Theory: connecting 2 solar banks with different times of light

[n4pgw]

Howdy folks,

I'm a newbie here trying to get a little better understanding of how solar energy works in this particular given situation.

I have my theories based on what I remember from what I read, but no experience working on this scale of solar energy. My theories are open for correction. It's the reason I am here.

My model is an OFF GRID home with the roof line running mostly north and south. There are 1000 watts of panels on the roof. The East side of the roof has 500 watts worth, the west side of the roof has a set of 500 watts of panels. As the sun rises, the MORNING sun strikes only the east panels. The MIDDAY sun lights both panels, and the EVENING sun lights only the west panels.

Part I

What are the most efficient ways to connect these panels together?

A: Series
B: Parallel

Part II: Separately

Here are my theories:
A) Connected in series, the dark panels will prevent the lighted panels from working, so I would see the power grow from 0 to 1000 watts starting when the sun first lights both panels. It will drop from 1000 to 0 watts as the sun drops off the east panel.

B) Connected in parallel, I would expect to see the power rise from 0 to 500 watts as the morning sun rises on the east panel. Then, it would rise from 500 to 1000 watts as it adds light to the west panel. Then, it would drop to 500 watts as the east panel becomes shaded, and finally to 0 when the sun sets below the west panel.

Part II: Separately

In this scenario, the east and west panels are separate systems. They each have a solar charge controller, but they share a common load.

How do I tie them both to the same load?

My theories:
A) Parallel: Tie the outputs of the two charge controllers together to feed one 12 volt source.

I believe I would end up with the same results as connecting the panels in series. The charge controller with the most output would supersede the other controller. The controller with the lower output would read the higher voltage of the other controller and think the battery was fully charged and turn itself off.

B) Series: Tie outputs of the two charge controllers together in series to raise the voltage. Feed the combined voltage output of the two into a third controller that charges the battery or feeds the load.

For best results, the third controller would need to be an MPPT type and the two separate controllers would have to be able to work together. For this scenario, I am assuming they are and they work together.

My theory is that as the east system will generate 12.6 to 14.6 volts with the current rising until it peaks at midday. Then, the west system will kick in an additional 12.6-14.6 volts to raise the voltage to as much as 24 to 30 volts peak. The MPPT will drop the voltage and charge the batteries or the load accordingly.

Since my Renogy PWM charge controller varies battery charging voltage to match the condition of the battery, I would guess my above theory is total trash as it would never know what the state of the battery is.


My guess is that the best scenario is connecting the east and west panel systems in parallel and feeding one charge controller.

[solarix]

Connect the two sides in parallel. The voltage seen on the east and west sides will be pretty close to equal. The current (amps) is what mainly changes as the sun moves across. Even when the morning sun is straight on the east panels and producing lots of current, the west panels will come up to almost the same voltage but have very little current. Good MPPT charge controllers can handle this quite well. Yes, two separate controllers would be a bit better, but not worth the serious extra expense. For a systems over 1000W, MPPT controllers are cost effective - don't be penny wise an pound foolish with a PWM controller.

[sensij]

Part I
B: Parallel

Yes, what Solarix said, except that the voltage of the array orientation seeing the sun will be slightly *lower* than the side in shade, because voltage drops as temp increases. However, that voltage difference should not be enough to seriously hurt the overall efficiency of the parallel strings.

Part II: Separately

Here are my theories:
A) Connected in series, the dark panels will prevent the lighted panels from working, so I would see the power grow from 0 to 1000 watts starting when the sun first lights both panels. It will drop from 1000 to 0 watts as the sun drops off the east panel.

It shouldn't go all the way to zero. There are typically bypass diodes in the panels that would allow cells that can't keep up with the current to be skipped, so the section of the array capable of producing would still be generating some output. However, the transition between functioning and bypassed isn't smooth, and intentionally designing the string that way is not a good practice.

Also, the combined power will never produce 1000 W. When the sun is overhead, it is not hitting either side squarely, which will result in lower output. The 500 W rating is achieved only at a very specific condition of direct light with specific spectral distribution at a controlled temperature.

B) Connected in parallel, I would expect to see the power rise from 0 to 500 watts as the morning sun rises on the east panel. Then, it would rise from 500 to 1000 watts as it adds light to the west panel. Then, it would drop to 500 watts as the east panel becomes shaded, and finally to 0 when the sun sets below the west panel.

This is simplistic, but a good start. The west facing panels can produce power in the morning even without direct sun, since there is diffuse irradiance. Neither sub-array is likely to ever sustain 500 W, even in direct sunlight, because they will warm up and power will drop. On cool spring days, you might get close, especially if there is snowfall and you get some reflection from the ground.

My theories:
A) Parallel: Tie the outputs of the two charge controllers together to feed one 12 volt source.

I believe I would end up with the same results as connecting the panels in series. The charge controller with the most output would supersede the other controller. The controller with the lower output would read the higher voltage of the other controller and think the battery was fully charged and turn itself off.

The bulk stage of charging is not voltage controlled, the battery will take as much current as you can provide. Both controllers will be reading the same battery voltage, and should transition from bulk to absorb at the same time. The absorb and float stages are constant voltage, so having more than one controller involved would be OK. There are some controllers that will communicate with each other and work together to best charge the battery, but even without communication, controllers should work reasonably well together in parallel.

B) Series: Tie outputs of the two charge controllers together in series to raise the voltage. Feed the combined voltage output of the two into a third controller that charges the battery or feeds the load.

For best results, the third controller would need to be an MPPT type and the two separate controllers would have to be able to work together. For this scenario, I am assuming they are and they work together.

My theory is that as the east system will generate 12.6 to 14.6 volts with the current rising until it peaks at midday. Then, the west system will kick in an additional 12.6-14.6 volts to raise the voltage to as much as 24 to 30 volts peak. The MPPT will drop the voltage and charge the batteries or the load accordingly.

This is nonsense.

Since my Renogy PWM charge controller varies battery charging voltage to match the condition of the battery, I would guess my above theory is total trash as it would never know what the state of the battery is.

The PWM doesn't vary the battery charging voltage, the voltage is whatever the array sees hooked up to the load of the battery. It switches very rapidly between connecting and disconnecting the PV array to the battery, so charge current is constantly being switched on and off. Some better controllers may smooth this out somewhat. The battery voltage it detects determines what percent of time the connection is made or opened. The controller programming will determine what battery voltage to target in each stage of charging.

My guess is that the best scenario is connecting the east and west panel systems in parallel and feeding one charge controller.

Good guess.

For what its worth, I think having 1000 W of array split up between east and west will give you a more satisfactory charging profile than if you had panels facing south, although there is definitely a hit on overall energy production as the direction faces further away from south. Really, splitting SE/SW is probably the best. By lowering the peak power, but increasing the number of hours that you are able to maintain useful current flow, the battery has a better chance of getting through the absorb stage.

[n4pgw]

Thank you both for your replies. I definitely need to reorient my mind to thinking current over voltage.

I learned a lot here.

This is nonsense.

LOL! I learned enough now to understand why you said that.

Now I have a better understanding of how things work and what all that nonsense means in my Renogy manual.

I'll have more questions soon.

Thank you both again.