string with equal illumination because there are less conversion losses in that scenario.

This is why commercial solar installations use series strings. It reduces the current in the wiring, which reduces the resistive losses in the wiring (I^2R losses), reduces the cost of the wiring (aka "Balance of System") and increases the overall reliability by reducing the number of inverters required.

AnOldUR

elbolillo said:

CatDaddy said
I can provide SPICE simulations that refute your "facts" if you're interested. Otherwise, you do you.

The real world is great for me, thanks.

I watched the video that you posted and it was convincing, but wanted to see for my self.

I have two 50 watt panels on the roof of my Jeep normally wired in parallel that measure about 20 volts at the connection to a Victron 75/15 controller. As would be expected, wired in series there's 40 volts at the controller. I left my T@B in our heated garage and ran 40 feet of 10 gauge wire out to the Jeep. The controller is mounted in the tub, close to the LiFePO4 batteries.

For my experiment, I read the amps going to my batteries through my Victron shunt with (1) both panels exposed, (2) with one panel covered and (3) then with the other panel covered. I did this three times with the panels wired in series and three times with them wired in parallel alternating to allow for changes in exposure during the time it took me to go back and forth.

Surprisingly, after averaging everything out there was little significant difference. Whether wired in series or parallel, covering a panel cut the amps feeding into the batteries roughly in half. My conclusion is that for the way we use solar to charge our campers batteries, series or parallel wiring of multiple panels will give us results too close to make any difference.

The only consideration is that in series the higher voltage means lower amperage making wire size less critical.

boathealer

AnOldUR said:

elbolillo said:

CatDaddy said
I can provide SPICE simulations that refute your "facts" if you're interested. Otherwise, you do you.

The real world is great for me, thanks.

I watched the video that you posted and it was convincing, but wanted to see for my self.

I have two 50 watt panels on the roof of my Jeep normally wired in parallel that measure about 20 volts at the connection to the controller. As would be expected, wired in series there's 40 volts at the controller. I left my T@B in our heated garage and ran 40 feet of 10 gauge wire out to the Jeep. The controller is mounted in the tub, close to the batteries.

For my experiment, I read the amps going to my batteries through my shunt with (1) both panels exposed, (2) with one panel covered and (3) then with the other panel covered. I did this three times with the panels wired in series and three times with them wired in parallel alternating to allow for changes in exposure during the time it took me to go back and forth.

Surprisingly, after averaging everything out there was no significant difference. Whether series or parallel covering a panel cut the amps feeding into the batteries roughly in half. My conclusion is that for the way we use solar to charge our campers batteries, either series or parallel will give us results too close to make any difference.

The only consideration is that in series the higher voltage means lower amperage making wire size less critical.

Thanks for the experiments. This is good info.

From my research, MPPT controllers vary in capability and sophistication.  Some of them will ramp up current from the lower-voltage (shaded) panel in order to maximize the (average) wattage provided to the batteries, when combined with a non-shaded panel(s). 

Likewise having blocking diodes in the panels vs. not can also change what is seen at the far-end of the system in terms of wattage delivered when partially shaded. 

As usual, there are a lot of apples and oranges out there....