diy solar

diy solar

How to handle Surplus PV Power?

swiftfall

New Member
Joined
Sep 2, 2023
Messages
14
Location
Dallas, TX
I'm trying to wrap my head around what seems like it should be fairly simple, but I'm obviously missing a link somewhere in all of this.

Let's first imagine a "perfect" scenario. All the days are perfectly sunny for exactly 6 hours then the sun just turns off, like a light bulb. My solar panels are perfectly aligned and 100% efficient.

Now imagine I wish to run a load 24 hours a day that pulls 100A during the 6 daylight hours and 1A during the 18 night hours.

With solar panels I need to produce 103Ah for each of the 6 hours they will produce power (100Ah to use immediately, and 3Ah to store for the night).

So I connect the solar panels (via MPPT controller) to a busbar with 103A cables. The battery connects to the busbar with 3A cables. The load connects to the busbar with 100A cables. And I'm good, I think (plus fuses, oversizing cables, etc, but, we're assuming a perfect world here where nothing breaks and everything always works exactly as expected).

EXCEPT... what if the daytime load is only run on weekdays? What happens that first Saturday when the sun comes up dumping 103A of solar power in and I only have a 18Ah battery to consume it? Is there some component I need to drain this surplus power? Do I need to wire it differently (PV directly to the Load, or something)? Or does it just work and I'm overthinking it?
 
Panels will only generate what the MPPT takes from them. Any excess sunlight just turns into heat and is radiated off as it would on any surface.
 
I'm trying to wrap my head around what seems like it should be fairly simple, but I'm obviously missing a link somewhere in all of this.

Let's first imagine a "perfect" scenario. All the days are perfectly sunny for exactly 6 hours then the sun just turns off, like a light bulb. My solar panels are perfectly aligned and 100% efficient.

Now imagine I wish to run a load 24 hours a day that pulls 100A during the 6 daylight hours and 1A during the 18 night hours.

With solar panels I need to produce 103Ah for each of the 6 hours they will produce power (100Ah to use immediately, and 3Ah to store for the night).

So I connect the solar panels (via MPPT controller) to a busbar with 103A cables. The battery connects to the busbar with 3A cables. The load connects to the busbar with 100A cables. And I'm good, I think (plus fuses, oversizing cables, etc, but, we're assuming a perfect world here where nothing breaks and everything always works exactly as expected).

EXCEPT... what if the daytime load is only run on weekdays? What happens that first Saturday when the sun comes up dumping 103A of solar power in and I only have a 18Ah battery to consume it? Is there some component I need to drain this surplus power? Do I need to wire it differently (PV directly to the Load, or something)? Or does it just work and I'm overthinking it?
Sounds like ya need to double the Ah of the bank...
 
Panels will only generate what the MPPT takes from them. Any excess sunlight just turns into heat and is radiated off as it would on any surface.
So the MPPT will determine that the battery is full, stop pulling from the panels, and then it would be as though the panels were just not connected?

If that's the case, what happens when I turn the load on during daylight hours? Like, if the load was off, the MPPT see the full battery and not be pulling power from the panels. But then, if someone turns the load on, does the MPPT see this and immediately start pulling what it needs?

Do I need the battery to act as a buffer (as I think @Supervstech might have been suggesting)? Since I know the load will never be more than 100A and I can pull at 1C with LiFePO4, if I had a 100Ah battery, there would be some buffer there. The load would have a battery to pull from and discharge enough for the MPPT to see that charging was needed and start pulling from the panels again? How much of a buffer do I need? If I doubled it (36Ah) is the MPPT smart enough to not attempt to charge this with 100A? It doesn't seem possible considering the load, the battery, and the MPPT would be directly connected via busbar. Can PV charge an inverter with no battery at all?
 
Yes, many inverters don’t need a battery to run . The panels provide the power it needs to run.
The charge controller will maintain a constant voltage. If there’s a load, current appropriate to keeping the voltage the same will flow as long as it’s within the capabilities of the panels and mppt
. The charge controller doesn’t care whether the current goes to your battery or load; it doesn’t know or care. Within the capabilities of the system ( power available/ impedance presented to its output) it will keep the voltage where the set point is regardless of who is actually drawing the current.
You will find it more helpful to think in terms of power instead of amps, because real-life systems the devices all run at different voltages, so knowing amps alone tells you nothing useful.
 
Yes a battery can act as a buffer. Think no sunlight for xxx days.
 
Yes, many inverters don’t need a battery to run . The panels provide the power it needs to run.
The charge controller will maintain a constant voltage. If there’s a load, current appropriate to keeping the voltage the same will flow as long as it’s within the capabilities of the panels and mppt
I've been warned several times in several places to hook up battery before panels. This seems contrary to that. If the MPPT controller doesn't have a battery to read.... how will it be able to do it's fancy math to determine the proper power point?
 
Yes a battery can act as a buffer. Think no sunlight for xxx days.
A battery CAN act as a buffer. But, in the above scenario do I NEED a buffer. If the MPPT controller just needs time to see the full battery, then a small buffer is enough. If it'll never stop pulling power from the panels, then I guess I need 2 days worth of buffer (which, in this scenario would be 1200Ah, a HUGE difference from the 18Ah I actually need for storage).
 
Most typical MPPT require a battery, connected in the order you mentioned but some fancy all in one systems can function without a battery.

