Solar Forecast for Adaptive Battery Charging

[hilt_ctn]

Hi ya'll

Does anyone use a Home Assistant solar forecast add-on/integration to dynamically manage their battery charge / discharge level ? I'd like to hear how well it works or doesn't work for you. Allow me to explain.

Like most, I don't get enough sun in winter to charge my battery fully ( need mooaaar panel powa ! ). So I have to limit its discharge. I've actually automated this to shift at different times of the year. For e.g : As the light improves in the warmer months, HA shifts the battery discharge limit to allow more power to be released. As the light decreases in the colder months, HA shifts the battery discharge limit to allow less power to be released. I collected the data manually over a year and set the limits accordingly. Works well.

Buuuut I'm curious. How far could one take this ? And is it even worth it ? For e.g : The next battery shift is 1st Sept. It will shift 5% to allow that amount of extra discharge. Today was a sunny day and the battery hit 100% at 15h30. So there was enough time to squeeze another 5% today and screw Eskom just a bit more.

Some guys out there have done some truly amazing stuff.

So this is just out of curiosity and learning. Interwebs....debate !

 

[Speedster]

I use solcast but that's mainly to determine whether to run my pool pump on a day or not. What's the use case behind changing your battery discharge level? Why not let it run to the set min and tap from the grid for the balance?

 

[hilt_ctn]

I use solcast but that's mainly to determine whether to run my pool pump on a day or not. What's the use case behind changing your battery discharge level? Why not let it run to the set min and tap from the grid for the balance?

I went with the default weather integration to determine if the pool pump runs or not. Rainy = no pumping ! I plan on tweaking this more at some point.

With regard to the battery discharge leveI, that's exactly what I do : The battery ( 14.3kWh ) runs down to the set min and then I tap the grid for the rest.

In the lighter months, the battery min is set down to 50% to keep some capacity for load shedding. During those months, we don't even use all 50% available, but it's there . We run for free. In the darker months it's as high ( or low depending on your perspective ) as 70% as a 30% charge is all the system can muster during these times. It's also like this because the power box down the road that feeds our entire block blows up what feels like every 3 or 4 months. Pseudo load shedding. The last time it shat the bed, that 70% saw us through 29 hours of no power and that included running the geyser.

That set min is exactly what I'm curious about optimising within an inch of it's life. On days like today, that set min could shift dynamically from 70% to say 65% and suck up all that sunny goodness, saving me a whopping R2.45 ! Hey it all adds up.

In truth I'm happy to wait. The set min will auto shift to 65% on 1st Sept, then to 60% on 15th. Assuming more sunny days in the next two weeks, it feels like I'm giving Eskom money for nothing...even if it is a pittance.

 

[B-1]

I integrated solcast but haven't used it. I have enough on sunny/semi cloudy days so what I've settled on is that I manage my geysers based on the battery level.

So if the battery isn't at x level I don't switch on the guest geyser and then if it isn't at y I don't switch on the main geyser. It works well for the most part but I have had a cold shower or two if I wasn't paying attention and its been raining for a few days. If I am paying attention I just switch the geyser on from my phone when I wake up and its warm by the time I shower.

 

[SauRoNZA]

I went with the default weather integration to determine if the pool pump runs or not. Rainy = no pumping ! I plan on tweaking this more at some point.

With regard to the battery discharge leveI, that's exactly what I do : The battery ( 14.3kWh ) runs down to the set min and then I tap the grid for the rest.

In the lighter months, the battery min is set down to 50% to keep some capacity for load shedding. During those months, we don't even use all 50% available, but it's there . We run for free. In the darker months it's as high ( or low depending on your perspective ) as 70% as a 30% charge is all the system can muster during these times. It's also like this because the power box down the road that feeds our entire block blows up what feels like every 3 or 4 months. Pseudo load shedding. The last time it shat the bed, that 70% saw us through 29 hours of no power and that included running the geyser.

That set min is exactly what I'm curious about optimising within an inch of it's life. On days like today, that set min could shift dynamically from 70% to say 65% and suck up all that sunny goodness, saving me a whopping R2.45 ! Hey it all adds up.

In truth I'm happy to wait. The set min will auto shift to 65% on 1st Sept, then to 60% on 15th. Assuming more sunny days in the next two weeks, it feels like I'm giving Eskom money for nothing...even if it is a pittance.

But battery percentage and discharge rate are two entirely different things?

So why change the discharge rate? Just change the battery reserves as you are seemingly already doing?

 

[Speedster]

I went with the default weather integration to determine if the pool pump runs or not. Rainy = no pumping ! I plan on tweaking this more at some point.

With regard to the battery discharge leveI, that's exactly what I do : The battery ( 14.3kWh ) runs down to the set min and then I tap the grid for the rest.

In the lighter months, the battery min is set down to 50% to keep some capacity for load shedding. During those months, we don't even use all 50% available, but it's there . We run for free. In the darker months it's as high ( or low depending on your perspective ) as 70% as a 30% charge is all the system can muster during these times. It's also like this because the power box down the road that feeds our entire block blows up what feels like every 3 or 4 months. Pseudo load shedding. The last time it shat the bed, that 70% saw us through 29 hours of no power and that included running the geyser.

That set min is exactly what I'm curious about optimising within an inch of it's life. On days like today, that set min could shift dynamically from 70% to say 65% and suck up all that sunny goodness, saving me a whopping R2.45 ! Hey it all adds up.

In truth I'm happy to wait. The set min will auto shift to 65% on 1st Sept, then to 60% on 15th. Assuming more sunny days in the next two weeks, it feels like I'm giving Eskom money for nothing...even if it is a pittance.

Nah, I aim to minimise grid use as much as possible. Batteries can run down to max if needed, no need to keep that much in reserve.

 

[hilt_ctn]

But battery percentage and discharge rate are two entirely different things?

So why change the discharge rate? Just change the battery reserves as you are seemingly already doing?

Ack :S I used discharge/SOC as interchangeable terms. When I said " limit its discharge", I meant limit the SOC so that the battery could only discharge x percent

So to be clear, I was thinking dynamically altering SOC not rate of discharge.

It's a nice idea that I felt an itch to explore ( there's probably a cream for that ! ). But I think being this dynamic when it's hardly required would be more trouble than it's worth. Well, for me anyways

 

[SauRoNZA]

Ack :S I used discharge/SOC as interchangeable terms. When I said " limit its discharge", I meant limit the SOC so that the battery could only discharge x percent

So to be clear, I was thinking dynamically altering SOC not rate of discharge.

It's a nice idea that I felt an itch to explore ( there's probably a cream for that ! ). But I think being this dynamic when it's hardly required would be more trouble than it's worth. Well, for me anyways

Yeah I think you’ll really just shift when you use Eskom a little bit and not really win out overall.

I pretty much just set mine higher by 5% to 20% for all of winter and then drop it back to 15% in summer when it almost never gets close anyway.