You are correct that once the battery is full the MPPT will stop charging automatically, based on the voltage limit that you set, even if sun is still shining. And if you then turn on a load, the MPPT will pull it from the panels and leave the batteries full.
 
Last edited:
You will find it more helpful to think in terms of power instead of amps, because real-life systems the devices all run at different voltages, so knowing amps alone tells you nothing useful.
I don't think that's true. Amps alone guides cable sizing, right? 1000V and 1A needs to be sized for 1A. 1V and 1000A needs to be sized for 1000A. In both cases, we have 1000W.

That being said, the situation is the same in Watts, it just doesn't help see the cable sizing picture or the battery charging picture. But, here it is....

I will consume, 4800W continuously during the 6 sunny hours. And 48W continuously during the 18 night time hours.

So I need a battery that will store 864Wh. I need to produce 4944W of power continuously through the 6 day light hours (4800 to use immediately, and 144 to store for the night). What happens on the the days where I don't run the day time load and therefore 4800W goes unconsumed? The first 864Wh will charge the battery (though, if we factor Amps back in, it'll be charging it way too fast, I think) and then... what happens with the rest of the power? At any given time the MPPT controller will see 864W (or less) available at the battery. When the 4800W load turns on, does the MPPT controller see this and reengage the panels? Or does the load try to pull 4800W from the battery which it doesn't have? Does it get 864W at once, because that would overload the cable (again bringing Amps back in to the picture since, above, I connected the battery with 3A cables but this is a 18A load, or more?)?
 
Most typical MPPT require a battery, connected in the order you mentioned but some fancy all in one systems can function without a battery.

You are correct that once the battery is full the MPPT will stop charging automatically, based on the voltage limit that you set, even if sun is still shining. And if you then turn on a load, the MPPT will pull it from the panels and leave the batteries full.
This is only the case if I have the "load" connected to the "load" terminals of the MPPT controller, right? If the load is connected directly to the battery (as the MPPT manual suggests for Inverter Loads) then it wouldn't be able to control that, correct?
 
Exactly right, but even if you draw from the battery directly the MPPT will notice shortly and will resume charging the battery/powering the load.

If you power the inverter from battery directly you would no longer be limited by the current limit of the MPPT load terminals but you would no longer have control to turn the load on and off via MPPT Bluetooth or WiFi ap.
 
Exactly right, but even if you draw from the battery directly the MPPT will notice shortly and will resume charging the battery.
Excellent. Okay, so that makes sense. Now how to I deal with the idea that the battery, nor the cabling that connects it, capable of handling the 100A/4800W daytime load?

Ideally, during the day the MPPT controller would power the inverter directly. And then the "left over" power would charge the batteries.
 
But then, if someone turns the load on, does the MPPT see this and immediately start pulling what it needs?
The MPPT only truly sees this exactly if it has communications to the inverter telling it what the load is.

If the MPPT is only reacting to the voltage it sees on the battery there's are many ways it can misbehave and either not provide enough charging current to keep so it keeps cycling the battery down a bit, or provide too much repeatedly going in and out of float or full charge off.

This is generally dealt with by having extra buffer all over the place. If everything was precisely sized as in your examples with a variable load you'd have a hell of a time trying to ensure you ended the solar production day at exactly 100% charge.
 
After some thought, and a few more videos I think it might work like this

PV (150V) -> DC-to-DC (48V) -> Diode #1 -> Load
^-------> MPPT -> Battery -> Diode #2 -----^

Diode #1 prevents the Battery from supplying current to the "wrong side" of the DC-to-DC converter. And Diode #2 prevents the Battery from being charged by the 48V from the DC-to-DC converter.

I'd still have to size the cable from the MPPT to the battery to handle the full potential battery output 18A unless there's some kind of current limiter I can place with Diode #2. Both the DC-to-DC and MPPT controller will only pull the current they need, I think. Basically, with the help from the diodes, we're create two circuits that have the same input and output but with every different paths.

The only other potential issue is, if the Load pulls from the battery during the day there will be excess PV current. As long as the MPPT then pulls that excess and recharges the battery, it'd be okay. But, adding a few extra Ah, which I'd do anyway, to the battery would help prevent that.
 
It's mostly an exercise in understanding anyway. There are no perfectly sunny days. If this was put into the real world, I'd either need Grid-Tie or at least a few hours worth of daytime battery to account for clouds, shorter days of the year, etc, etc. So, in the end, it'd just be a pretty typical setup.
 
So the basic idea is to use a relatively small battery, just to keep the MPPT running. But when the sun is shining, you want to MPPT to power something large from the load terminals?

While it may be capable to send 100A to the battery, the DC load terminals might be limited to 30A. The good news is you can program the battery charging current to just 3 amps if you like.
 
So the basic idea is to use a relatively small battery, just to keep the MPPT running. But when the sun is shining, you want to MPPT to power something large from the load terminals?

While it may be capable to send 100A to the battery, the DC load terminals might be limited to 30A. The good news is you can program the battery charging current to just 3 amps if you like.
Keep in mind... the mppt doesn't SEND power...

Loads DRAW power
The mppt will sit there and allow it to be pulled.

The battery will only pull what it needs and can use.

Unless the battery is pretty big, 30A isn't going to flow into it, just .2C so, a 10Ah battery will pull 20A max, etc...
 
Back
Top